tl;dr: Director Severiano Martinez has been awarded a $15,000 grant to produce his film “Room Tone,” a chamber drama about an actress’s final session at an AI voice-cloning company. Production is expected to complete in September with a premiere this winter. Cycle 2 of svfilm.fund will reopen in the Fall.

The Creative and Technical Worlds

The technology and creative industries are two worlds that rarely talk to each other. They run on almost opposite logic. Tech runs on frameworks, abstraction, speed, and scale. The creative world runs on the human particular, the unrepeatable, slow contemplation, and complete attention to perceptive detail. One uses the cohort, the other, the singular individual. Velocity, versus waiting for “the thing” to emerge. They center in completely different geographies (SF vs LA), attract 2 completely different kinds of people, and have 2 completely different worldviews on what spending a life is for.

I spent the past decade as an engineer. Since the age of 16, I’ve been creating software, starting first with mobile development on Android, before moving to web engineering and eventually product engineering when I moved to San Francisco in late 2023. The only world I’ve ever known was the world of abstraction. Code, mathematical logic, the technically verifiable. The perceptive world, the world of felt sense, was something that was mostly invisible to me.

Studying Film

At the start of 2025, I felt something pulling at me to explore filmmaking. For a year before moving to San Francisco, I had been experimenting with creating cinematic shorts. I didn’t think much of it other than as a hobby. I became intensely interested in capturing pieces of reality and giving them a shape anyone could experience.

Coming from engineering, I struggled. Both with picking up the craft of storytelling, as well as the identity shift. Unsure what to do with me, I saw my former network of technical peers slowly fall away. I could not see it at the time, but there were 25+ invisible barriers ahead of me, or anyone attempting this transition, to overcome. I did not understand that it would be a long road (almost a year and a half of study) before I had an even basic understanding of how motion pictures worked on people.

Missing Structures

At the start of this year, I felt I was on more solid conceptual ground. Still struggling with the identity shift, I decided that there would be no other choice for me than to pursue this as my career. When I surveyed the film scene, the architecture that would make this transition more possible (peers, resources, opportunities), I found several essential pieces missing.

There was no serious space for working filmmakers in tech to meet, so I started a monthly film club. There was no development pipeline for scripted stories, so I started The Silicon Valley Film Fund. The first 2 months of the fund being open, I received no scripts, so I began writing my own. Throughout this period, still learning, acclimating, adjusting to a completely new field while building the pipes around me.

Cinema

The technology industry’s understanding of cinema, and most creative work that doesn’t contribute to product, is as a hobby (derogatory). Something completed in free time usually to signal competence or perceptive insight to go do the actual thing which is product-building. Movies are to be watched on weekends with popcorn and snacks. Recreation, leisure. This is one way to view it.

I see it differently. Cinema, the medium of sight and sound suspended in time, is the closest technological approximate in media we have to re-simulating human experience and moving it around with infinite scale. It is a permanent record of the world’s experienced understanding of a place and time.

In my original memo introducing SVFF, I cataloged the current cinematic canon defining the tech industry. The Social Network (2010) is the obvious cornerstone, with other films in loose orbit around. After a year of studying 200+ films, across eras and movements — Italian Neorealism to the French New Wave, New Hollywood to the Taiwanese New Cinema, Bresson to the Dardennes to mumblecore — I saw a massive gap in how the culture of the tech industry is represented. In fact, it’s almost entirely absent in its exact texture.

Stable References

This is a culture-scale or civilizational-scale problem, not an immediate individual problem. Quarter to quarter, year to year, in no direct way will it be apparent that the canonical film on early-stage startup life has never been made. Life will go on as it does. But something will be missing. During moments of moral crisis, private existential uncertainty, there will be no shared thing to point at that metabolizes that experience.

Films act as portals for these experiences. Stable references that a culture can gather around in conversation and contemplation. A film is more than what happens or who is in it, those are only circumstantial arrangements within the form. A viewership’s interpretations unlock an ecosystem of thought. A perpetual bridge opens for shared understanding. And the conversation lasts forever.

A film that metabolizes working class precarity? Rosetta (1999). The Dardenne brothers explore working class Belgium throughout their films (it was so influential that Belgium named a youth labor law the “Rosetta Plan” that same year). A film that metabolizes the relationship that can’t happen, real connection that arrives at the wrong time. Past Lives (2023). The full arc of ordinary life: birth, school, work, marriage, death. Yi Yi (2000).

What references does the technology industry have? Almost none. The Social Network is based in a courtroom, the films that need to exist are based in a Series B office.

SVFF

Around March, as my mental library of films grew, it became clear to me that institutional intervention was required. I was aware of only 1 film in development (Again, Tomorrow — Nikhil Ganesh, Rainbow PARC*) focusing on the culture’s interior. So I wrote a memo and posted an open call for scripts with a $15,000 grant attached.

Film as a medium has a substantial starting capital requirement. I wondered, if you removed even the start of that, who would surface with stories they were writing? I quickly found out that the answer was almost no one was writing stories from inside tech. In the first 2 months I received 1 submission that was topic-adjacent, with 13 other scripts coming in late-April and June.

Sifting through the scripts, I was pessimistic at first. There is a structural gap where those who best understand the experience of building in tech will not have acquired the cinematic formation (which takes years to a decade) to form. And those who have the formation and craft will be too far outside the world to render it accurately. But I read every script, line for line, with full attention. Returning handwritten notes to every writer.

After 5-6 scripts my hopes were falling. Had I made a mistake? The locations were wrong, the characters weren’t fully realized, the plot mechanics were missing, the dialog stiff and typed. I kept reading, 2 writers were promising, but one script stopped me in my tracks. “Room Tone” by Severiano Martinez. By page 10 I had a slight tear. It wasn’t based in San Francisco, but the tonal register, and its exploration of tech’s implications were spot on for what I predicted was out there when I started looking in March.

Meeting Severiano

We arranged to meet at Cafe Réveille on Waller & Steiner (ol’ reliable for Lower Haight / Hayes Valley 2023 to mid-2025 twitter people). I brought my original copy with notes, and a clean copy for him. When he walked up, he surprised me with 2 copies of his own. v1.4, printed single-sided.

There’s an immediate connection you can feel when talking to people who straddle the gap between the technical and creative realms. Conversation can take on the form of systems and the abstract, then shift into a completely perceptive register with sentences still somehow landing, no meaning or excitement lost. We talked about the story, the characters, the location. Everything was ready to go and he knew everything he had written, beyond what I had even anticipated.

It felt like 2 separate decades were meeting in the middle, at this very specific point in time. His in the creative world: working in LA as a director and moving up to SF to work on an AI startup in voice and generative worldbuilding (VYBE Labs). Extensive production experience with tech’s abstraction mode. And mine in the technical world: a career engineer, starting various things, working a few places, before ultimately being pulled to the creative realm.

By the end of the conversation, I was confident in the material and his ability to execute it.

A Catalog in San Francisco

My longer-term vision for the fund is a catalog of films that revolve around the geography of San Francisco, exploring the tech scene from multiple angles honestly. I am also interested in tech’s implications around the world, in family life, work life, identity (Room Tone exists here), and domestic matters.

The tech industry’s current understanding of film is as marketing collateral. The launch video, the VC podcast, fireside chats, developer experience videos, event recaps, product demos. What I want to show, is that its highest use is for reshaping culture and how civilization sees itself.

Paul Thomas Anderson shaped the San Fernando Valley. Edward Yang shaped Taipei’s boom years. Ozu and postwar Japan, Satyajit Ray and Bengal, Fellini and Rome. Who is that for San Francisco tech?

On Submissions

Tech is Eating the Creative World Alive

While reading scripts, on pen and paper, it became apparent to me that I was sitting in the same geography that was trying to make the creative industry disappear. After writing about what I look for in scripts, I realized that I had just written training data that took a year of intense watching to acquire for the next big lab model. For the first time in my life I felt the visceral estrangement from tech that the creative industry is feeling.

I began to ask myself what separated me, a fairly slow script reader, from an AI model that can already do a pretty good job at writing high-level script coverage? There are a lot of details on tone, performance, cast & plot architecture, and formatting a model might miss. A reader might have an awareness of the current cultural climate, memories of a specific geography and time. But these are all data problems that can be worked around.

Only original stories in the long tail will survive. Human fingerprints on them, personal stakes attached, director Q&A after. Selection will remain a bottleneck, a script reader/writer will need to know what’s good, and that only comes from reading hundreds of scripts and watching thousands of films. I never thought of it this critically, but the fund exists to fund just that. Stories on tech where the author is the only one who could have written it.

Submission Review

I want to encourage more writers in San Francisco to write these stories. If I could fund 4 films per cycle and had the funding, I would. But few are writing them, and the scripts that do exist aren’t yet ready for production. The infrastructure to develop them is nascent and proving itself.

Of the stories that were submitted (14), I found that…

Overall

Nearly every script was not ready for production, most reading as rough first drafts. Many shared anxieties about AI eating creative/knowledge work. 1 script was essentially ready to be shot (Room Tone), with 2 other writers who are early but very promising. Only 3 scripts were in SF proper (2 loosely indicated it, but were not direct). Only 1 script featured an entire cast of startup founders (other scripts adjacent, less focused).

Overwhelmingly scripts contained dense action lines, literary dialog, and 1-dimensional plot lines (axis of conflict). Generally, underrealized in the cinematic form. Dialog containing technical details struggled. AI was consistently an antagonizing force. A common failure mode was carrying too much drama through screens (Slack messages). A few scripts were complete outsider thriller narratives with no concrete grounding in SF.

Frontier labs are entirely absent. Research is almost entirely absent (other than 1 script with a few researchers).

Looking Forward

The goal of the fund is to offer a development pipeline for writers to take their stories from drafts to ready for production. Everything is so early right now that I am learning alongside the same writers I am reviewing submissions of. This first batch had a thematic focus of exploring AI’s implications on replacing creative work. I hope to see a wider set of themes in future cycles.

I am vigilant to keep a healthy mix of themes. I am also keen on staying anchored to the SF geography + characters (tech founders, early-stage, VCs, etc). I am not interested in the satirical register. I am looking for the sincere, grounded, restrained, and observational.

I hope more writers surface in the coming months and years as the fund progresses. It is incredibly early.

Cycle 2 will reopen in the Fall, scripts@svfilm.fund will remain open for submission.

Since I left engineering in February 2025, I’ve been filming in various offices around San Francisco to cover basic living expenses while studying film. Podcasts, fireside chats, feature announcements, etc. Random things popping up here and there. A year away from the industry and a time when I essentially lived in these spaces has given me enough distance and perspective to notice how strange they look and feel.

I’ve been carrying this feeling for a year, but recently, after coming back from a fireside chat I just filmed, it has become so strong that I wanted to look more closely at it. I will draw on my memories from interning at Twitter in 2019 (Market St), working as a product engineer in 2024, and filming in a few dozen offices around the city in the past year. For obvious reasons, I won’t be using any real images from spaces; instead I will create a new office layout and walk us through it, pointing out what you’d see and hear.

Space as Ideology

In the past year I’ve watched over 500+ hours of cinema, studying how filmmakers carefully design spaces to hold meaning. Spaces drenched in story information, designed to live forever, where every object means something. This has developed a perceptual sensitivity in me that naturally looks for meaning in objects, distance, light, color, how people move, where they exist in space, and what is omitted from frame. Nothing is by chance, everything holds a point of view.

Spaces hold an enormous amount of information. They constrict where we can move, what we can look at, what we spend time around. They encode an organization’s purpose and what individual time is for. These decisions become invisible the longer you spend in a space.

My Own Space

Over the past year, I’ve watched myself slip deeper and deeper into filmmaking, to the point that my own living space has been invaded by visuals and objects of the field. I could see my value system visibly realigning around me over many months as a function of where my attention has gravitated.

During this study period, some images carried so much density that I wanted to see them all the time. I began to hang them on the picture rail above me. Single images that I felt summarized a whole film or an idea a director spent exploring their whole career. Touching on a few:

Left to right:

• Nostalgia (1983): A wife, grandmother, daughter, dog, horse in front of a Russian farmhouse dissolving into fog. Tarkovsky’s contemplation on life on exile while in Italy.

• A Woman Under the Influence (1974): Peter Falk sitting in the back of a wagon with his children. A moment of calm and unity in an otherwise turbulent marriage. Cassavetes explores human chaos and aliveness throughout his films.

• Boogie Nights (1997): The final scene where Dirk Diggler (Mark Wahlberg) chants to himself in the mirror as the camera creeps in. I actually have this scene taped to my wall (last page in the script).

• Tangerine (2015): Outside Donut Time. A place no one but 1 filmmaker was willing to point a camera at to bring that world to screen.

Images extracted across 220 films. Filmmakers exploring deep, haunting themes where a lifetime of compression led to their creation. Media designed to be alive. Made to leave a residue on a watcher.

This space constrains me to stay inside this density of meaning almost all the time.

The Conference Room

My eyes began to wander. I began to notice how I couldn’t find anything in the room that meant something. Something with a story. Everything that functionally needed to be there was there. An L-shaped green sectional, 2 barrel chairs, a table with 6 faux-leather seats, carpet, 2 tv screens, sound bar, black linear suspension light above, plants.

Then various knick-knacks on the floating shelves:

But the room felt thin. I could not sense a distinct point of view no matter where I looked. Evidence of individual presence, discernment, preference. Everything pointed at meaning something, but didn’t mean anything in itself.

I glanced outside the glass walls of the room to the open office. Flags on the wall with company logo. Rows and rows of desks. Monitors. More plants. Sofas.

After the recording finished I walked out to the snack area. I looked and looked and looked. In the office of a company worth $20B+ I could find nothing that meant anything other than a few polaroids on the wall.

Production Design Challenge

I began to get curious. If I had to explain this space to a production designer, someone whose job it is to imbue meaning into an image through the design and presentation of physical space, how would I provide precise instructions? If I were writing a screenplay with precise spatial and visual detail on objects, what would I tell watchers they see? What is their ambient sense of being in the space?

Synthesized across visiting dozens of offices over 4 years, I have a clear answer to this question.

Office Layout

I’ve sketched 4 office layouts I commonly see. They all reduce to the same 10 areas, each serving a specific function. Offices have a:

• Reception Area: Guests are issued badges, packages are received.

• Coat-Hanging Area: A coat closet or hangar, clusters of shoes on the floor. Some offices are strict about not wearing shoes, some mixed. Positioned close to the open workspace.

• Open Workspace: Essentially a farm of desks in open space. Areas are usually grouped by department (engineering, sales and marketing, legal, etc).

• Meeting Rooms: There is usually 1 large meeting room for all-hands (holds >30), a few medium-sized rooms (holds ~10), smaller rooms (4-5), and small call booths (found in linear groups of 3-4).

• Open Work Area: Alternative to the open workspace. Sofas, coffee table, inlet areas where 4-5 can sit, work, eat. Furniture is more varied, seating is softer.

• Kitchen, Kitchenette, or Snack Area: Either a full kitchen with cabinetry, fridge, a 20’+ kitchen island w/ sink, or a smaller kitchenette with a snack cubby and a drink cooler. Offices may contain 1-2 kitchenettes, one on each floor.

• Lunch Area: Open area for lunch. In a 2-story office this is usually visible from the work area above. A serving counter with metal trays to keep food warm. Picnic-style seating, tables for 4-6. Doubles as event + all-hands space, sometimes with a set of tiered platforms leading down to the lunch area (where people can sit).

• Games Area: The “joy” section. Only in 1 location. Either a small cubby with board games, or freestanding games like a basketball hoop. Something to poke at between walking out of meetings or after 7.

• Bathroom(s): Usually 1 per floor.

• Quiet Room (optional): A silent room (like a library) where speaking is absolutely prohibited. For deep focus.

1 — Square Layout

The first layout is 1 area, shaped in a square. Meeting rooms on the edges and corners. 2 entrance/exits come in from the south. There is 1 kitchen, and a central lounge area.

I’ll use ChatGPT to generate quick mockups to get the rough idea across:

There is no reception area or quiet room in this layout. This is usually companies post-Series A, pre-Series C.

2 — Rectangle Layout

Contains a long open work area with exposure coming from windows on the left. Sofas sit opposite, a gym in the corner, bathrooms in the far back. A large all-hands meeting room on the opposite far end, with an event area adjacent. Storage in the corner.

3 — Loop

Everything we saw before, except in 1 continuous loop. Elevators located central. Meeting rooms line the inner walls. Offices at 1 Post Street are in this form factor.

One perspective from the SW corner looking east across the kitchen:

4 — 2 Floors (we will walk through this)

Finally, a 2-floor layout with the main work area upstairs and an open dining area downstairs. Soft work area adjacent to dining area, kitchen on each floor, soundproof library for focus downstairs, small cafe upstairs (right off the steps).

This is a composite of 4 offices, I’m not an architect, so forgive anything that seems nonsensical or breaks building code.

Starting the Tour

We Stand Outside

Facing glass double doors. A woman at the reception desk notices us. A loud click and long buzz. We open the right door and step in. The air is cool, light is even and sourceless. We walk over to the reception desk where 2 kiosks sit, stepping to the right kiosk.

We enter our information, a few pauses to glance around. We notice how smooth everything is and how few materials & textures make up the space. Maybe 4 materials and 4 colors. Concrete floor (2 shades of gray), wood on the reception desk (white oak veneer), white walls, and lights in long beams (silver linear pendants or long linear black suspension lights). Nothing on the walls other than vertical tv screens.

This color language will hold throughout the office, we will see a lighter cream color for seating (which we will get to).

2 badges print. The receptionist cuts them, hands them to us, walks around the desk, and asks us to follow towards…

The Right Entry Door

To the open lunch area. As we approach it, we begin to pay closer attention to what we hear. Above, a coffee machine hisses from the cafe. Straight ahead, 2 engineers emerge from the meeting room laughing, they briskly walk past us on the right. Behind us, a FedEx worker exits the same double doors we entered with a hand truck full of packages.

These sounds settle as the receptionist leads us through the door into the…

Open Dining Area

Where we stop for a moment to look around.

The space is mostly empty (it is slightly past 3), staff are cleaning up the buffet area, moving metal trays. 3 employees stand at the full kitchen in the corner, 2 chatting, 1 scanning the open fridge. 1 engineer sits in the open work area on a beanbag, beaming at his screen, laptop on his thighs, crunched up. 6 employees sit in the open seating area, 1 team of 4 still eating, then 2 individuals at different tables, each on their laptop.

Light chatter is heard from the 2nd floor. Looking west, we see 1 or 2 people pass through the glass safety rail. 1 engineer works on the sofa flush against the railing, we see the back of his head. The receptionist explains that the company just moved to this new office and is still orienting to the space.

As she speaks, we glance at the east walls. Next to the open work area, we see 4 pieces of 5’ x 6’ art placed eye-level, each 20’ apart. An image of downtown San Francisco, an abstract painting, a geometric piece with 2 shapes, and a spirited away piece with wooden edges.

Purchased from a gallery or wholesaler. No one in the office could explain what the expressionist piece means if you asked them. The receptionist begins walking us towards the library. We follow.

Library

The door shuts behind us and the silence is deafening. The carpet is soft under our feet. As the receptionist whispers “so…this is a quiet space…” we both continue to test the floor underneath us. 2 engineers working in the far corner glance up briefly, one wearing a silver AirPods Max. They return to complete focus on their screen.

We glance around. Warm light illuminates the shelves containing a sparse number of books and various succulents. Picture frames with important academic and pop culture figures hang next to the engineers working. The OpenAI library offers a current example.

Notice how this library creates the impression of being in a library, without actually sharing the same weathered accumulation of preference (this book not that) and time that you would see in a traditional library. The Rose Main Reading Room @ NYPL for example:

Something about our library lacks weight versus the “real thing”. The dust is missing, the air is thinner, the plants are too perfectly placed, the seating is creaseless, the books chosen but untouched. Something feels uncanny. We can’t pin down a specific consciousness staring back at us saying “look at this, this is a good thing, this is from me to you”.

Before we can think further, the receptionist hurries us back out of the room and turns us to walk up the…

Bleacher Steps

We pause. We look back down at the first floor.

Then glance to our right at 3 tiered platforms leading to the 2nd floor. Plants are found in a higher concentration and hang over a pair of teammates taking a meeting without headphones. By this point we have pinned down 4 primary plant species:

• Snake plant (Sansevieria trifasciata)

• Bird of paradise (Strelitzia reginae)

• Areca palm (Dypsis lutescens)

• Fiddle leaf fig (Ficus lyrata)

The color of the platform is beige veneer wood and beige seat cushions. A few more steps lead us to the top where we stop by…

The Cafe

For a breather. The barista hands a Swedish woman her cup, as she walks away the engineer next in line takes a small step forward before pausing to read the menu intently. The barista waits patiently, glancing at the counter. The menu has 6 items: drip coffee, espresso, latte, cappuccino, cold brew, matcha latte. The barista is a man.

We round the corner of the steps and step towards the east work area. In the far distance we see all hands is in session with about 20 participants in the room. Half of the room looks down at their laptops, the other half looks intently at the tv screens on the east wall. 3-4 participants nod gently as they listen.

To our left are 2…

Meeting Rooms

With clear glass walls. Inside room 1 sits 4 beige chairs in a barrel style, and in room 2 sit 4 beige chairs with a bouclé fabric surface (the fur) and open backs. In room 2 the walls are lined with wood slat acoustic panels, in room 1 the walls are brushed concrete. No one is inside either room.

We step further into the east working area.

Work Area

Exposure comes from above via skylights positioned directly over the open lunch area. We see 8 rows of desks, pairs facing one-another, each row 4 desk-pairs deep. Standing desks with white legs and light oak veneer tops. Only 2 in the area are raised to full height, all others stay down to seated height. Herman Miller aeron chairs alternating black and silver. Dark hoodies and light jackets are draped over 4-5 of the chairs.

Desks

On desks are Dell UHD monitors (ol’ reliable). We glance down to the row closest to us. Most desks are sparse with objects. The desk closest to us has a laptop stand, a bottle of caffeine pills, 4 ITO EN drinks (2 closed, 1 open and half-finished, 1 open and finished), a mac mini, and a capybara plush. O’reilly Desigining Distributed Systems and Designing Data-Intensive Applications sit under the monitor base, along with 4 other books. No family photos.

Hierarchy

We notice that the rows have no visible hierarchy. Hierarchy just moves into Slack and people’s private understanding. Who you ping for specific questions, which avatar you know has the answer for what, who goes quiet and how the quality of attention in the room changes when someone is speaking in meetings. Everyone knows where everyone else stands and does not say it. The flat, open layout gestures at egalitarianism, but human power structures do not change…

An engineer at the far end of the row we stand near looks over at us and smiles. Over his left shoulder, we see 4 Polaroid photos suspended on string and appended to wooden clothespins.

The engineer goes back to scrolling a JSON payload on the right of his screen. 6 terminals are open screen-left. The all-hands in the distance concludes and loud chatter fills the floor as people exit. The receptionist looks at the group, then back at us. She offers us a snack at the kitchen. We agree. We turn around, walk around the corner of the steps, past the games corner, past the cafe, and over to…

The Kitchen Area

As 5 participants arrive for a post-meeting snack and debrief. As we approach one of us needs to use the bathroom. A quick excuse me and a quick jog to the bathrooms on the left. Enter the male bathroom.

The lighting is fluorescent and awake. Surfaces are pristine and spotless. You relieve yourself and move to the left of 2 sinks to wash your hands. Between the sinks is a basket of sanitary goods. Lubriderm lotion, Listerine mouthwash, Dove Men+Care spray deoderant, Colgate Wisps. To your left, a Kleenex and diffuser reeds suspended in yellow oil. To the far right, Q-tips, mint green floss picks, and small mint packets.

Something about seeing sanitary goods in a public bathroom rings as strange (you only recall this happening in hotel rooms). You exit back outside to the kitchen.

The receptionist and I are making polite chatter. Rejoining, we move to the snack cubby. Companies with the best-designed products usually have the best snack and drink selection. Most literally a test of the founder’s or Chief of Staff’s taste. We scan the snack cubby to see where they stand:

• Wonderful pistachios (shelled): those are usually really expensive at the store, we each grab 1.

• Snyder’s pretzels: meh, those make my mouth dry

• Welch gummies

• Goldfish

• Beef jerky

• KIND bars, YES bars, CLIF bars

• Sahale Snacks

  • Banana Rum Pecans

  • Honey Cinnamon Cashews

  • Pomegranate Pistachio

  • Pomegranate Vanilla Cashews

  • The highest effort nut mix brand that exists. We each grab 1.

• Seaweed sheets: leaves a bad taste in my mouth, not for me until I give it a 2nd look

• Cookies & Cream and Caramel Cashew Barebell Bars

• GoMacro Organic, Plant-Based Protein Bars

• Doritos

• Kettle Brand Sea Salt Potato Chips

• Chocolate Pockys

3 winning snacks we would return to in a cubby of over 30 snacks.

With those in hand we turn to the fridge:

• Harmless Harvest organic coconut water: immediate winner, very expensive in the wild

• ITO EN: Organic Japanese Green Tea: these stay on the top shelf

• Spindrift (various): grapefruit, island punch, pineapple

• Waterloo (various)

• Poppi prebiotic soda (various): I will drink it because it is soda

• Olipop (various)

• Health-Ade Kombucha (various)

• Coke

• Diet Coke

• Gatorade (icy blue, red, orange, white)

• Fairlife protein shakes (chocolate, vanilla)

• La Colombe Oat Latte

• Celsius (various)

• Soylent: not in stable supply, but on occasion

• Bottom drawer:

  • Uncrustables: what my friends brought for lunch in middle school

  • cheese sticks

  • hummus

  • hard-boiled eggs (pre-wrapped)

There is no plain water.

As we finish our snacks and wash the remains down with coconut water, the receptionist says “this concludes our tour”. She leads us to the stairs, we toss out our trash, down the east steps, curl the corner, down the hall, back into…

The Lobby

The receptionist trails off towards the reception desk and we head to the door, pushing it open as a pair of engineers on a walking meeting slide past inside.

A rush of road noise and the cool San Francisco evening breeze. What immediately strikes us is the variance and apparent chaos of the outside world versus the one we just exited. We feel the sun on our cheek.

After a moment of adjusting, we begin northwest towards Market Street to catch a bus home.

End of Tour

Go to any serious startup in San Francisco and you will see the same office. The colors might change, the layout might change, textures may vary, materials may vary. But the underlying force that shapes how these spaces look does not change. The principle is not aesthetic. It is the minimum aesthetics required to keep bodies in space, sustained and focused, for as long as possible.

The Cocoon

I now understand it as a cocoon or a butterfly’s chrysalis. Just enough to keep your body seated and mind focused, but not so much that you are overstimulated such that your attention goes elsewhere, away from working (which is what aesthetic things do, they take your attention and return to you beauty, which has no legible output).

Structurally, most individuals in the office are as interchangeable as the plush beige sofa in the all-hands room upstairs. If you gave a moving team of 30 under an hour to clear both floors, they could do it. Every furniture item has the same “weight” and staying power of a Monobloc chair from your childhood YMCA. This is not a personal judgement, it is a conclusion the space poses to its inhabitants.

And an entire industry spends most of their adult life in these spaces believing they are “clean” and fully realized, when in actuality they point to being fully realized. This distorts the technologist’s sense of what aesthetic “good” is, because if you spend long enough in a space designed away from close scrutiny, where every object means nothing, your muscle for attention and aesthetic discernment collapses.

This has consequences in public and private life.

Thin Interiors

Redfin

Look at any Redfin listing over $2M+, where the San Francisco equity-owning class is shopping, and you will see the same sparse design patterns repeat. Let me grab one (I can’t guarantee staged vs. unstaged listings, although staging still reflects on buyers’ preferences).

875 California St #705, San Francisco, CA 94108

Beige color palette, the knot pillows, books that don’t mean anything, wall art that no one can explain nor anyone has personal ties to.

Barrel chairs, brushed steel lighting, plant.

A lobby that mirrors the library we walked through. Radiating with light and smooth textures, but ultimately saying nothing.

Corgi Cafe

A 24/7 cafe for builders (which is actually a really great service to the community) where costs were cut at every corner to suspend bodies in space to generate UGC on the timeline and a city news cycle.

A gust of wind could blow the furniture away. A moving crew could clear the place out in under an hour like it never even happened. There isn’t an ounce of love to be found in any object, just pure commerce. Pure ROI. And it achieves that nicely.

Dress

Taste Labs blew up the timeline in June with the goal of building a “taste layer for AI”. I haven’t been tracking the company or exactly what their offering will turn out to be. But I saw this quote tweet on the Taste Labs team photo captioned “tasteful team. join us” that was perfectly accurate:

Remember back to the color palette we saw when we first entered the office. A few simple, neutral colors. Smooth materials. Nothing that draws your attention. Nothing you could argue with. Nothing communicating any sort of individual sensibility or identity you could reject.

In the above photo, is anyone wearing a single item that couldn’t happen to belong to anyone else?…

To someone steeped in tech’s aesthetic language, this ironing out of character seems tasteful. This is just how things look. This is just how people dress… until you go outside the industry.

Balboa Theater

I frequently visit Balboa Theater, 1 of 3 theaters in the Cinema SF local theater chain. I began visiting in early January when I found the film selection at AMC Metreon 16 in downtown almost never overlapped my private study watchlist. It’s the classic commercial vs. aesthetic divide where AMC plays mass-market films with commercial objectives (and weak story quality) and Cinema SF plays films with cultural objectives (historical & craft-based) at the cost of finding a smaller, more devoted fanbase.

Initially, I was bothered by the smaller screen sizes, the sound system wasn’t as powerful, and the seating wasn’t as comfortable (no recliners). But the film selection kept bringing me back. The stories were just better. The films had more craft and actually mattered to people. Eventually the conceded amenities didn’t even matter.

Over the weeks, I began to notice how people standing in line for popcorn dressed. You could never guess the next person’s outfit. The shoes, the pants, the shirt, the jacket — all operated independently, as if each outfit was a unique organism that had shaped to the person individually. They took some time to parse and “read” initially, but once you understood what each outfit was doing it made complete sense. It felt inevitable for that specific person.

It seemed that “taste,” what the timeline had spent weeks torturing itself over…these people had it, they were doing it effortlessly.

Taste Arises Prior to Consensus

What are these people doing differently? What allows a group of strangers with no communication beforehand to each come to reasonable and aesthetically interesting outfit choices? Did they need to consult the X timeline for what taste meant? Did they search for the latest fashion framework? Did they look at what their peers were wearing and copy that?

No. They weren’t asking anyone about what felt good to them. They just did and wore what felt good. And new, creative, “correct” choices naturally arose from not even reaching the point of questioning their perceptive sense (they have good taste in movies, after all). Judgement about color, skintone, visual rhythm, materials, body shape — happened automatically. Constraints were triangulated to come to a decision. All options stayed open.

Most probably aren’t even on Twitter. Most have no idea what Taste Labs is or what the latest X discourse on taste is. They just have it, because they didn’t think their way there. They felt their way there. And how you feel is not up for consensus.

The Tech Industry Runs on Consensus

People in tech want to belong. A lot of us didn’t really fit in early in childhood, maybe we thought about things too deeply, were too inquisitive, were too quiet, were not good looking enough, or were just not sociable. This inferiority complex sits latent in most tech people, hiding behind a mask of ambition and skill-building. Some people genuinely love to build, for others it’s the only outlet that would work for them.

This ambition and technical skillset leads to a career that happens to pay handsomely. And real, genuine friendships can be built with co-workers. But the belonging the industry produces is contingent on maintaining a status quo with peers.

Writing the next definitive framework on X. Saying the smartest thing. Replying quickly. Signaling you were early to something that mattered. It is fragile and contingent. You are one step away from the outcrowd. It is tied to professional and financial consequences if you “get it wrong”. This is not real belonging.

In this social climate, when making aesthetic decisions you know your peers will scrutinize, the natural choice is to sand down what would make you unique. Get rid of friction the same way you remove friction in a UI tens of millions of users will use. But aesthetics don’t work like that. The point is the friction. The point is the choice you would make that no one else would. That is your whole identity as an individual.

Beauty

Once you see it, you can never unsee it. Every office you’ve ever been in, every tech event you’ve ever gone to, every peer you’ve seen purchase a black leather jacket — they all filter through the same framework-dominant mode that nods towards some consensus and invents nothing new.

In our imaginary office, those polaroids are specific people deciding that specific, unrepeatable moments mattered. So much so, they put them on the wall such that they, and others walking by, could replay those same moments. A warmth and a place and a time would transmit.

Everything else was placed on bulk order. There is no person behind it.

The tech industry doesn’t hate beauty, it just isn’t on the roadmap. It has no legible output you can point to on a dashboard.

It is a specific person transmitting life to another. An accumulation of choice, consideration, pressures, that produced something real.

Tech spaces have none of it.

tl;dr: The first cycle of the Silicon Valley Film Fund, running from March 10th to June 10th, recently closed. 13 scripts total were submitted. I cover the 14 things I look for in a script: strong character identity, protagonist identification, spatial awareness, sonic resolution, emotional awareness, dialog rhythm, control of pace, tonal control, formatting, premise & originality, axes of conflict, and opening & ending. Weaving in exemplary films and a sample from the hit movie Obsession.

I recently closed the first round of SVFF (the “Silicon Valley Film Fund”) which ran from March 10th to June 10th accepting scripts for short fiction films based on the technology industry. Over 3 months, scripts@svfilm.fund received 21 emails, 17 releases were sent, and 13 releases were returned. So 13 scripts are sitting on my dining room table waiting to be read after I post this.

I started the fund with a question: “if you remove the financial barrier to making a minimum viable film based in tech, what would happen? who would surface? who is writing stories and imagining this world anew?” I quickly found out the answer is almost no one. In the first 2 months of the fund being open, after 50,000 impressions on the announcement tweet, I received 3 emails, and only 1 was a tech-adjacent script. This sort of made me panic, was no one trying to create serious film work on the most consequential industry on Earth?

Afraid of the first cycle closing with no scripts, I began writing myself. For a while I’ve been seeing a specific character and specific scenes where they are saying and doing funny things and at some point while pacing around my place I decided to get Highland Pro (screenwriting software) and write them down.

At this point, I had only read 6 feature film scripts (Anora, Pulp Fiction, Boogie Nights, Sinners, Fruitvale Station, Moonlight), so I got the general gist, but that’s very different from having screenwriting grammar internalized and knowing how to write.

I spent 5 weeks through the end of March and all of April writing the first draft. Writing 2-4 hours every morning.

Before this, I’d only written a few (6) small 1-minute commercial scripts in January for a marketing campaign for @arlanr. So I just had to learn things as I went, reading a few more scripts (The Social Network, Punch-Drunk Love, Whiplash, The Breakfast Club, Cigarettes & Coffee 1993, Her) and watching a lot of YouTube videos on writing.

At no point was I sure that the writing was going anywhere, nor did I understand the full story/through-line until the last few pages. I was just following scenes that started as stems, extended them, and eventually they began to connect.

There’s a Stephen King quote from “On Writing” that I kept regurgitating to anyone during this time:

… during the course of an interview for The New Yorker, I told the interviewer (Mark Singer) that I believed stories are found things, like fossils in the ground, he said that he didn’t believe me. I replied that that was fine, as long as he believed that I believe it. And I do. Stories aren’t souvenir teeshirts or GameBoys. Stories are relics, part of an undiscovered pre-existing world. The writer’s job is to use the tools in his or her toolbox to get as much of each one out of the ground intact as possible. Sometimes the fossil you uncover is small; a seashell. Sometimes it’s enormous, a Tyrannosaurus Rex with all those gigantic ribs and grinning teeth. Either way, short story or thousand-page whopper of a novel, the techniques of excavation remain basically the same. — Stephen King, On Writing: A Memoir of the Craft (pg. 163-164)

Creative writing requires a non-linear mode of thinking that I’m entirely not used to coming from engineering. You carry half-formed character identities, motivations, locations, scenarios, for long enough a span of time until something appears. The best writing comes from long spans of doing what would outwardly look like nothing (going for a walk, a bike ride, staring at a wall) and coming back to the script with the problem solved and 3 more leads on where to go next. The process is entirely invisible and almost the exact opposite of engineering work, where output can be approached linearly & scheduled — fast, verifiable, visible, shareable.

Reading Obsession

Near the end of May (the 20th) I watched the movie Obsession (released May 15th). I didn’t think much of it and honestly walked out thinking the characters & story were thin and the writing very barebones. But Inde Navarrette’s performance shook me. There was a point while watching it, about 1 hour in, where I couldn’t really locate myself or my sense of place because her behaviour was so strange, so beyond what you’d encounter in a normal human being, that I had to do a double-take and really pay attention to get a fix on who this person was. I knew there was something special about that performance specifically.

The film is now on track to become one of the top #20 highest-grossing horror films of all time, standing at $286M after it’s 5th weekend (source: The Numbers).

Around this time, I had just finished reading the script for Her (2013), so I went looking for this next. I was curious how much of Indie’s performance was found on-set versus known on the page (and the answer is a bit of both). I found the script here. (many people are surprised to find out that the scripts for their favorite movies can be found for free online, and…yeah they are just online and you can print them)

Given the first cycle of SVFF has closed and I have scripts to read, I thought that this would be a good time to talk about what I look for in writing, using Obsession and other scripts from great writer-directors as examples.

What Makes a Great Script

As a disclaimer, I’m not a professional script reader or writer, but I do think I have a strong sense for a good story that moves me when I see it. From the 10 feature film scripts I’ve read so far, a short film first draft I’ve written, and 500hrs+ of cinema watched in the past year, I can pin what I look for down to 14 things:

• Strong Character Identity: The writer has fully realized the characters as full people in their heads and maintain that distinct identity throughout the script.

  • Character Backstory: The writer understands their characters’ full history, what went wrong, what went well, deep desires for the future, deep regrets of the past — the history that makes them, them.

  • Mannerisms: Whether it be a nervous tick, a specific way their eyes move between people, cutting people off mid-speech, or walking a certain way.

  • Physical Traits: Are they tall? What ethnicity are they? What does their face look like? How does the way they look make people feel?

• Protagonist Identification: When watching visual media, you will identify yourself with 1 or more characters as the film runs. This identification process can be deepened by the central protagonist doing relatable things: falling in love, getting abused, wanting to be accepted, fearing something. This anchors the watcher to the story because they see themselves in it somewhere. Stories that fail to complete this process early will struggle to hold viewers throughout the film. Stories that complete this process can do almost anything with watchers, make them feel jolts of pain, euphoric highs of finding love, etc. (and of course it’s not a hard rule to have a single protagonist…there really aren’t any hard rules…ever)

• Vivid Spatial Awareness: The best writers have a vivid sense of space. Where are characters positioned? How far are they apart? Which way is their body pointed? How wide and tall is the room? How large are objects? How fast is the car moving? What is the map of the city/neighborhood/courtyard? All of these will be resolved as if they were seen (but this does not mean that the words on the page are any more complicated).

• Sonic Resolution: Can the writer hear the setting where a scene takes place? What ambient sounds are heard? Does the room echo? What’s their character and texture? Do they know when music cues and fades out? This is more optional. Moonlight (2016) is an example of an action-driven, sonically precise script, w/ Barry Jenkins being a more poetic writer.

• Vivid Emotional Awareness: Throughout the story, who wants what? Who is on their back heels? Who is powerful? Who is hurt? Who is carefree? Who has a specific edge over who? Under what conditions would character X do what character Y wants? Under what conditions would character Y not do what character Z wants? The writer has a sense for every character’s inner world at all times.

• Dialog Rhythm: Does the dialog sound like the writer is trying to write, or does it sound like a conversation overheard? Real speech is messy and not step-by-step. Every exchange of words carries rich subtext that stretches far beyond what’s heard (hidden desires, power changing hands, backstory). The best dialog sounds like music, information speeds up and slows down, moving in waves. Bad dialog reads jagged and mechanical, with an odd meter to it. There are whole books just on writing dialog.

• Control of Pace: The writer has a strong sense for the “speed” of the story at any point in time, and can speed up or slow it down at will. Pacing interior to scenes and across the script are controlled. Stakes escalate cleanly across the entire story. The revealing of information speeds up and slows down instead of staying uniform.

• Tonal Control: One of the most difficult things to write and even more difficult to direct across a whole film. Does the writer have granular control over the tone watchers experience line-by-line? Scene-by-scene? Act-by-act? This is knowing the “taste on watcher’s tongues” at any point in the film during writing, directing, and editing. Parasite (2019) has strong tonal control; Obsession has it too. Takes a decade+ to internalize at the highest level.

• Formatting: Similar to “Control of Pace” but specific to how pages are formatted. Are sections sparse (reading speeds up)? Dense (reading slows down)? This changes how the picture unfolds in readers’ minds (actors, crew, collaborators, etc) and ultimately changes what ends up on screen. Every character has a story effect, down to the em dash “—” used in dialog or leaving out a period mark “.” at the end of sentences. This doesn’t mean perfectly typed sentences; it means carefully considering the effect of each character. The perfect formatting could lead to an “ugly” page that reads and paints the picture perfectly. Whiplash (2014) is basically a perfect script in formatting.

• Premise & Originality: Does the film create a new world that hasn’t been seen? What else does it look like? Is there direct imitation instead of original thinking (copied vs. synthesized w/ coherent logic)? Is this giving watchers something new? Something they need?

• Axes of Conflict: Conflict is what forces characters to make decisions under pressure, driving the story forward to a meaningful conclusion. How many “lines” of conflict can you draw between characters? Does it look like a complex web or a single straight line (one-dimensional)? Is the conflict profound and story-motivated? Or is it contrived? The best stories have several layers of rich and meaningful conflict. Flat stories have 1 axis of conflict that anyone could think of.

• Strong Opening & Ending: Does the ending land? Did every scene build to it and stay consistent to the through-line? Is the opening strong (relates to “Protagonist Identification”), visceral and cinematic (not necessarily violent or dramatic)?

• Commercial Viability: Can this actually be filmed? By this specific producer-director with this budget? A period drama in the 1800s is going to be very difficult to make because the whole world has to be built: costume, production design, hair & makeup.

• Heart: And most of all, does the writer have a personal stake in this story? Did they suffer in a particular way to give them this perspective? This is not to be hand-wavey; it concretely translates to whether a room full of hundreds of strangers, or a fast-scrolling timeline, with no one telling them what to think and feel, end up being moved and changed by the story. You can’t access people’s deepest, darkest memories unless you reach yours first and move them to the page. It seeps into every line of action and dialog, changes how you motivate collaborators and crew, and ultimately, who decides to finance the project. Great stories leave an emotional residue that remains long after the story has been consumed (good films that do this Hamnet 2025, Past Lives 2023, Y Tu Mamá También 2001, A Man Escaped 1956, Come and See 1985).

This is a long list, but most of it boils down to: did the writer see what they were writing, instead of writing to extend the page. What the writer could not see is left open-ended.

The Absolute Basics

It’s easy to forget how strange film is as a medium. You might think you’re watching real people going through real locations deciding what they are saying in real-time, as if you were there. But when you remember that it’s entirely a constructed illusion that feels real you begin to see the seams and get curious. If the image projected in front of me isn’t the actual thing, how am I constructing the world that feels so real? Who are these characters to me specifically? Why does cutting between shots work and not disrupt me? The illusion unravels.

This would completely sidetrack this article (and whole books have been written on what cinema is), but what’s really going on is you are bringing yourself to the film as you grapple to understand it. Characters merge with associations of people you know. Locations merge with childhood memories. Actions that don’t make sense are creatively interpreted. Every film plays as a unique experience in the hundreds of minds in the cinema, in unison on story texture, but experiencing completely different inner worlds. We will SIDESTEP this complexity to say…

A screenplay is a blueprint for describing how this new world flows in time. In the most economical way possible, it describes what viewers will see and what they will hear. The answer to “when” they will see and hear specific things is metered in the writing itself (or through action).

A script is made up of 3 building blocks:

1. Location: Where we are as watchers. (written in the “scene heading”)

2. Action: Events that happen on or offscreen.

3. Dialog: What characters say.

That’s it. These 3 building blocks allow you to describe the visual world that you are seeing/want to realize and transfer it to readers. There are no hard rules other than achieving that goal. Anything else about formatting can be searched, and learning proper story texture just comes from reading a lot of scripts and watching a lot of film.

In the above, we have 2 characters in a moving car at night. Nicky and Bear. Nicky is intoxicated. We are on notice to expect it to worsen as dialog proceeds. By the end of reading this snippet, you are in the car, seeing any one of the following:

1. Facing Nicky (who is sitting in the passenger seat) from behind Bear’s right shoulder. (camera behind his seat)

2. Facing Nicky from outside Bear’s window, in a 2-shot with both in-frame

3. Outside the windshield facing in, in a 2-shot with both in-frame

4. Close to Nicky’s face, then switch focus to Bear right before he speaks his first line

But those details are omitted from the page. Some writer-directors “hardcode” camera position & movement, like Paul Thomas Anderson, but most stick to the story (the essentials of what happens to whom and where).

An analogy I like to think of is that screenplay form is like an empty glass cup. It provides the rigid structure for holding liquids (audio-visual experience suspended in time), but torturing and studying glass shape and properties (subtext, compression, pace) doesn’t bring you much closer to putting juice inside it (having something unique to say, something specific and necessary to look at). This is ultimately the hard and rare thing.

Back to the Points

We’ll speed-run through each point in no particular order, moving between Obsession and other exemplary scripts to illustrate.

Scripts you’ll want handy: Obsession (2026), Fruitvale Station (2013), The Breakfast Club (1985), Moonlight (2016), Thief (1980), Rosetta (1999)

Vivid Spatial Awareness

Screenplay prose is dead simple. So simple a grade-schooler can read it (typically lands around grade 6-8 in reading level). It’s simple sentences chained together to paint a picture. Unlike a novel, which is the final artifact, a script only exists to communicate a film to the people who will make it.

Despite reading simply, there is a mountain of visual and temporal complexity that has been worked out sitting behind these simple phrases. And it’s often indistinguishable at first glance whether a writer has inhabited the world they are writing about with high resolution or whether they have written characters just to extend the page.

Fruitvale Station (2013) — Subway Car

A scene I find myself returning to to illustrate this is this scene from Fruitvale Station (2013), written by Ryan Coogler (script here). Coogler is a writer who writes simply but sees space with extraordinary detail. In the following scene, he tracks the position of 14 characters in a BART subway car as they cross to San Francisco for New Years:

After reading, you may see a layout roughly like this:

And the passage materializes in the film like this:

What follows is dialog that rests on this spatial foundation. Characters passing items to one another, looking in various characters’ directions, laughing and facial expressions in reaction. None possible without a clear idea of who is where.

Coogler would go on to win an Academy Award for Best Original Screenplay for Sinners 14 years later (which I also recommend reading).

The Breakfast Club (1985) — The Library

Another great example is this snippet from The Breakfast Club (1985), it’s a very economical film, using a cast of 5 students caught in high school detention held in the school library (1 location). Despite being lean, the film’s power comes from cast design and performances.

After reading, you may see a layout roughly like this:

And the passage materializes in the film like this:

Tonal Control

One of the main tasks of the writer at writing time, and the director on set, is to understand the tonal texture of what they are writing/capturing at any given moment, both local to a scene and global to the story as a whole. This invisible emotional texture is like a quilted fabric that runs across the entire script, changing scene by scene, act by act, line by line.

Strong writers/directors see the whole quilt (or very large regions of it near to where they are) and can project the effect of their words, or an actor’s performance, on the viewer's emotional state at any given point in time. What are they anticipating? Where are they confused? How long can they stay confused? What do they fear? Where should things speed up? Slow down?

Obsession is a quilt that depends almost entirely on 1 actor’s performance, Nicky (played by Inde Navarrette) escalating in intensity across acts and varying more and more wildly within scenes themselves. The most signature scene in the film happens when Bear confronts her about her Dad allegedly having cancer; it turns out she had lied about it.

After reading the script, it became clear that most of Nicky’s performance, in concept, was on the page, and only magnified with Inde’s interpretation & Barker’s control in direction.

The film is essentially this same tonal whiplash tightened and tightened until it becomes unbearable to watch by the end. Strong tonal control in writing and a great achievement in direction. Essentially a chamber piece study in controlling tone, at least on one dimension (fear), in the horror register.

Sonic Resolution

When sonic details become significant enough to the story, you will see writers write them in. A good example is Barry Jenkins in Moonlight (2016). In the following passage, the main protagonist “Little” is being chased by a group of neighborhood boys where he takes shelter in a crackhouse for cover:

And the passage materializes in the film like this:

This ties into Protagonist Identification (which we won’t individually look at), but you can see how by page 3, we are bonding with the character and identifying with their abuse, wanting them to be rescued. We become them for the rest of the runtime.

Sonic detail in this high a resolution in writing is not mandatory since readers can usually fill in the gaps, but for this scene specifically, it was exactly what was intended to be shot, so it had to be written with such precision.

Dialog Rhythm

Film is similar to music in that both are time-based mediums, susceptible to the rhythms and properties of story information being interpreted moment by moment over a fixed span of time. Once image and sound are captured and played back, their consumption takes on a natural cadence (in effect regardless of whether you notice it).

A script can be thought of as a score for life.

Dialog is a compressed form of speech. Regular everyday speech is too inefficient to translate directly to a film/script, since every block, every sentence, of dialog is ideally doing 2-3 things:

• revealing character (how someone talks reveals who they are)

• advancing plot (information exchanged, decisions made)

• establishing relationship (power dynamics, history, intimacy)

• creating subtext (what’s not said, what’s said wrong)

• setting tone (the emotional register of the world)

• building or releasing tension

• delivering theme (indirectly)

Speech Rhythm In Everyday Life

In an interview, director Ryan Coogler said that he likes to listen to the way people talk in public or when riding on the bus, so that those natural cadences help inform his writing.

I’ve begun to do the same, and I now naturally find myself in public listening closely to the cadence of how people speak, observing how light falls on their face, what is revealed about relationships when the speech stops (body language, distance, facial expression). A lot of story information moves non-verbally.

The work of Robert Bresson (simplified performances) & John Cassavetes, particularly Faces 1968 (as a canonical example on communication through pure facial expression in close-up), have been very influential on this practice of observation. I think watching so much film just trains you for this mode of attention in day-to-day life.

Thief (1981) — Car Ride

There are more overtly lyrical writers like Quentin Tarantino that push their dialog towards a musical rhythm (the “foot massage” dialog from Pulp Fiction for example). But a director I particularly like that mixes dialog control with sharp performances is Michael Mann. Here are a few passages from his 1981 film Thief, where he captures powerful but measured performances where the dialog finds its own meter.

First, an argument in the car between Frank (James Caan) and Jessie (Tuesday Weld) in the car:

Which materialized in the film like this:

Step by step:

• Jessie’s non-replys create a vacuum that needs to be filled

• The pressure of Frank’s desire builds, begging for a reply

• Jessie’s reply relieves some of this pressure momentarily

• Frank crescendos for the signature “I’m a thief.” line

• Jessie crescendos (a bit after this clip) saying “so what, I don’t care”

Pressure builds gradually until it erupts, an enormous change of charge in under a minute, under careful control. Two characters who talk past each other the whole film, Frank fundamentally unable to give Jessie what she wants, and Jessie unable to receive.

Thief (1981) — Meeting Leo

Another more lyrical scene, Frank meets Leo, a mob boss:

Which appears in the film like this:

The back-and-forth is much faster, pinging between the two, Leo with quick replies and Frank more measured and to himself. Despite speaking quietly, the emotional charge between the two is enormous. Leo is a dangerous person.

For a 3rd example there is the adoption agency scene which crescendos smoothly as well, but will omit for copyright purposes as it is a longer scene.

Heart

The last point I’ll touch on in detail is whether a script/film has “heart” behind it. Does the writer/director/crew have a personal stake in the story they are telling? Is something at risk, is there exposure in the telling of it? This may seem optional, but the best stories all have this. You cannot move an audience of strangers sitting in a dark room with no one telling them how to think or feel unless you put something profoundly true in front of them without hiding.

In his book “Sculpting in Time” Andrei Tarkovsky (one of the greatest directors of all time) writes:

… Artistic creation demands of the artist that he 'perish utterly', in the full, tragic sense of those words. And so, if art carries within it a hieroglyphic of absolute truth, this will always be an image of the world, made manifest in the work once and for all time. … These poetic revelations, each one valid and eternal, are evidence of man's capacity to recognise in whose image and likeness he is made, and to voice this recognition. Moreover, the great function of art is communication, since mutual understanding is a force to unite people, and the spirit of communion is one of the most important aspects of artistic creativity. — Andrei Tarkovsky, Sculpting in Time (pg. 39)

(continuing on…)

… Art is a meta-language, with the help of which people try to communicate with one another; to impart information about themselves and assimilate the experience of others. Again, this has to do not with practical advantage but with realising the idea of love, the meaning of which is in sacrifice: the very antithesis of pragmatism. I simply cannot believe that an artist can ever work only for the sake of 'self-expression'. Self-expression is meaningless unless it meets with a response. For the sake of creating a spiritual bond with others it can only be an agonising process, one that involves no practical gain: ultimately, it is an act of sacrifice. But surely it cannot be worth the effort merely for the sake of hearing one's own echo? — Andrei Tarkovsky, Sculpting in Time (pg. 40)

(finally…)

… True artistic inspiration is always a torment for the artist, almost to the point of endangering his life. Its realisation is tantamount to a physical feat. That is the way it has always been, despite the popular misconception that pretty well all we do is tell stories that are as old as the world, appearing in front of the public like old grannies with scarves on our heads and our knitting in our hands to tell them all sorts of tales in order to keep them amused. The tale may be entertaining or enthralling, but will do only one thing for the audience: help them pass the time in idle chatter. — Andrei Tarkovsky, Sculpting in Time (pg. 188)

Looking back on it, I aggressively underlined almost every sentence in the book, it’s just that good (and I’m not aware of any other book that contains a stream of thought so detailed from such a prominent director in history, it was written while Tarkovsky was on exile from the USSR in Italy, Sweden, and France between 1982-1985).

Rosetta (1999) — Whispering Hope

I’ve recently started going through the Dardenne brothers filmography (Luc Dardenne and Jean-Pierre Dardenne), and recently watched a film I really enjoyed called Rosetta (1999).

In it, a desperate teenage girl claws her way through dead-end jobs in a deindustrialized Belgian town, dragging her alcoholic mother along with her. The Dardennes made it from the inside, from years of documenting working-class Belgium, and you feel that weight in every frame.

The most moving scene in the film happens when Riquet, a waffle wagon worker who becomes Rosetta’s only real human connection, lets her sleep on his cot in a spare unit:

Which appears in the film as this:

There is a claustrophobia to the space and an intimacy to seeing someone chant encouragement to themselves in bed. It’s the character at full exposure, and the audience has seen enough of her getting abused to fully embrace her as themselves.

The best films aren’t trying to make a point, they’re trying to make someone, or some circumstance in the world, seen — in such a way that people can find themselves in the characters. This requires a connection to the material that is personal, so you can reach the personal in others.

This is the core reason why cinematic storytelling on the technology industry is non-existent. Aside from a massive creative labor shortage & enormous financial incentives pulling creative talent away — it requires direct contact to human experience without abstracting it into frameworks or meaning. The people with access to the experience can’t hold it raw. The people who can hold experience raw don’t have access.

The Rest

I won’t belabor the rest of the points, so quickly:

• Strong Character Identity: Read The Breakfast Club (1985), it will teach you everything you need about cast design. Every character forms around a distinct archetype found in high schools that everyone can relate to (jock, academic, outcast, badboy, pretty girl). Great cast design has individual and highly functional characters. This leads to strong identities where characters don’t sound like they could be “anyone,” they sound like real people.

  • Through specificity, the story becomes about much more than high school. It becomes about the search for identity in youth, the struggle against authority, performance of identity under social pressure, parental damage, and the limits of empathy.

• Formatting: If you scroll this article, you will see portions where your reading slows down, and you have to pay more attention (like the 14-point bullet section) and portions where you can fast scroll past (the video samples). The same goes for reading, when dialog exchanges thin out and go vertical on the page, your eyes speed up, and the image speeds up. Thick walls of action lines slow you down and force environmental context to get built. Same goes for dialog. Whiplash (2014) is a good script to study for format and character economy.

• Vivid Emotional Awareness: Similar to tonal control but focused on character motivations and internal states instead of watcher experience. Is the writer writing from inside their character’s deepest fears and wants? You’ll be able to sense it.

• Control of Pace: Stories speed up and become more intense as they near the ending. Pace is accelerated. “Speed up” doesn’t necessarily mean frantic action or faster events (it can mean that); it just means more happens within the characters and with more charge, such that reading speed quickens and the reader is eager to reach the end. Paul Thomas Anderson does this well across acts (see Magnolia 1999).

• Premise & Originality, Axes of Conflict: Many indie films I watch have a single axis of conflict and a weak premise. Just because it happened in your life and it generalizes to others, doesn’t make it a good story (i.e. family issues, childhood memories). These stories can work, but the characters just need to be compelling, performance compelling, something has to be different about them. If the chart of character conflict is just 2 bubbles and 1 line, or 4 bubbles and 2 lines, the story isn’t good.

• Strong Opening & Ending: See Whiplash (2014), Kill Bill: Vol. 1 (2003) (or any Tarantino film), Anora (2024), Y Tu Mamá También (2001), endlessly many more

Writing Better Films on Tech

I just wanted to write a post to collect my thoughts on a criterion that had been invisible to me until writing this, and that I will use for SVFF’s first cycle while reading. It is also the measuring stick I use to assess story quality for non-scripted stories (documentary, social content, etc).

So when I say “we need better stories” I’m running through the tech film canon, which contains nearly 0 films from the interior, and thinking point by point. What is the quality of the stories (cinema) made from inside this culture? And the problem is, there are few to even assess quality on.

tl;dr: There are <30 active filmmakers creating inside the technology industry right now. Versus 200,000+ software engineers in the Bay Area. There is only 1 feature film in development from the culture’s interior. Tech, one of the most consequential and complex cultural forces in the world, has almost no narrative storytelling capability in film. I map the first steps to solve this problem.

A filmmaking scene is emerging inside the technology industry. It’s still small and fragile, but there’s a group of serious people who want to make a career at the intersection of tech and film. This couldn’t have happened pre-2023, but the cultural room now exists for it.

The following is a list of studios, media operators, writers, directors, producers, & creatives who are creating at this intersection. I’ve gathered it over 3 years of doomscrolling X and texting myself any time I see a film set or film-adjacent person (and also from running a film club in San Francisco).

The directory is organized into 3 sections: studios (named entities), individuals, and infrastructure. The goal is to provide clarity on basically everything going on. When you can see the edges of something, then it can grow.

Studios

Narrative Lean

Building towards making fiction films.

Story Company website @storyandscience instagram

• by: Jason Carman (100+docs w/ S³, 2 feature-length docs, feature-length trilogy + book adaption in-development) (@jasonjoyride)

• north star: creating feature-length science fiction films

• commercial work: cinematic + docu-style spots for deeptech companies

• clients: deeptech companies and institutions

RAINBOW PARC* website @rainbowparc instagram

• by: Nikhil Ganesh, Krish Mysoor, Neil Thekdi (@nikhilganesh001, @krishmysoor, @producer_neil)

• north star: fiction films based in the technology industry. “Again, Tomorrow” filming 2026.

• commercial work: cinematic spots for early-stage startups

• notes: one of the only studios I’m aware of creating fiction from the industry’s interior (via the team’s South Bay background). Nikhil Ganesh is a director to watch.

Commercial Lean

Building towards strong commercial execution.

Offscript website @wentoffscript instagram

• by: Juliana Glodek, Alli Gooch, Cameron Dennis (@juliazniv, @alli_gooch, @docdropout)

• commercial work: cinematic storytelling for startups/tech ++ original series, shows

• clients: early-stage startups & media operators

Osmo website @osmo_studio instagram

• by: Will Hoppin, Basil Alfaro (@willyhopps, @basilvalfaro)

• commercial work: cinematic storytelling for startups

• clients: early-stage startups

Hard Launch website @hardlaunching

by: Jacky Huang, Rylee Stanton, Lucas Quan (@Jackyhuang, @rylinaz, @lucasquan)

commercial work: launches for startups (have worked w/ Eigen, Flora, Listen Labs)

clients: early-stage startups

Afterimage (nyc) website @afterimage

• by: Josh Lee, Cyrus Duff (@lee94josh, @cyrusduff_)

• commercial work: creative + film for startups

• clients: early-stage startups

Really Good Videos website @reallygoodvids

• by: Ben Aguilera (based in spain) (@ben_aguilera)

• commercial work: feature + product launches (have made banger work for raindrop and others)

• clients: startups

Skale website @SkaleSolutions

• by: Mark Vassilevskiy (@MarkKnd)

• commercial work: feature + product launches (some film pieces, usually motion spots)

• clients: startups

The Launch Video Company website @subahwadhwani

• by: Subah Wadhwani (@subahwadhwani), runs Atomik for podcasting

• commercial work: launches + podcasts (Atomik)

• clients: startups + venture

Alto Visuals (la/nyc) website @altovisuals instagram

• by: Loucas Tsilas (nyc), Andres O’Beirne (nyc) (@LoucasTsilas, @andresobeirne)

• commercial work: launch videos (did launch for Cluely, Poke)

• clients: early-stage startups

Camp Studios (Creator Camp) website instagram @creatorcamp_

• by: Max Reisinger, Simon Kim, Chris Duncan (@Maxbrsee, @thesimonkim, insta:@chris_dunc)

• commercial work: full-stack creative

• clients: startups, global brands

Native Four (la) website instagram

• by: Dilan Mistry (insta:@dilanmistry)

• commercial work: commercial + branded

• clients: global brands (outside of tech), EP w/ Offscript

Individuals

This section will be organized into 11 groups: directors, writers, producers, creative directors, in-house, independents, head of media @ company, X “influencers”, operators, venture capital, and those close/adjacent. Within categories, names will be organized alphabetically.

CD = creative director

* = repeating from studio section

Directors

Arvin Farsyad (toronto) website @arvinfarsyad instagram (w/ Simply Living Studios)

Donald Jewkes website @donaldjewkes instagram

*Jason Carman (tech × film origin point, ’23) website @jasonjoyride instagram

Joel Yoon website @joelyoon_ instagram (operating under dogeared)

Joshua Bonzo (nyc, creator, director) website @wordisbonz instagram (operating under The Editor Company)

*Nikhil Ganesh website @nikhilganesh001 instagram

Souvid Datta (london, some work w/ @fiftyyears) website @souvid instagram

Steven Xiao website @stevenxiao_ instagram (operating under Montlake Studios)

Writers

Austin Peirson (writer, editor) website @AustinPeirson (Austin has the early intangibles of a strong writer and strong editing skills, potentially moving to LA soon)

Shani Zhang (artist) website @sha_zng instagram (vivid cinematic sense, example, think she’d be a great screenwriter, currently draws part-time)

Producers

Adam Miller (Sachin & Adam) @_AdamMiller_

Adil Mania (Silicon Mania) website @adilmania instagram (one of the first to take X-first shows and formats seriously, very show-oriented rather than cinema)

Arielle Swedback (nyc — Substack) @aswedback

Ben Slevin (A16Z) website @bnslevin instagram

Benyam Ephrem (writer, director, producer) website @bephrem instagram

Courtney Bankhead (Cheeky Pint @ Stripe) website

Ebaad Rehman website @0xEbaad (started capi, now side questing — I believe it petered out, the premise was great, just lost momentum, imo should still exist, maybe not the podcast form factor but something more dynamic like docs)

Eliott Mogenet (Expresso Studio) @eliott__mogenet

Gabriel Ramans (Silicon Mania) @gabrielramans

Hunter Weiss @Hunter_Weiss instagram (WE!SS studios, serving global brands + startups)

James Lin website @jlinbio (research/deeptech lean)

*Juliana Glodek (head of studio) @juliazniv instagram

Kenny Tjay website @kennytjay instagram (video and podcast producer)

Randa Sakallah (Substack) website @mostlyvibing (has really good story sense)

Rob Giani website @robcgiani instagram (cinematic deeptech lean)

Sachin Shah (Sachin & Adam) @Sachin_Shah_

Tamara Winter (Stripe) @tamarawinter (2 episodes of “Tacit,” they did well, then series disappeared. maybe budget / ROI issues, fate of most good productions in tech when twitter fervor dies down)

Creative Directors

Alec Bahta (nyc, Whop) @quietluke instagram

*Alli Gooch (la) website @alli_gooch instagram

Jeremiah Warren (la) website @jeremiahjw (looking for role as CD in tech, based in LA)

Nefise Akcakir website @nef_isa instagram (visiting sf, working on move from canada, either director or CD)

Reggie James (General Catalyst) @HipCityReg instagram (new at GC)

Richard Zheng (A16Z) website @richardzphotoz instagram

Yacouba Gnacko (Alif) @YGnacko instagram

In-House

Aaron Bisla (Dedalus) website @44ron44 instagram

Aiden Blumenstein (Founders Inc) website @super8n (strong cinematic sense)

Arman Ajoomal (Core Memory) website @armaanajoomal instagram

Hugo Ahlberg (Opal Camera) website @HugoAhlberg instagram

Sunday Ajak (The Residency) @sundayajak instagram (head of marketing)

Independents

Abhz (australia) @Aabbhhz

Alex DeNuzzo @xdenuzzo

Brandon Lu (nyc, creative producer) website @brandonjlu instagram

Chloe LeValley website @ChloeLevalley

Daniel Kwon (nyc, prev. Wabi) @dantioxidants instagram (consulting now)

Edmund Tian @edmundtian instagram (AI filmmaking)

Eliott Mogenet (writer, director) @mogenetEth instagram (I saw a production a while back but now I think running a startup)

Ivan @leomeethewoo (operating under Intently, fintech launches)

Jaden Kwan @findingjaden instagram (operating under Tastefully)

Jordan Mitchell @byjmitch (operating under New Renaissance)

Nick Gerwitz (cinematographer, editor, producer) @nicholasgerwitz instagram

Shubham Tuteja (delhi, CD) @ShubhamTotu instagram

“Side Quest Ryan” @sidequestryan instagram

Thomas Burton website @tomredhairmedia (operating under Red Hair Media)

Vikrant Patankar website @vikpat instagram

Will Phillips website @willsclips_ instagram yt (moving stories to the “outside world,” very consistent at cinematic quality, essential work so this moment is not lost to time, analogous to Jason Carman’s work on S³ except software-focused vs deeptech)

Head of Media/Product/Comms/Growth/Strategy/Brand/DevX @ Company

Adel Wu (Reducto, HoG) website @adelwu_ instagram (animation, illustration)

Brian Kinnes (Whop, HoM) website @briankinnes instagram

Dar Sleeper (1X, HoP) website @radbackwards (first brand to consistently produce serious film work in tech)

Denton Baird (Zipline, Head of Comms) website @DentonBaird (zipline is doing very solid work on comms)

Fynn Comerford (Until Labs, Head of Strategy) website @FynnComerford instagram

Hannah Ahn (Superpower, HoM — OG startup branding person ’23/’24) @hannah_ahn instagram

James Reina (New Media @ A16Z, Growth, nyc) @thejamesreina (prev. Mr Beast)

Julia Fedorin (Composio, HoM) @juliafedorin instagram (very content forward)

Liam Hinzman (Exa, HoP) website @LiamHz instagram (illustration, 3D)

Louie O’Connor (Ethereum) @lou3ee instagram

Manveer (Prime Intellect, HoP) website @manveerxyz (PI one of the few companies doing brand well, distinct & immediately recognizable identity)

Sarah Chieng (Cerebras, DevX) @MilksandMatcha (Big Chip Club, has done a film screening or two)

Steven Pham (Y Combinator, HoM) @svkpham

Tiago Sada (Tools for Humanity//World, HoP) website @tiagosada instagram

TJ Krusinski (Content @ Chroma) website @tjkrusinski

X-fluencers

Anu Atluru website @anuatluru instagram (cultural commentator/critic)

Cole Lee (“covacut,” creator) website @covacut instagram (dropping new tech × art intellectual frameworks regularly)

Jayden Clark (MOTS pod) @creatine_cycle instagram (longform pods documenting sf underground, very good at the saxophone)

Jenny Gao (“50,000 Art Hoes”, australia) @taotechic instagram (very perceptive, writes screenplays I think)

Maja Wilbrink (Narrative Strategist) website @majamediaco (extremely talented writer, extremely strong narrative & story sense, moving from australia to sf soon)

Paula Dozsa @paularambles (likes film)

Scott Belsky (partner @ A24) website @scottbelsky

Operators

Patrick McGuire (CEO @ Soon) @patrickmcguire

Eoghan McCabe (Intercom, Chairman @ Soon) @eoghan

Venture Capital

Brent Liang (New Media @ A16Z) @liangsays instagram

Brett Berson (FirstRound) @brettberson (film@nyu)

Claire Zau (Partner @ Lightspeed) @clairejyz instagram (explaining tech and vc moment by moment in plain language)

Drew Moxon (Head of Platform @ 50y) @moxon (prev. interactive media @ usc)

Erik Torenberg (New Media @ A16Z) @eriktorenberg

Erin Gleason Lane (CCO @ Founders Fund) @eag

Josh Machiz (CMO @ Lightspeed) @Machiz instagram (thesis: venture should produce content & real film work, mindshare, etc etc — obvious, no one is doing it well)

Julian Weisser (ODF) @julianweisser (cinephile)

Lester Chen (sf/la) @Chen (very strong sense for story, commissioning work for speedrun)

Monica Lim @mon__lim (has done a film screening or two)

Shari Doherty (prev Lucasfilm ’97-’00) @sharidoherty

Adjacent

Chris Samra (Waterloo) website @crsamra

Elizabeth Qiu (University of Maryland) website @elizqiu instagram (moving to sf soon I think)

Karina Bao (photo/film) @repkarinabao (has taken still photos for SF Alexandria, @aadilpickle blog)

Max Prokopp (prev. producer) @maxprokopp instagram (in south bay // doing research I think)

Notes (Studios + Individuals)

I never really saw the full picture until sitting down to write this list. Looking on it now I see:

• 0 professional writers: Almost no screenwriters from inside the culture. Austin is working on moving from Canada to SF, and Shani works in art (might begin writing). But no one is formally trained. Without writers, stories can’t get made.

• 2 professional (culture-interior) narrative directors: Nikhil Ganesh + Jason Carman. Donald Jewkes has all the intangibles and the strongest interiority, but isn’t writing anything right now. Jason Carman is working on a trilogy, but won’t be addressing culture-interior sf tech stories.

• 1 culture-interior feature-film in production: RAINBOW PARC* is the only studio coming from inside Silicon Valley currently producing a film (“Again, Tomorrow”) on the tech industry’s interior life. To me, this should be 3 or 5 studios.

• 2-3 serious, career producers

• 5-6 creative directors

• various independents doing independent things

• various heads of media with sense

• some X influencers, mostly commentary

• 1-2 serious venture firms

• then others adjacent

Tech, one of the most consequential and complex cultural forces in the world right now, has almost 0 narrative storytelling capability. And has fragmented & weak labor depth in almost every creative category film requires. Almost all storytelling activity is commercial-aligned vs. culture-aligned.

Infrastructure

Infrastructure is everything surrounding individuals to allow them to be productive. This includes access to 5 things: capital, community, studio space, festivals (to showcase work), and distribution (publications, critique).

* = things I have started

Capital

*The Silicon Valley Film Fund (svfilm.fund): A $15,000 grant for narrative filmmakers creating stories on San Francisco tech’s cultural interior.

OSV Fellowship (website): A one-year program for researchers, builders, and creatives advancing civilization. Grants up to $100,000 (Jason Carman awarded in ’24).

Future Vision XPRIZE (website): Looking for 3-minute sci-fi film trailers portraying a more abundant future (“show us a future worth building”). $3.6M in prizes total, $2.6M to winner to make a feature film, $1M to runner-ups.

Notes: The main bottlenecks right now are clarity on what stories should exist and supply of those stories (re: writers, directors), not capital.

This clarity and the cinematic training take years to acquire. It should just be tablestakes that those with a vision and clear plan should be given capital to remove that barrier (given the story shortage and capital surplus).

Community

Once talent is concentrated in a certain geography, there needs to be physical meeting points for people to convene, share ideas, and see that others are doing the same thing (otherwise people will get discouraged, leave, or quit).

Ideas need others to bounce off of and amplify. As a software engineer in SF this comes for free, everything is building, conversations about it are everywhere. If you open your X timeline, 90 out of 100 posts will be about investing, mental models, or fundraising/user milestones. For those creating stories, a new infrastructure has to be created.

*SF Film Club (Luma): A small salon series for filmmakers creating at the intersection of tech & film. Cinephiles + full-time filmmakers.

Design Film Club: Monthly screenings crossing over w/ Design Eng Club. (a loose collective runs this, have seen @mathurahravi, @ftnabeelah, @tinathetechie, @floguo involved).

New Systems: By Vin Verma (@internetvin), a 3rd space in Toronto (83 Walnut Avenue) for technologists & artists w/ a mix of tech & media events. Also create films.

TIAT: By Ash Herr (@imempowa), creating a home for creative technologists. Gallery at 151 Powell St.

Localhost: By Kei Hayashi, Suhas Sumukh, & Hardeep Gambhir (@KechoHayashi, @suhasasumukh, @hardeep_gambhir). A startup accelerator + media effort w/ 4 hubs around the world (Bangalore, Tokyo, Paris, SF launching soon).

Machine Cinema: By Minh Do (@minhsmind), exploring AI filmmaking’s intersection with traditional cinematic storytelling. Meetings primarily in LA (Luma).

Nautilus: By Zelda Poem (@zeldapoem), pioneering the patronage model in ’24 for artists in SF in a 3-month accelerator format. Not film but adjacent.

Notes: SF Film Club is the only space in the city that working filmmakers in tech with serious cinematic ambition can convene. There is opportunity for a larger community center like TIAT (151 Powell St) to exist just for film, as a screening and meeting space (similar to what the Cinémathèque Française was during the French New Wave). I don’t think the community is large enough for that yet though.

Localhost is great, but it’s a hybrid accelerator + media space & globally distributed. The community space, should it exist, should be 100% dedicated to the creation of stories instead of product, and fully identified against the SF geography.

The best analog for what is needed is New Systems.

Studio Space

Film requires capturing people in physical space. So, a new wave of filmmakers will need a physical studio that they can film subjects in.

There aren’t many filming spaces in San Francisco that look great. And of those that do exist, they aren’t catered directly to founders or venture capitalists.

In the absence of spaces that look good to film in, you get companies designing their own spaces, which leads to a reversion to the current aesthetic default:

Although many startups will prefer filming in their offices (due to time constraints), there still exists an opportunity to build a space that caters to both commercial demand and a community of filmmakers. With an actual production designer.

This would serve as a home for filmmakers to bring their own clients and film personal projects that serve the culture.

Festivals & Distribution

A bit ahead of the curve (since work has to exist first), but there will eventually need to be a specialized festival dedicated specifically to short, narrative, and experimental film projects on the tech industry.

In late ’25 there was enough density of work from the launch video craze that Playback (by Offscript) offered an outlet. This flow of capital and cultural interest has died down, and the cultural pressure has gone dormant (partly due to narrative momentum slowing, partly due to less interesting things getting launched since mid-’25). But it will build back up again as filmmakers begin to make more work.

This festival would run annually. For now, SF Film Club is a small outlet where filmmakers can screen work.

What’s Happening in 2026

So the things to pay the closest attention to this year:

• RAINBOW PARC* is shooting their feature film “Again, Tomorrow” this year

• Jason Carman is shooting an adaptation of the book “The Idea Factory” by Jon Gertner

• SF Film Club is running monthly & SV Film Fund submissions close June 10th

• Offscript is pushing the commercial register, with eyes on reshaping industry-wide narratives

• I am currently writing a short film, shooting in 2026

The Reality

I see a tweet every few months of people expressing frustration around the lack of stories addressing tech honestly, I think about this post from @VirtualElena on tech’s thin film canon:

Or this one from Keller Cliffton:

…Hollywood writers don’t know what building in tech is like, and we have 0 professional writers here. And it will take 5+ years to train new writers, who wouldn’t have worked in tech in the first place…and if they were in tech and succeeding it would be so lucrative they wouldn’t leave…a big inescapable loop!

So, it’s important to realize that tech is structurally incapable of producing stories on film, and I hope this list makes that clear. We can’t rely on Hollywood to understand a culture that is so different from a completely different geography. We have to build the storytelling capacity here.

So before firing off a tweet about SF or tech not producing stories, or Hollywood misrepresenting tech (tweets are fine…but), we have to look at the reality. The reality is above. To change reality, you have to move time and money towards talent and a vision.

You either build this capacity, or the stories do not get made.

If you’re missing from this list or any information is incorrect, please dm @bephrem on X, and I’ll add or correct.

I’ll be working with exactly 2 startups on custom studio builds in Q2, including end-to-end content production. DM @bephrem if interested. Otherwise, this article gives you all the basics to get started yourself.

The full shopping list is here. You can jump to the build section if you’d like to skip the intro (it’s lengthy).

The Problem

Every startup has the same basic problem. You ship features every week, and you want to talk about them to the public. To do this, you need to commit your voice and image to film in a compelling way, so you can speak to thousands of people. Potential customers, employees, and investors. This problem repeats every month. It never goes away.

You consider your options. You could:

(1) Hire a Crew, Shoot In-Office: A camera crew will come, spend 2 hours setting up lighting and 2-3 cameras, then charge a flat day rate (maybe ~$3-8k) just for footage. Across 4-5 videos/mo $12-40k in content spending. This is unsustainable. I’ve observed this setup process dozens of times. It requires a location scout every time, with new lighting & sound considerations. You get 1-2 pieces of content, w/o integrated consideration of a communications roadmap for the company. And most offices don’t look good; they look like startup offices.

(2) Hire a Crew, Rent a Location: You can rent one of the few filming locations that look good in San Francisco, then run your own camera setup, or work with a crew. This is also not ideal as it takes you out of your office and the flow of your day. You face the same setup / teardown problems, although you get good production design for “free” (but even then you get a generic look that isn’t tailored to your company & the tone you want).

(3) Buy Gear & Film Yourself: This seems like an enticing option; you just need to buy an FX3 & some RODE Wireless Go’s, right? Until you realize that compelling film work actually has many moving parts that make it watchable:

• Writing: What are we actually going to talk about (topic selection + framing)? At what timestamp do certain important points land (estimating watcher attention span + runtime)?

• Cinematography: There are many invisible principles to making an image “easy to read” to the eyes & pleasing to look at. Almost no one is getting them right on X & LinkedIn. You need to know the rules behind lighting and composing an image, which are non-obvious.

• Production Design: Every pixel of the frame must be load-bearing. Almost no one is designing their sets with a unique identity that is an extension of their product’s visual language. The same care put into a digital design system (typography, color, spacing) — should extend to the physical world.

• Performance (Acting): The most important visible component is how the people on-screen make the watcher feel. Does the presenter transfer good energy to the watcher? Does this host-guest combo have good chemistry? This is what keeps people watching and brings them back. This takes an eye for what isn’t there in a performance.

• Sound: Most videos on X & LinkedIn have mediocre sound. Not that important if the content is good, but gives a presence boost for the watcher if they can clearly hear the speaker, and the spatial logic of the room can be felt. Good sound lets people watch longer and more easily. Getting dialog polish right can completely change how an image is felt.

• Music (Score): The music needs to be in conversation with the rhythm of story information unfolding. It will need to hit at certain moments, and supplement the overall tone of the performance.

• Editing: Picking the best takes that cohere. Graphics. Motion.

• Coloring: Footage needs to get colored. This seems optional until you see a before & after of footage that has been properly colored. Below is a before & after from a recent shoot I did with @arlanr, with my coloring treatment notes. Coloring can profoundly transform how an image reads.

You can see how the needs of the medium stack up. Founders with growing companies don’t have the time to run a production process end-to-end like this.

(4) Hire a Freelancer or Head of Media: Find a pluralist visual thinker who is also mission-aligned with your product. Not impossible, but difficult. Most freelancers don’t have rigorous film training. And most of the best visual artists don’t care about B2B SaaS or AI. Some will. The best ones will be thinking about the above.

(5) Hire an Agency on Retainer: If you’re at the stage where you can afford an agency, you can hire them for larger brand campaigns or recurring content. Once a large brand campaign is delivered, agencies operating at that level will have to context-switch to other large productions for other companies. Recurring content producers work with many clients and won’t embed with you at an organizational level. What is really needed is a producer deeply embedded in the product, tracking changes every week, suggesting new pieces — fully in the loop.

(6) iPhone: Just use an iPhone. If you have something really raw & true to say, iPhone and room audio (“low production value”) could connect even better. But for most regular content, this just makes it harder to look at & hear you.

(7) Do Nothing: Many startups just don’t produce content when it would be low-hanging fruit to do so. I see this all the time. Founders with good engagement on text posts & compelling personalities that would do well on-screen — that just never have the time or consider making videos regularly.

…

Most of these aren’t permanent options. 5/7 are temporary fixes. The ideal scenario is that you own the means of media production. A permanent studio (whatever the scale) built around your company’s identity, producing content on your schedule, indefinitely. To understand why this matters, let’s look at what media even is, from its cultural to its commercial applications.

Media and Story

Media is how we form and shape our understanding of the world. As much as we’d like to believe we are “independent thinkers,” our thoughts and beliefs are primarily constrained by what we have seen and what we have heard throughout our lifetimes. Born ignorant, through a series of media transformations, we come to an understanding of the world. We are perpetually at the mercy of what we are shown and not shown (narrative constraint).

Media’s existence is necessitated by a hard physical constraint. We are born to separate bodies & minds that need to constantly sync to a relatively aligned view of the world for things to get done. This alignment problem is as old as human cooperation. Storytelling is the first technology ever invented to solve this coordination problem.

Story sits inside media as latent structure. It encodes the patterns we impose on experience to make it meaningful. A beginning, a middle, an end — causality. Character. Stakes. Story is a cognitive structure that makes experience transmissible. Without story, there are just events without a prescribed meaning. When media with story information is consumed, experience is transmitted.

Forms of Media

There are many forms of media, each shaping how meaning moves through the world:

• Oral (the first): Spoken word. The original medium — no technology required. Every civilization developed it independently.

• Image (the 2nd): Painting, drawing, photography. Static visuals making the seen world persistent.

• Written (the 3rd): Text on a surface. Thought made portable, permanent, and copyable.

• Audio (Recording): Music, podcast (longform speech). Recordings freed oral storytelling from real-time constraint and made it portable for the first time.

• Cinema (Film/Video): Moving image + sound + time. The most complete simulation of lived experience humans have ever built — addressing sight, sound, motion, and emotion simultaneously.

• Broadcast: Television and mass radio. One-to-many at scale for the first time. A single signal reaches every home simultaneously, creating shared monoculture and centralized narrative power.

• Print: The Gutenberg press made writing infinitely copyable and cheap. For the first time, one person’s ideas could reach millions without distortion, enabling mass literacy, mass politics, and mass religion.

• Digital/Internet: All previous forms containerized and delivered over a near-instant network. The first truly bidirectional medium at scale.

  • Social Feed: The current dominant form — an infinite, algorithmically sequenced scroll optimized purely for attention capture.

• LLM Chat: The newest layer. Many now treat AI conversations like their closest advisor. Extraordinary narrative power that’s still under-discussed.

I’m probably missing other ones (games, live interactive media, and on and on). But these are all ways that meaning moves through the world. Your company needs to be fluent in at least one of them.

Media is Infrastructure

I wanted to lay the above foundation because most of the technology industry has a very narrow understanding of what media is for or what it can do. The industry term for commercial media is “content”. A container word treating every medium as a pipe, and whatever flows through it as an interchangeable commodity. It reduces the most complete simulation of lived experience humans have ever built to a unit of marketing output.

As someone who has spent 400+ hours watching cinema in the past year, it is clear to me that media is far more than marketing collateral. It is a transformation mechanism. Taking people from point A to point B in their understanding of your company, the product, and the world. Most companies reason about their narrative on a very narrow time-horizon (<3mo) and scale of impact (“let’s share this feature we just made”).

Founders will usually finish a feature, realize they need a video for it, reach out to their immediate network for a videographer, record a video, publish it, and then be done.

Venture capitalists will usually have a portfolio of interesting founders they can talk to, bring them onto some sort of long-form show or podcast, record a 1-hour video, publish, and then be done.

If we’re honest, this is the state of storytelling in tech for 80-90%+ of companies and venture capital firms. Wholly uncreative, yet functional. It works (and really well sometimes), because (1) the quality bar for storytelling on X & LinkedIn is incredibly low and (2) inventory is incredibly low relative to demand (attention surplus).

Few companies are treating their media as infrastructure. Just as product infrastructure is planned and roadmapped with total seriousness and permanence, a company’s storytelling infrastructure needs to be equally as considered.

Most startup problems are downstream of storytelling:

• Acquiring Talent: The best talent works for companies worth joining. The companies worth joining are telling the best stories about why they exist and why they’re going to win.

• Attracting Capital: Venture investors pattern-match on conviction before they take a meeting. Public media is your first pitch. Most founders don’t realize how much due diligence happens before anyone reaches out or replies to inbound.

• Acquiring Customers: People buy from companies they trust. Trust is built from repeated contact over time — not a single video. The accumulation of consistent presence is what moves someone from aware to convinced.

• Retaining Employees: Employees rally around a strong why behind the work they’re doing. Stories about what you are building create this shared bonding mechanism and keep people tethered to why they’re building.

• Public Perception / Positioning: Your public narrative shapes how the industry positions you relative to competitors over time. You either shape it or it shapes itself without you.

Good media strategy is simply an organic multi-year conversation you have with the world about who your company is and why it exists. It is a cloud, a consciousness, built and pointed in your direction. Every point of contact makes people feel what it is like to use your product and work with you.

Non-Linear Effects: While product functionality is immediately measurable (it works or it doesn’t), the impact of stories a company tells about itself are only observable in a very long-range, non-linear way. You might release a podcast episode that someone listens to in 3 months, then returns to your venture firm in 3 years for an eventual investment. There was no way to track this ROI because the effects were impossible to measure and they spanned years. You just have to trust it’s happening and “read the room”.

X & LinkedIn Are Underproduced

This is hard to see if you aren’t a video publisher, but X & LinkedIn are heavily underproduced for video, and the X & LinkedIn algorithms are very generous in awarding reach.

3 months ago, right as I was thinking this, a post from @_kaitodev came up in my feed that captured this well:

Here is the tweet and here is the transcription:

“Content creation is vastly unexplored. You have a billion users on LinkedIn, and like 3 million active posters weekly. That means that’s like 997 million are not posting at all. Lots of people want to consume. Not a lot of people are producing. Please, make content. You have big attention that you can get. And money :)”

And I quote tweeted:

I don’t have extensive data to confirm this attention surplus, but I have empirical evidence just from…observing how videos have performed on X over the past 3 years. I closely study how views are served minute-by-minute, hour-by-hour — where posts top out, how topics surface and subside, how the “For You” feed builds itself. And it’s pretty clear that reach on X is pretty easy right now, even without highly polished film work.

And these aren’t fake views, these are real people becoming convinced and changing their opinions about whole industries & companies. This is a new media platform sitting right under people’s noses, and it’s invisible for those w/o the data to see clearly.

TBPN Owning Broadcast

Somewhere along the journey, TBPN must have realized this gross attention surplus. They started the show informally (in Oct 2024), but on some level, they probably felt there was a missing media platform they wanted to see. This platform was specifically broadcast. A whole industry opened the X timeline daily, and there was no real-time news source with trusted face(s) to help the industry (and the world) make sense of it all. John & Jordi were perfectly suited to do this work due to their backgrounds, but even at the time, no one was taking tech media seriously.

It was sitting in plain sight. To own the narrative of the most powerful industry in the world you just had to (1) be the right people to do the job (of course), Jordi & John have the exact chemistry, charisma, and level of insight to even host and make the show what it is (2) hire a producer and a ton of interns in LA to rip content into the X & LinkedIn feed (while posting on YouTube, Instagram, & TikTok as tertiary platforms).

Fast-forward 1 ½ years to a $100M+ acquisition (“low hundreds of millions”) by OpenAI.

TBPN is an X-first media platform, with 333k+ followers on X, 30k+ on LinkedIn, 70k+ on YouTube, and 30k+ on Instagram. This is the first time that a big dollar sign has been placed on the trust and attention of a whole industry. Now with a market cap, X-first, “tech media,” is legitimized for the first time in its nascent 2-3 year existence.

The Conclusion

If you are a startup where your audience is primarily on X and LinkedIn, you should begin producing video content for the platforms. There is real money in trust, customer attention, and positioning being left on the table that is far lower than the cost of building a studio and producing.

The following will explain how to do this.

The Studio

Foundations

The best way to explain a studio concept to you is through example. Showing you what good writing (dramatic structure), cinematography, production design, performance, sound, & coloring look like. If you understand the basic principles of a fixed studio setup, you can extend it to various visual problems you’ll face.

I can’t cover 4 years of film school here, and none of the following is even a fraction of exhaustive. But the following sets a foundation for understanding the kit list later.

Cinematography

First, we will talk about the very bare basics of cinematography. We will work with the below sample video I shot with the great @jia_seed. The video is a personal narrative delivered in a sit-down format, compressed into a 1-minute Twitter attention span. Graphics are intentionally minimal; it is oral storytelling in a cinematic container.

Cinematography is the art of telling a story through moving images. When watching a video, people are “reading” the image, the same way they read a novel page by page. This reading happens very fast, and instead of reading letters, we read faces, body language, motion, color, exposure, objects, visual patterns, and on and on. What is around characters (production design) & how they are making us feel (performance) are all absorbed subconsciously.

Most of the story for our studio will be told through the production design of the set. Above, we filmed in Jia’s workspace in San Francisco. When choosing the frame I kept 6 key attributes in mind:

1. Area of Interest (Focal Point): What is the precise piece of the frame that the watcher’s eyes will be fixated on? Everything in the frame either works with or against this visual gravity.

2. Planes of Contrast: Our eyes are drawn to movement and contrast in a frame. Where are the areas of contrast in the frame (where sharp changes in color, exposure, or shape happen)? How do they interact with the focal point of the image? If you rewatch some of your favorite movies, you will find that actors’ eyes routinely get placed on hard lines of contrast in the image because our eyes move there.

3. Exposure: Exposure is the amount of light a region of the image receives. We want the face well-exposed. Effortless to read. Above, Jia’s face is very easy to read, the sky accentuating/highlighting the top region of the frame we are looking at.

4. Depth: How far back does the image reach? To reveal 3-dimensional space so we can inhabit it? The more depth an image has, the more story information we can absorb, which naturally creates interest. Most studios in tech lack depth.

5. Color: Jia’s wardrobe is coincidentally (and rightfully) muted to allow the exposure on her face to shine, making her easy to read and listen to.

A full breakdown video can be found here, as well as a breakdown PDF.

For our studio, the main principles for cinematography will be:

1. Make the face easy to read: Expose well for the face, if the space justifies, light for the space first then the face. Since we will just have 1 or 2 people talking, being able to read facial expressions will be critical. It’s the main mode of communication.

2. Find depth: We don’t want a flat photography studio look. Depth gives people visuals to chew on so they want to stay in the world they’re watching.

3. Control wardrobe: Keep it neutral and make it work with the overall set color.

Production Design

Production design is the intentional construction of space, furniture, light, color, texture, and objects within a frame to carry meaning rather than just occupy space.

Almost no one in tech is doing production design well. The current state of production design in tech is: laptop, table, books, wall.

Exhibit A (the tweet)

Exhibit B (the tweet, did 1M impressions mainly because of the announcement itself, but video definitely played a role in extending the reach)

Exhibit C (the tweet, the most common studio design perpetrated — wood acoustic panels off Amazon + Shure mics on table + overhead lights flooding the room with exposure)

Exhibit D (anthropic youtube — anthropic is doing a good job)

For our studio, we will want to intentionally choose every object that ends up in the frame. The chairs, table, cups, walls, & visible lighting elements. No room for visual clutter, everything is load-bearing. And somehow avoid being visually trite, while avoiding overinventing for the sake of it.

Writing & Performance

Every story has a beginning, middle, and end, which ultimately connects to a meaning. There is a problem we stay to watch get resolved; when it is resolved, we are done watching.

Let’s look again at the Jia piece. It’s so short and simple that it is a good learning ground for basic dramatic structure.

A dramatic “beat” is the smallest unit of change in a story — a shift in emotion, relationship, understanding, or stakes. These beats are what keep people watching.

Here’s where they are located in the short Jia video:

• 2s — Dramatic Premise: “Moved to San Francisco and didn’t know anyone.” The story is set in motion.

• 6s – Escalating Conflict: Dropping internships at Disney and Intuit to chase startups. Stakes and desire appear.

• 20s – Deep Empathy: Real stutter kept in — vulnerability makes her relatable.

• 26s – Breakthrough: “And surprisingly this worked…” We feel the character progressing.

• 47s – Resolution: “All that matters is just who you are as a person…” Everything pays off with meaning.

If your video is missing some sort of dramatic spine, it will sit inert and be unwatchable. You have to create stakes that are true to the thing you’re talking about. They often won’t be that compelling. And that’s ok. They just need to be there, and you need to have considered it.

Sound

Sound is deceptively difficult to do well. There’s a difference between dialog that is legible, easy to hear, and dialog that puts us in the room with the speakers. For our studio, we need clear and crisp dialog, while also creating a sense for the spatial logic of the room.

Here’s a sample where the room’s spatial logic is expressed in the sound:

Notice how every piece of the frame is controlled. The chairs, table, glasses, clothing, my clipboard — are all in conversation with each other. Everything has been designed to support the story and accentuate focus on the speakers. Also, notice how the footsteps pan & have spatial integrity to where they are coming from in the frame. These sound choices are logical to the scene and, most importantly, put you in the room.

Score (Music) & Color

Finally, we’ll touch on score (and throw in coloring as a quick note). I have a sample from a 6-part series I shot with Arlan (again). Below is a 2-minute commercial spot for Nia’s research paper parsing functionality. The ambient score accentuates the depth of the room, creates curiosity, and has a slight volume scoop at the start when the camera smoothly snap-tilts up to reveal Arlan’s eyes.

Performance, cinematography, and score all cooperate to create a viewing experience that moves in lock-step. For content coming out of our studio, we’ll be shooting on static cameras (no movement, unless an operator is on-site), but the same principles of score coordinating with transitions / setting / story apply. It’s all 1 cohesive thing.

For color, here’s the before and after (with notes):

Principles, Not a Blueprint

The above just covers the core principles that make contained film work like this compelling. Even in a simple, fixed, studio setup like this, good performance, production design, cinematography, score, & sound — can make something worth watching.

The Build

So now let’s build our theoretical studio. It will just be a person sitting in a chair with 2 cameras pointed at them. A medium shot (to capture environment and body context), and a medium close-up shot (to capture facial expressions and character interior). Really simple.

Our studio needs to solve 9 fundamental problems:

1. Image Capture: We need cameras. With the right lenses. On tripods. With continuous power and memory to record to.

2. Monitoring: We need to see the image we are capturing on a big screen. On a tighter budget, this is optional, but a display monitor gives us the ability to see multiple camera feeds at once, from 1 physical standing position. We can clearly judge:

   1. framing

   2. composition

   3. exposure

   4. contrast

   5. color

   6. focus

   7. background separation

3. Sound: We need to capture dialog with clarity and presence. Both are achieved via proper mic placement.

   1. Clarity is ensuring every word is intelligible without effort. Listeners shouldn’t have to work to understand what’s being said.

   2. Presence means the voice feels like it’s coming from a real person in a real space. The dialog quality creates a sense of “being in the room”. Many clip-on mics produce functional audio, clear and intelligible, but you feel like you’re listening to a clip-on mic. A properly placed boom will put you in the room with spatial bearings.

4. Lighting: We need to (1) expose the speaker’s face with soft light, (2) add texture to the backdrop / back wall to accentuate the focal point of the image, & (3) add any other practical lighting to the scene to distribute sources of exposure throughout the frame, creating more depth to the overall image.

5. Production Design: For our minimal studio, we’ll just have 2 production-design elements (1) a hand-painted muslin tarp backdrop, held up by stands, and (2) a chair. just a straight-up single chair, to sit on. That’s it.

6. Power: We will need to distribute power to our cameras (for continuous power), monitor, sound mixer, and lighting.

7. Safety: We will need sandbags for any stands (so they don’t tip over).

8. Set Management: We’ll want spike tape to mark the position of cameras, lighting, furniture, and chairs (so we can restore the frame at any time and just hit record). Gray card for white balance, zipties for cables, clapper for audio sync.

9. File Storage & Transfer: We need a continuous system for getting footage from camera cards to post-production reliably after every shoot. Hard drives for on-set storage (SSDs if you have the budget). Optionally, a Mac Mini can live at the studio for full-resolution cloud file upload (more on this later).

Each section will start with the shopping list, then I’ll explain why I made the choices I made.

Image Capture (shopping list)

shopping list

1. Sony FX3A (2x) (link)

2. Sony 100mm f/2.8 STF (link)

3. Sony 85mm f/1.8 (link)

4. Sachtler 75/2 Mark II Aluminum Tripod (2x) (link)

5. Sony 128GB SF-M Tough Series UHS-II SDXC Memory Card (4x) (link)

Camera + Lenses

You don’t need a professional cinema body like an ARRI Alexa, or Sony Burano. And you don’t need cinema lenses.

The highest you need to go in camera body is an FX3, and for glass, Sony lenses. Most character in color, contrast, and texture can be added in post-production. We just need a clean, flexible image.

Lenses

The 85 & 100mm both flatter the face (as telephoto lenses do) and have a controlled, compressed look that create focus on the subject. Both are close in focal length, so that cuts between the angles feel controlled and have similar environmental compression. The 100mm STF renders the background with unusually smooth falloff and can be used for either the medium or medium close-up.

Tripods

Sachtler is a long-standing camera support brand that makes excellent tripods. There are cheaper options on Amazon, but if you’re going to place a $6,000+ camera + lens on something, you want it to be very, very stable and reliable.

Cheaper tripods break down over months as you open and close them. Although our studio cameras will never move, better to just buy it right once. Or…get a cheaper tripod if you want. It will break on you eventually, but it will be delayed relative to how much you open and close it (I’ve had 3 cheap tripods off Amazon deteriorate on me over ~6-12 months of use).

SD Cards

The Sony TOUGH cards are molded into a single durable piece and are unlikely to break. There are cheaper options for cards that have write lock switches (which are undesirable). Really up to you, the studio isn’t moving anywhere, but I just like buying what will last (especially because this is a professional/commercial setup).

Monitoring (shopping list)

shopping list

1. Lilliput 31.5” 4K HDMI Monitor (link)

2. Neewer C-Stand w/ Casters (link)

3. VESA Mount (link)

4. 35’ HDMI Cable (2x) (link)

5. HDMI Right Angle Adapter (2x) (link)

The Stand

The VESA mount grips the back of the monitor (4 screw holes), then clamps to the top of the stand. The stand is rollable so people can move the image around so anyone can see.

The Cables

SDI is the professional standard for moving video over cable. Teradek for wireless transmission. The FX3 does not have SDI, so we use HDMI. HDMI will come out of the cameras and go into the back of the monitor.

We use right-angle adapters to reduce strain on the FX3’s HDMI port and keep the connection secure.

Sound (shopping list)

shopping list

1. MixPre-6 II (link)

2. Sennheiser MKH 50 (link)

3. K-Tek Boom Pole (link)

4. Matthews C-Stand (link)

5. Matthews Grip Head (link)

6. K-Tek Boom Cradle 2 (link)

7. XLR Cable (25') (link)

8. Sony 2x Lavs + Receiver (UWP-D27) (link)

9. Sony MDR-7506 (link)

10. Gaffer Tape (55yd) (link)

11. AA Batteries (20-pack) (link)

Overview

All sound goes into a mixer. It can record up to 8 tracks (support multiple speakers). XLR connectors are the professional standard for moving audio. Boom mics (the ones you see operators with poles on movie sets move around) provide superior audio with a rich frequency range. When you think “that audio sounded like a movie,” it almost certainly came from a boom microphone. The FX3 supports 2 XLR inputs and one 3.5mm input (which you can’t use at the same time as the XLR inputs). We will not use any of these and instead rely completely on the mixer to record audio.

Boom Mic

The Sennheiser MKH 50 sounds natural, rejects room noise well, and makes voices feel like they exist in a real space. I personally love the DPA 4017b’s for their superior clarity and detail, but they’re on the pricier side.

The boom mic attaches to a mount that comes with it (you can get a shock mount if you want, but optional) → that mount has a female screw, the boom pole screws into it → the boom pole holder cradles the boom → the boom pole holder is gripped onto by the c-stand grip head → the c-stand grip head goes on top of the c-stand (5/8” baby pin) → XLR goes from the boom pole (which has a female XLR port) into the mixer.

Boom Mic Placement

The most important thing for sound quality is mic placement. Worth repeating because it will save you hours and hundreds of $$$ of post-production headache and disappointment. The most important thing for sound quality is mic placement.

You will position the boom above your subject (yourself, probably), with the boom pointing down from above at a 45° angle. Pointing at the chest / mouth area. Slightly in front of the mouth, where the sound is traveling. 6-12 inches from it. The closer the better. Every inch matters.

This has real commercial consequences. A wrong mic placement capturing too much room audio will lead to lower video retention and lower reach simply because it is harder to hear the speaker. Rich sound capture stands out on a timeline of videos with poor sound capture.

I can’t stress how easy this is to get wrong. If I’m not there or a good video person you know isn’t there to check the audio, it is deceptively easy to let mediocre audio capture slip by unnoticed. Because it will sound functional and “just work,” but it won’t wow you. It should wow you and call you to attention.

Lav Mic

I added a lav mic for backup, so you’d run both a boom ++ lav mic for redundancy into the mixer. The Sony UWP-D27 has a receiver that will run a 3.5mm jack into the mixer.

Lav Mic Placement

Place the lav on the chest, centered, 6-8 inches below the mouth. A lot of great videos on YouTube on lav mic placement. It’s truly an art.

I included gaffer’s tape for securing the mic wherever you put it.

Monitoring

Headphones for monitoring (industry standard). Another rule that will save you a lot of pain and despair: always have someone monitor the mixer live while filming. Anything can happen while recording; you have no idea what sound is being written to those cards unless live ears are listening. A cable can malfunction, the mixer can get unplugged, the mic placement might drift, there might be ambient noise running you don’t notice, but the mic will capture it clearly.

If you’re filming solo, at least have eyes on the gain levels of the mixer, make sure things are flashing and alive, and…working.

Dialog Editing

Additional note — going through an additional step of dialog editing can make a night and day difference in the clarity and presence of your final dialog (think “sounding like a Super Bowl commercial” vs “sounding like a YouTube video”).

Here is a before (no score) and after (dialog polish + score + mix) of the “research paper” video with Arlan. Notice how before, you feel like you’re in a big echoey room. It pulls you away from focusing on Arlan’s words, and instead, your attention goes to your position in the room. This could be the right story decision in other cases, but in this case, it only detracts from getting our message across.

After dialog polish, you feel like you’re “inside his voice,” receiving it directly. You’re locked in on his words, and your attention does not drift elsewhere.

Lighting (shopping list)

shopping list

1. Aputure 600c Pro II (link)

2. Matthews C-Stand (link)

3. 6x6 silk diffusion rig

   1. Matthews 6x6’ Silk (link)

   2. Matthews Frame (6 x 6') (link)

   3. C-Stand (10’) (2x) (link)

   4. Grip Head (2x) (link)

4. OR, softbox

   1. Aputure Light Dome (5’) (link)

5. Amaran 60x Bi-Color (link)

6. Impact Light Stand (link)

7. Cinefoil (link)

Key Light

So let’s revisit the medium shot on our studio subject (I pulled this still from Shotdeck, a shot reference site filmmakers use).

When lighting this scene, we need to think about 3 planes of exposure (1) the subject (torso + face), (2) the backdrop, & (3) the overall room (ambient).

There are 2 lights active in this setup. (1) There is a big diffuse “key light” coming in off screen-right to expose the subject’s face & torso, and (2) there is a light coming in off screen-left to add texture to the backdrop (near the focal point of the image, which is the subject’s eyes, siphoning attention center-frame).

The Aputure 600c will be our key light for the subject. It is powerful enough to overpower strong sunlight coming in from windows to expose our subject well. It will emit a hard light that we need to diffuse. We have many options for diffusion.

Here are 2:

1. 6x6 Diffusion Silk: The wider the surface area of a diffusion fabric & the farther back the light source before it hits it, the softer and more diffuse the light. You can check out this diffusion fabric test on YouTube to see how different diffusions create different qualities of light on skin.
   
   I chose silk because it fully diffuses the light behind it, eliminating hot spots, while preserving the warmth and color accuracy of the source. The Matthews 6 x 6’ frame holds the silk. It is on the pricier side, so you could get away with just hoisting the silk between 2 c-stands (grip heads would bite into the fabric & you’d spread it taut).
   
   There is a big qualitative difference between light diffused by a big silk cloth vs using a softbox. It is more ethereal and cinematic.

2. 5’ Softbox: This is an alternative. A softbox attaches to the front of the Aputure to soften the light that comes out of it (it twists and locks to the front of the light, into what’s called a “Bowens mount”). The light this produces is very soft, softer than any smaller softboxes. Still not as great as silk, but it’s a more portable option.
   
   Use the grid that comes with it to control the light spread, keeping it focused on a subject. Keep the grid off if you want to boost ambient exposure in the whole space + behind your subject + your subject themselves.

Foil

The cinefoil is this rigid, aluminum foil-like foil (but much stiffer). Similar to barn doors, it allows you to shape the light coming out of fixtures.

Backsplash

We only need a small light for the backsplash. We can use the cinefoil to shape it how we please.

Hairlight

This isn’t on the shopping list, but you can add an extra light, place it behind your subject, and aim it at the back of their head to create a “rim light”/“hairlight”. This creates an accentuated “halo” of exposure behind your subject, creating further background separation and making the image easier to read.

You can also add a touch of light to the shadow side of the face for what’s called a “kicker” to create more dimension on the face. Good video on all this here.

Reading the Face

Our main goal with lighting is to make the face easy to read and give room elements a proper relative exposure. If other elements in the frame are competing for exposure, it will take a longer time for a watcher to parse the image.

Comprehension should be instantaneous, without suppressing other exposures to the extent that the room becomes dark and moody.

Working with a Cinematographer

If it’s in budget, I strongly recommend consulting a local cinematographer. Have them walk through your office, and ask them where placing the studio would be best. They will help you plan around the sun relative to time of day, potential framing, and just overall make the image look great.

You want decisions passing through a real artist’s eyes (which is why most studios and productions in tech look bad; they aren’t passing through a real artist’s eyes, someone who literally lives off making beauty).

Production Design (backdrop + furniture) (shopping list)

shopping list

1. Hand-painted Muslin Backdrop (10x12ft) (link)

2. C-Stand (10’) (2x) (link)

3. Grip Head (2x) (link)

4. Paracord 550 (link)

5. chair options (quickly grabbed these)

   1. Past Life Chair (link)

   2. Gimlet Chair (link)

   3. Haiku Dining Chairs (link)

   4. Armchair (link)

Backdrop

We can get the hand-painted muslin backdrop from Etsy.

The main idea is that we want to create a sense of depth while neutralizing background components that are not essential for viewing in the image. Every item in the frame must be interrogated for meaning. Does it improve the frame’s balance? Visual rhythm? Distribute colors better? Reveal environmental information? Or is it just there to be there? In the final case, it should always be removed.

Chair

A chair. As always, color, texture, material should work with the overall set.

If you want, we can add a table, then put a laptop on it.

Working with a Production Designer

The production design is most of the viewing experience for a fixed setup like this. If you want a custom-tailored solution that looks cinematic, you need to consult with a production designer. If you just place stuff around yourself, it can work…but it usually won’t look great.

A production designer will bring a trained eye to the space — someone who thinks about color, texture, and object placement as visual language. Worth the consultation fee for the first build.

Power (shopping list)

shopping list

1. Power Strip (13 outlets, 15A) (link)

2. Extension Cable (25’) (link)

3. USB C to USB C (10’) (2x) (link)

4. USB-C Charger (65W) (2x) (link)

Pretty Simple

We get continuous power to the 2 FX3s via USB-C cables connecting to power bricks. Batteries stay inside the cameras in case the power goes out.

We have a power strip to power the monitor, the 2 lights, and sound mixer.

Safety & Set Management (shopping list)

shopping list

1. Sandbags (20lb) (8x) (link)

2. 18% Gray Card (link)

3. Spike Tape (link)

4. Hook & Loop Cable Ties (50x) (link)

5. Clapper Board (link)

Sandbag C-Stand Legs

For safety, we want to place at least 1 sandbag on the biggest leg of any C-stand (the side that is bearing weight). The “biggest” leg will be the one highest up in the “ring” of legs (you’ll know what I mean when you’re looking at the stand). This is most important for the backdrop & our big Aputure key light, those falling on anyone would be very bad!

By the way, here is how you use a C-stand.

Other Things

Gray card to set white balance, spike tape to mark positions of set items (for repeatable frames), slate to sync audio (or you can just clap your hands), and zipties to manage cables.

File Storage & Transfer (shopping list)

shopping list

1. Sandisk 2TB SSD (2x) (link)

2. LaCie 4TB (link)

3. Dropbox Plus (link)

4. Frame.io Pro (link)

5. optional

   1. Mac Mini M4 (link)

File Transfer (Streaming Proxies)

Finally, we come to file storage and transfer. Pretty simple. We will use Frame.io’s “Camera to Cloud” to stream low-resolution files (called “proxies”) straight off the cards (literally as we are recording) to our editor. By the time the shoot is over, they will have all the files they need to begin the edit. Editors work with proxies until a final conform step, where low-res files are swapped for full-res so that color-grading can be performed.

Full-Res Storage + Transfer

Someone in the studio will have to take the cards out of the camera and move the full-res files to the cloud. Dropbox can auto-watch a folder on the studio machine to auto-trigger cloud upload. The editor will download the full-res files from Dropbox.

Locally, we will make 2 copies of the footage onto SSDs. This is standard redundancy when taking footage off a set. The stakes are lower here and you can be less cautious if you’d like keeping only 1 copy.

SSDs are much faster than HDDs. You can use hard drives if you want; they are much cheaper.

Recap

So to recap:

• Proxies via C2C: Stream low-res files to Frame.io during the shoot — editor can start before you’ve packed up.

• Full-Res Offload: Cards pulled after the shoot, copied to 2 SSDs for local redundancy.

• Cloud Upload: Dropbox auto-watches the studio machine and uploads full-res files automatically.

• Final Conform: Editor swaps proxies for full-res masters from Dropbox before color grading.

The above minimizes time to first rough cut.

Network Attached Storage

At some point, you will outgrow using individual hard drives. You will want a NAS (Network Attached Storage) — a device that sits on your local network with multiple drives in a RAID configuration, automatically mirroring data so a single drive failure never costs you footage. Synology is the standard brand. I haven’t set one up myself yet, so when that day comes I’ll just watch a YouTube video or ask Claude what to do.

Studio Grand Total

So the total (on just equipment) is close to something like this:

1. Camera: $12,700

2. Monitoring: $1,900

3. Sound: $4,800

4. Lighting: $3,600

5. Backdrop: $500

6. Furniture: $500

7. Power: $100

8. Safety: $200

9. Accessories: $60

10. File Storage & Transfer: $1,600

Grand Total: ~$26,000

To put this into perspective, $26,000 is roughly what a single mid-level marketing hire will cost for 3 months — before benefits, equity, or ongoing salary. This studio produces content for years.

You can find ways to reduce costs, recheck the list, and remove what you don’t need.

After the Shoot

The above just covers production costs (there’s no pre-production since you’ll just walk in and shoot from a light script). For post-production, you need to budget for original score (or music libraries), dialog editing, sound mixing, color, graphics & motion, and subtitles / captions. Post-production is out of scope for this piece.

I’m planning to write a future piece, “Film Production for Founders: What It Costs,” that goes in-depth on what every piece of production costs for both the producer and the end-client. So you can know exactly what your producer is budgeting against.

Good Content ≠ PMF

This is sort of an obvious disclaimer, but if you’re pre-PMF, building a content studio shouldn’t be at the top of your list; it should be talking to customers. Good storytelling is important at any stage, but permanent infrastructure is a longer-term bet when there is a product worth telling a story about.

AI

Will AI replace the need to do all this to capture image and sound for something this simple? Yeah, probably. But the deeper you go into the stack of creative decision-making: cinematography, production design, writing, performance, score, color, editing rhythm — the more the combinatorics compound against getting a satisfactory result end-to-end.

Performances in physical space, with someone coaching you through how you come off in real-time, will also be hard to replace. But it’s precarious to be too sure about anything. For now, all you need to know is this is how you’d do it today, and this is what it would cost.

Go Build

So that’s it. If you’re a startup founder or venture fund that’s been considering building your own content studio, this is everything you need to consider or buy. It’s a deliberately simple setup. No complex considerations, no crew of 20, no Hollywood budget. All that’s necessary is a considered space, the right gear, and a commitment to show up consistently.

I’ll be working with 2 startups or venture firms to build their studios in Q2. Each build will be unique — I’ll work with a cinematographer and production designer for the initial build, then stay on as a creative lead for ongoing content. If you want to be one of them, DM @bephrem on X. If you’d rather build the studio yourself, everything you need is in this piece.

tl;dr: The Silicon Valley Film Fund is funding the next great cinematic works on the technology industry. Our first grant will be a $15,000 award to a filmmaker who wants to make a short film. The fund’s short-term goal is to jump-start a new film industry in San Francisco, its medium-term goal is to train a new generation of tech-native writers, directors, and producers, and its long-term goal is to shepherd a theatrically successful film into theaters. We have no portfolio. We have no agenda. Our agenda is the culture’s. Submission criteria can be found below. Submit your script to scripts@svfilm.fund. We will review every script.

We Need Better Stories

Silicon Valley tech is one of the most culturally dominant forces in the world. It has reshaped how humans work, communicate, relate to each other, and understand themselves. The founders, engineers, and builders of this world have changed civilization. And yet there is almost no serious cinema that comes from inside it.

Not caricature. Not satire. Not outsider reconstruction. Genuine interior cinema — stories told by people who lived inside the industry, built in it, and have been shaped by it. And the window to make it is open right now. AI has made building so accessible, and the industry’s implications so far-reaching and consequential, that those building now have the time, and are forced, to look up and ask: why are we building?

Here is the current film canon for the innovation industry:

cinema

• The Social Network (2010): 15 years old, zeitgeist story (Facebook), explores high-level themes of obsession, disconnection from reality, and betrayal. Shaped by David Fincher & Aaron Sorkin. Canonical cinema. This is the best portrait of startup life we have.

• Her (2013): Technology filling human voids. Isolation. A specific modern loneliness. Strong touchpoint, but ultimately observational.

• Steve Jobs (2015): Biopic. A single person and snapshot in time.

• BlackBerry (2023): The rise and fall of RIM. A portrait of founders and a company in real time. But, it’s a documented story working from a historical scaffold.

• Pirates of Silicon Valley (1999): A TV movie. Also on Steve Jobs & Bill Gates.

• Artificial (coming soon, 2026): Sam Altman’s 2023 ousting from OpenAI, another zeitgeist, point-in-time story.

• Jason Carman is working on a sci-fi trilogy (coming soon, 2026-2027)

documentary

• New Space (2025): Jason Carman’s first feature-length documentary.

• Too Cheap To Meter (2025): Jason Carman’s second feature-length documentary.

• General Magic (2018): Close to interiority. A retrospective.

• The Inventor: Out for Blood in Silicon Valley (2019): Theranos. Elizabeth Holmes. Observational.

• Lo and Behold (2016): Herzog treats Silicon Valley as an alien civilization to observe. Fascinating. Not interior.

• The Internet’s Own Boy (2014): Aaron Swartz. A martyr’s portrait. Not a world.

• Something Ventured (2011): Founders celebrating themselves. Hagiography dressed as history.

• The AI Doc: Or How I Became an Apocaloptimist (coming this month, March 2026): The inevitable “regular guy trying to figure out what is going on” documentary. From trailer, looks like it will be a balanced, high-level account. A handful of expert interviews, equal parts fear and equal parts optimism. A perspective that should exist. Still not from inside this world.

television

• Silicon Valley (2014–2019): Satire. Brilliant. Keeps the world at arm’s length by design.

• Halt and Catch Fire (2014–2017): Period reconstruction. 1980s tech. Interior but historical.

• WeCrashed / The Dropout / Super Pumped (2022): The scandal wave. Fraud as prestige TV.

• Black Mirror: Tech anxiety from the outside. British. Speculative.

• The Audacity (coming soon, April 2026): From the trailer looks like more outsider reconstruction with dramatic license.

Biopics about legends, documentaries about fraud, a TV movie, one Herzog oddity, a 15-year-old film made by two geniuses who weren’t from this world, and a timely AI documentary. Seven decades of building the most consequential technology in human history, and this is the only cinematic record that exists of this industry.

Compare this with the finance industry, just looking at cinema: Wall Street (1987), Rogue Trader (1999), Boiler Room (2000), Too Big to Fail (2011), Margin Call (2011), The Wolf of Wall Street (2013), The Big Short (2015). A rich canon of seven films spanning four decades from multiple perspectives. Greed, collapse, moral compromise, human cost.

A thorough cinematic rendering of a world.

Why This Matters Now

Storytelling is the transmission of experience. Cinema defines collective memory. It is the only medium that can make you feel another person’s interior life in your body. It is the antidote to cultural misunderstanding. It is the ultimate unifier. Eras are processed & remembered through the films made about them.

We don’t need to make better films about ourselves in the same way that we don’t need great art. It’s not imperative to our life or death. To this working or that working. But the greatest works of film, the greatest works of art, make it hard to imagine life without them. And you only know after they exist.

When inside the tech Twitter bubble, it’s hard to see that we’re in an era worth documenting cinematically. A time-period where the work this geographic region produces is so consequential and widespread, that not taking agency over its authorship would be a mistake.

AI’s implications are becoming real. Block laid off nearly half its workforce — whether we know it or not, we are slowly entering into a class war. As a film producer, I work with artists (editors, cinematographers, composers, colorists, production crew). It is hard to appreciate, as someone in tech / on tech Twitter, how much artists hate AI. We’re entering one of the most culturally charged moments in tech history, without the ability to shape how this industry is understood. In fact, we’re structurally incapable of telling our own stories on film.

This is not a problem in an immediately apparent way, but more in a death by 1,000 cuts way. Culture operates in widespread, invisible, and impossible-to-measure ways. A Senate bill might get passed because a congressperson saw a film a decade ago that subconsciously influenced their vote. A venture capitalist might decide to fund a specific deeptech startup because they saw a Jason Carman documentary 2 years ago without recalling it.

Collective consciousness decides what is and isn’t possible. What is and isn’t true. It decides what this place is. We are currently not writing to it.

Interior Works

While writing this memo, I began to wonder if other points in history had felt like this, an artistic stagnation that led to a guy writing a memo and starting a movement. It turns out there are 2 similar movements: Italian Neorealism (1945-1952) and the French New Wave (late 1950s and 1960s), where François Truffaut actually wrote a critical memo before studying at Cinémathèque Française for 4 years before making The 400 Blows (1959). Let’s look at Italian Neorealism first.

Italian Neorealism (~1945-1952)

In 1945, Rome was in a state of rubble after WWII had ended. Germany occupied Italy for 2 years, and the fascist period — twenty years of Mussolini — had just collapsed. The city was physically destroyed. The economy was in ruins. Unemployment was catastrophic. And people were genuinely desperate in a way that’s hard to imagine — not inconvenienced, not struggling, existentially precarious. There were no resources and the studios were bombed or broke.

2 directors, Vittorio De Sica and Roberto Rossellini, made films with the resources they had. Their cameras, the streets, and non-actors. Rossellini started the movement with Roma, Open City (1945) using whatever film stock he could find on the black market. It’s not a polished film. It couldn’t be — Rome’s studios were bombed and its entire production infrastructure had collapsed. It featured a cast primarily of non-actors, it featured city streets in rubble, capturing the environment as it existed. He filmed inside the world as it was happening.

Bicycle Thieves (1948) by Vittorio De Sica featured a cast of all non-actors. A man needed his bike to go to work every day and feed his family. It gets stolen and his whole livelihood is threatened. At the time, this would have been catastrophic. Losing your sole means of transportation meant losing your job. The bike represented the father’s dignity, livelihood, and ability to provide for his son. De Sica was offered Cary Grant (an international star) for the leading role of the father with full financing, but turned it down. Instead choosing Lamberto Maggiorani, a real factory worker who had never acted. Lamberto went on to put on a surprisingly and totally actor-like performance as the Dad desperately looking for the bike. It didn’t look like acting because it wasn’t. He was that man.

Bicycle Thieves is a perfect example of a culturally interior work with authenticity that could not have been faked or researched. Rome in poverty, non-actors, a desperate time. Nothing was imagined with cinematic license. It was all real. And a simple story of a man searching for a bike with his son, shot with simple visual grammar, told you everything about what that time and what those people were going through. De Sica knew this world from living inside it.

French New Wave (late 1950s and 1960s)

François Truffaut, a film critic, observed that French cinema at the time had an inert, literary quality (he called it a “Tradition of Quality”). It featured prestige adaptations of classic novels by craftsmen who were following rules established by their predecessors. In 1954 he wrote “A Certain Tendency of the French Cinema”, an indictment naming names, attacking the French film industry. He believed that the director’s personality and worldview should be visible in and inseparable from the work itself. And he believed the films that needed to be made, weren’t being made.

His manifesto would seed the French New Wave movement and solidify auteur theory. Truffaut wrote the memo before any of the films existed. The films followed. 4 years later he would make The 400 Blows (1959) after years of obsessive film study at the Cinémathèque Française (the Paris film archive where you could watch films all day every day for almost nothing), while working as a film critic. The 400 Blows was semi-autobiographical — the main character Antoine Doinel was essentially Truffaut as a boy. Neglected by his parents, in trouble with authorities, finding freedom only in cinema and in running. Deeply personal to the point of exposure. It premiered at Cannes 1959. Won Best Director. Truffaut was 27.

3 Takeaways

(1) A Moment of Cultural Intensity: The common thread between neorealism and the New Wave is that they both arose from moments of cultural intensity that necessitated their existence. Both were made by people so inside the world that the camera had no distance from it.

(2) Thesis First vs Films First: Rossellini (who made “Rome, Open City”) just picked up a camera and filmed the world outside. Truffaut wrote a manifesto before the films came. Different approaches, same result.

(3) Interior Truth: Both movements asked the same question: who gets to tell the truth about this world, and what happens when nobody does? They were started by people who refused to accept that the stories that needed to exist weren’t being made. Neither waited for permission or resources. The absence was unacceptable to them. That’s why this fund exists.

What Current Tech Cinema Doesn’t Answer

The Social Network (2010) is the closest example to a deep interrogation of founder psychology we have. But it’s on one person, a 2003 Mark Zuckerberg, made into mythology. What it does well is capture a person before scale, but it’s just 1 portrait.

Here are characters that have yet to be sufficiently explored:

Here are characters that have yet to be sufficiently explored:

• early-stage founders

• early-stage employees

• late-stage founders

• late-stage employees

• partners to founders / employees

• venture capitalists

• researchers, scientists

• creatives in tech

• those unemployed (and the culture supporting exploration)

• regular residents of San Francisco (and their relation to people in tech)

• the person off tech Twitter with 200 followers that tweets insanely accurate takes from a Victorian in Lower Haight that gets 50 views but is usually right

• permutations of the above that doesn’t fit a specific mold (the above are very linear archetypes)

Here are questions that have yet to be sufficiently explored:

• what does loneliness in this world look like? what are people here really like? what does it look like to go from a nobody to somebody here?

• what does delusional self-belief look like? in its most positive form? in its worst form? what level of faith and foresight does it take to build something new? what are the consequences taken to its extreme end?

• what does love look like? romantic relationships between founders, employees, or others not in tech? what are the tensions? desires unable to be met?

• what does greed look like? optimizing for TC at the expense of everything else? where does that work? where does it fail? what does the striving look like?

• what does the social scene look like? how do tech people interact with the physical city they are in? what does tech twitter as the primary communication model look like?

• how do people talk? what does being terminally online really look like? what does functional autism look like? how do people walk and carry themselves? what do they avoid doing? what are they drawn to?

• what does overoptimization look like? neuroticism beyond reason? a singular focus that is confusing and unreasonable to onlookers.

• what does sacrifice look like? not abstractly. in a body. in a marriage. in friendship. what does growing through that look like?

• what does it feel like to build something that becomes something you didn’t intend? to watch the thing you made with genuine idealism cause harm at scale? and deciding who you are in relation to that.

• what does a high-growth startup look like? what are the ramifications on psychology?

• what does leaving the industry feel like? the engineer who walks away from salary, mission, identity, and has to reconstruct themselves outside the only world that ever made complete sense.

• who are these people really? not the myth. not the caricature. the actual human beings underneath.

None of this exists cinematically. That’s what we exist to fund.

What We’re Looking For

A script for a short film, 15-20 minutes long.

The criterion is not a résumé. Not a zip code. Not a job title. It is interiority. A relationship to this world that got into you before you knew it would matter.

You don’t need to have been a founder or an engineer. You just need to know this world from the inside — its specific people, values, contradictions, and way of being. The script will know if you do. The details that are too specific to have been invented. The silences that assume shared knowledge. The things that are in the script because they had to be — not because they were written but because they were lived.

You might be a director already producing commercial work for startups. Or a freelance videographer in San Francisco. You might be a founder who moved away. An engineer who writes screenplays and studies film as a hobby. A producer from LA who moved here and is figuring out the cultural landscape you now work in. There is no exact shape to who the recipient of this grant could be.

We’re uninterested in funding outsider interpretations. Dramatic work written from guesses or “wouldn’t that one line be funny.” Hacker thrillers. Cautionary tales on social media. If you have to make anything up, related to the core characters and story, without it being grounded in true life experience and observation, we are uninterested. Please do not submit your script.

Good scripts will/may have:

• details that are too specific to have been invented

• scenes included because they really happened, rather than just being dramatically effective

• characters only people in tech would recognize immediately

• locations that are too specific to the world that they had to have come from lived memory

• some core unresolved question the filmmaker wants to answer

• interior stakes that are too specific to have been invented

• dialog rhythm that is too specific to have been invented, it had to have been heard

• basically, a lot of things that are too specific to have been invented

If you have a script like this, we want to read it.

The Fund

Creative talent with the interiority to tell these stories exists, but no capital structure, void of private interest, has formed around them to take their craft and the stories they hold in their heads seriously. To take storytelling as a career seriously.

This fund is that missing structure. To greenlight stories that need to get made right now. To create structure around what is right now a nascent commercial filmmaking scene.

We are not a startup with a product we need to market. We are not a venture capital firm. We have no portfolio. We have no favorites. Having a product is fine. Having a portfolio is fine. But because all of these interests rarely align in the best interest of broader culture, a neutral institution needs to exist.

We simply look at gaps in the technology industry’s storytelling, and aim to fill them with truth. That’s it.

The Grant

$15,000. Awarded to 1 filmmaker to make 1 short film.

on budget

You don’t need a big budget when you have something true to say, and the craft to say it well. Medicine for Melancholy (2008) was made by Barry Jenkins on $15k (~23k today). 2 characters, handheld, San Francisco as backdrop. Through the rawness of the performances, you could still sense something very true being said, and you could clearly sense an early & powerful directorial voice in his shot grammar. Constraint can create great things.

Submission

You know if this is you. You’ve known since the first paragraph. You have a story that couldn’t have been written by anyone else. We’re building this fund for you.

Send your script to scripts@svfilm.fund. I will review every script submitted.

on submitting

Include (1) your script as a PDF, (2) a short note on who you are and why you’re the one to tell this story — one paragraph, no formal cover letter needed — and (3) any previous work if it exists. Previous work is optional. The script and the note are not.

Submissions close June 10. I will read every script. I can’t reply to every submission, but if your script is the one, you’ll hear from me. All submissions are subject to a standard submission release agreement, which you’ll complete before I read your script.

At the start of this year (February), I decided to get serious about my interest in film. I’ve always been interested in cameras and media, but coming from a lifelong background in engineering, I never really saw it as a viable career path.

Story Company launched in early February, and a media frenzy ensued. Every startup and venture capital firm began to question their media strategy on X & LinkedIn (if it even existed). Video content began to flood the X timeline.

In April, Cluely released their viral launch video, earning 13M+ impressions on X. The frenzy furthered. The X timeline began to fill with poorly lit venture capital podcasts, 1m10s clips, emotionally distant talking-head announcements, and even more launch videos.

The production swell continued until roughly August, by which time launch video quality had significantly declined (and marketing budgets for experimentation dissipated).

At the start of 2025, I predicted a renaissance in film and the arts, but all San Francisco (the innovation center of the world) got was viral launch videos and podcasts.

Moving to San Francisco

I moved to San Francisco in September 2023. I was working on a voice-AI-enabled platform for software engineers to live-practice interviews with AI interviewers. It was the natural extension to the course company I was running at the time, called “Interview Pen” (a clone of my previous company, “Back To Back SWE” which I exited prematurely).

After arriving in SF, I unrolled my $99 foam mattress from Amazon in the corner of my co-founder’s studio apartment in the Marina. I was excited to stay lean and build the product out.

Little did I foresee, 2 guys living on the floor in a studio apartment together, one full-time on the product and another part-time, turned out to be a problem.

Aside from the expected squabbles about living situation, my co-founder had a very high revenue expectation for the company ($40,000/mo+) before he would consider going full-time. Or, he stipulated, we had to get into Y Combinator.

Eventually, I got kicked out. I ended up moving 2 more times that Fall and Winter within the city. Spending 1 month in SoMA, 1 month in Lower Haight.

In the midst of this chaos, I started seeing some videos popping up on Twitter & YouTube from Jason Carman. At the time, video was only a hobby for me; throughout 2023, I was filming vertical shorts, trying to make them cinematic.

I reached out to Jason over DMs and actually ended up as a camera operator on this tweet.

Saturday Startup Stories

Jason Carman single-handedly carried visual storytelling in Silicon Valley from late ’23 to mid ’25, documenting the AI boom as it happened with his series “Saturday Startup Stories,” which featured a new deeptech startup every week.

I didn’t realize how insane a production cycle this was until I began attempting to produce work in February. To publish weekly, you have to be in all 3 production stages at once, at all times:

1. Development/Pre-production: Have a new story located or being researched.

2. Production: Filming the same day another episode is published.

3. Post-Production: Editing next week’s episode.

Maintaining the quality bar S3 had at this speed, while holding full-time work, is commendable (made more possible by Jason’s background as an editor).

The Cultural Core

Throughout this time period, all the way up to Story Company’s launch at the beginning of this year, no one was doing film in Silicon Valley in a big and serious way. It really was just 1 guy vs. the world.

Various YouTube influencers were touching here and there, but very few at the heart of innovation as it was happening, the cultural core of who was doing it and what this place is about.

Throughout 2024, I began to ask myself why this was the case. At the time, I was recovering from my co-founder leaving in January, faltered around in the Spring, before joining The San Francisco Compute Company as a Product Engineer in June.

So much was happening, but why were so few people filming it?

New Media

Jason’s work was the seed for more serious media operations to emerge in a culture that has historically denied image:

• News: TBPN (Oct ’24, John going full-time Apr ’25)

• Print: Arena Magazine (Aug ’24)

• In-House: A16Z New Media (Apr ’25)

• Personality: Ashlee Vance/Core Memory (Jan ’25)

• Podcast: Sourcery (Feb ’24)

“New Media” has always been a funny word to me because it’s just…regular media. We just didn’t publish our own narrative until ’24/’25. Most people in tech (including me) didn’t even know that narrative was deliberately created. It turns out there is a whole industry & city dedicated to only this.

Summer 2025

Between February and June, I was stumbling through documentary attempts and interviews, feeling and discovering my way through every aspect of filmmaking from first principles.

Why didn’t you like the video? The images are nice.. Story structure? Oh, what is that? Ah, screenwriting. Why does the image on that camera look washed out, and the other look crisp? Ah, cinematography? Logistics are a mess and I’m moving too slow. Ah, producing. The clip-on mic doesn’t sound as great as the overhead boom…where do I place it again? Ah, sound. I keep using music libraries and it doesn’t really fit the shape of the video. Ah, score composition. I keep visiting offices, moving plants around, but it never looks quite right. Ah, production design. One person coordinates all of these disciplines? Ah, directing.

Oh, there’s a whole school for this? Film school.

Lighthouse

In May, I began a documentary for Lighthouse (they help workers secure skilled immigration visas, like the O1). I had a general idea of what I wanted to do:

• cut together multiple customer interviews

• anchored by Minn’s personal narrative

• with supplemental commentary from employees on their personal journey getting visas

But I didn’t really understand what I was doing. The overall message, structure, & direction of the work.

I conducted 6 interviews, where each interviewee thought they were giving a testimonial for Lighthouse. But walking in, I wanted to go deeper and explore a narrative of personal struggle, the real truth of the subject (which is just challenging and legally touchy to talk about, especially when in a position of power, talking to someone you just met, when you don’t know where the final video will go).

So the interviews captured watered-down replies, not really going anywhere, not really saying anything.

In the edit, I realized how clueless I was. Piecing together a narrative that I didn’t know, over poorly directed interviews that didn’t have the informational or emotional content I needed to make the video work (and I was doing everything myself: the producing, cinematography, sound, directing, editing — for free).

And after 8 production days and 1 ½ weeks of editing, the project was impossible to complete.

Film School

At this point (roughly in May), I realized I needed to actually know what I was doing. I resolved that I’d learn everything about every subject that touched the medium. So I started reading…

Cinematography, writing, directing, producing, production design, editing. I wanted to learn everything so my inability to “do the thing” would go away. After the Lighthouse shoot I read a whole textbook on directing.

I began watching 4 films a week, covering almost 80+ films since late April. Internalizing story structure, shot grammar, cinematography principles, etc…

I didn’t realize it, but I was back in school again—this time as a first-year film student.

Offscript

Around May & June, I felt a vacuum start to form on the X timeline. Tech Twitter begins to circle certain keywords for 2-3 weeks spans, and the keyword at this time was “storytelling”. No one knew how to do it, or what it really meant. Narrative pressure was building.

One day I saw the post I was waiting for, from Juliana announcing Offscript, which went viral. Offscript’s team composition fits the perfect (only) shape that will work for commercial film production in Silicon Valley:

• Juliana: Bridges the creative and technical world. As a former software engineer at Google + formal MFA in film. Can chop it up with people on tech Twitter. Can think both in impressions + reach + long-term signal, and artistic + aesthetic terms. Without both, everything falls apart.

• Alli: Comes completely from the creative world. Providing creative direction and institutional production power.

• Cam: Silicon Valley moves much faster than the normal production world. Productions need to turn around in days to a week to keep up with tech Twitter. Cameron comes from a rigorous post-production background.

It felt like a mathematician seeing a theorem they’d long suspected finally proven. That’s the only way it could have worked.

It was clear to me that Offscript could not have emerged from within SF, and it deepened my curiosity on why that was the case.

Barriers to a Storytelling Industry

The backstory above sets the context for what it looks like for a lifetime software engineer (with a recently acknowledged strong interest in film) to attempt this rare career switch.

Since February, I’ve kept an Apple Note listing the challenges I’ve faced and how they might generalize to why Silicon Valley’s stories, the ones at the core, have been largely absent from the audio-visual medium.

These will read as a quick list with a conclusion at the end.

#1 — Education

A film education is as extensive as an engineering one. I went through 4 years of Computer Science at the University of Maryland. It involved extensive readings, practice, and hands-on projects.

A film education is no less rigorous. Film students go through a curriculum of 250-400 films over 4 years, getting hands-on experience collaborating with every department on full narrative sets, shooting student films, building countless relationships to specialties the craft requires.

Film as a medium is just as complex and demanding to have control over as the hardest programming assignment I ever did while in school. The complexity is just pushed to different problems (people, organizing ideas, logistical uncertainty, noticing what isn’t there in a performance that should be).

Those who study Film rarely pursue a dual major in Computer Science. Those who study Computer Science rarely pursue a dual major in Film. One is a field of the mind, the other, the heart.

#2 — Exposure

To progress in directing (& producing), you have to have mass exposure to every kind of audio-visual content. Feature films, documentaries, television, commercial spots, YouTube videos (Internet), etc.

This allows you to subconsciously absorb visual grammar (how a visual story converts to meaning, with what efficiency, pacing, rhythm, emotional texture), and build a mental library of preferences you can articulate in your own work.

People in tech (until culture has shifted in the past 1-2 years) don’t watch movies. They don’t watch TV. They’re building products heads-down, 9-9-6. There are many exceptions, but on average, your best backend engineer isn’t caught up with the latest show on HBO Max.

It’s only this year that I’ve started to study feature films, and I have a backlog of 500 films to study.

Studying feature films has been massively helpful for me to internalize visual storytelling. Showing, not telling. Writing with images.

Competence in the medium comes from exposure, which career engineers do not have.

#3 — Thinking vs. Feeling

Narrative film is about emotion delivery and creating change in the viewer. Software engineering is about critical thinking and system correctness + reliability. These are polar-opposite skillsets, attracting very different personality types to either profession.

Film favors the outgoing and personable, software favors the overly cerebral and pensive. Film is based on intuition and feeling; software is based on raw intellect and knowing.

Much of film work is based on not knowing, and instead finding. Almost every interview shoot I’ve walked out of where I “knew I got it” turned out to be an emotionally inert flop; those that I thought were a mess and unsalvageable often turn out to somehow work.

I’ve watched the movie Sinners 30+ times in theaters. Every time, before the show begins, I look around at the people behind and in front of me to collect an image of their emotional journey I can see as the film plays. I contemplate why we’re all there. Why do we pay $10-20 to sit in a dark room, staring at a large screen of moving images?

It is to experience emotional satisfaction. A beginning, middle, and end—that connects to a meaning.

Whether it be emotional catharsis, watching a horror movie, or finding true love, watching a romance. The product is invisible, existing only in viewers’ minds. Like a distributed database of emotional values. Your career as a director advances when you move as many of those distributed values in any desired direction.

On the other hand, startups are about creating tangible products. Those users can interact with and gain value from. Whether software UI or hardware, effort can go into seeing a tangible result.

The result of a film is intangible. You can never touch or see culture. You just feel its shape from your set of experiences with the world.

#4 — Talent Density

World-class technical talent moves to San Francisco to work at or build companies. World-class creative talent moves to LA or NYC to work at the highest level for their specific medium (in film, this is narrative features/commercial work/documentaries).

Any given film work, even a simple talking-head announcement, may touch 6+ creative specialties:

• production

  • directing (observing and managing performance)

  • cinematographer (will handle both camera & lighting)

  • sound recordist

• post-production

  • editor

  • colorist

  • graphics (design + motion)

The density of creative talent for any given discipline you’re hiring for will be a harder search here in San Francisco. Most creative talent physically isn’t here. The best are booked out 1 month. Some of the best cinematographers in the Bay might find themselves in the middle of the pack in strong talent markets like LA or NYC.

The best talent moves to where the best talent is. Where the opportunity to go as high as possible in your career is. And for creative people, that is (currently) not San Francisco.

This creates a vicious cycle. Without concentrated talent, standards stagnate. Without rising standards, there’s no visible career path. Without a career path, talent leaves—and the loop tightens.

Jenny Gao’s timely “50,000 Art Hoes Will Save San Francisco” talk playfully gave a nod to this serious labor shortage. (apparently the talk wasn’t filmed in San Francisco, rather, Sydney, Australia)

#5 — Collaboration

Film is a highly collaborative medium. It involves communication between many parties: subjects, clients, vendors, locations, crew, and your own creative staff. You have to talk to a lot of people. You must develop strong interpersonal skills.

Most people in tech (and San Francisco) are introverts. They can start beef on Twitter, but not go outside for 2 days. The city attracts those who were quiet or outcasts in some way early in life, and who are now on a revenge arc (which manifests as company-building).

A majority of software work can be completed individually (with more or less cross-functional collaboration required, depending on org and project size). An engineer can code up a backend endpoint, deploy, and it can stay operational for 4+ years—without talking to anyone.

Film work is a process of constant handoffs and clarifications. From project inception to completion, a never-ending alignment between stakeholders and creative staff. Being a clear communicator is a requirement.

This makes film talent less likely to emerge from those who naturally gravitate to San Francisco for a tech career.

#6 — Visibility

Film is a very visible medium. People bet on people, names they trust. Trust comes from the belief that someone can survive 10,000+ creative decisions and still get it right. So name & face become as much the product as the product itself.

Chain of Visibility: Actors are the most visible, since they are on-screen. Directors are visible as the creative authors of the work. Producers are visible for making the thing happen (except more in industry circles than general public awareness). Almost every other role: cinematographer, composer, etc. also maintain industry visibility.

Visibility is how you survive. It’s how producers get sent projects to produce. It’s how directors get projects to direct. It’s how composers get projects to compose. The industry runs on being seen—being easy to locate & say “yes” to.

Software engineers like to be invisible. They leave their GitHub avatars unset. They set their Twitter avatars to an abstract gradient. Say the most with the fewest words. Efficiency. Economy. In an ideal world, they want to make $500k+ TC with no one knowing their name. Why would you need to know my name?

Striving to be visible signals incompetence; the best produce the most while being completely invisible. Being invisible is high status.

This cultural attitude, this survival mechanism—having been ingrained across a lifetime, is almost impossible to unlearn.

#7 — Income

Software engineers make a lot of money. Over $200,000/yr for good roles, up to $500,000/yr+ into senior roles. Your first year as an independent director, you might be lucky to take home $20,000 (a 10x decrease)—after living expenses, investing in gear, building a portfolio of work, investing time learning, etc.

Throughout the year, I would help Sonith (from ZFellows) and 50y film their podcast + a mess of other freelance work. This allowed me to just get by and spend time studying and developing my own projects.

Arlan

By late July, I realized I wasn’t being intentional about the projects I was taking on. So I decided to produce a proper interview with an actual set to show I was capable of working with a crew.

I had a concept for an infinite white room I’ve wanted to do for a while. I knew the project would take some investment so I waited for the right person to do it with. Eventually I stumbled across Arlan (from Nia) on my timeline. I DM’d him.

In August, I produced the project, assembling the crew and going way over budget.

It cost ~$6,000 to produce. Just a simple infinite white room concept. Film production is more expensive (and inefficient) than most people in tech realize.

The below is a breakdown:

Across the final full interview (61k+ impressions) + 4 clips (124k, 30k, 14k, 1k…) the content earned over 230k+ impressions. At $15 CPM (not sure on Twitter’s exact pricing), a ~$3,500 return in earned attention. So not that terrible. I wasn’t producing this for the reach, it was a portfolio item to show I could think cinematically & direct a crew.

Leaving a high-paying, predictable, stable career for a capital-intensive, emotionally intensive, unstable career you’re not even good at yet—doesn’t make much logical sense for anyone with a strong tech career.

It especially doesn’t make sense during one of the greatest wealth creation events in technology history (’22-’25).

I’m ending the year still barely getting by in income (mostly because I hadn’t decided to commit to the career until a month ago and kept turning down paying projects).

#8 — Switching Costs (Skill)

I started programming when I was 16. I wrote Android apps for my Samsung Galaxy Note 4. I remember the first time I ran Android Studio, and a blank screen with “Hello World” text in the center popped up. It was magic.

During high school, I would take lessons on Treehouse, learning JavaScript, C#, Java, frontend frameworks, backend—I wanted to know everything.

Most programmers spend years honing their taste in code and patterns through self-education, projects at work, and code review. An eye for good code takes at least 4+ years to develop. Frameworks and languages can take 6+ months to pick up and become professionally proficient in.

Switching to film, you throw all of that specialized knowledge away. None of it matters. It’s a completely different set of skillsets (most interpersonal), in a completely different medium, with completely different goals.

I’ve “thrown” a decade of specialized knowledge away in full-stack engineering. From 16 to 26. “Thrown away” a 4-year Computer Science degree that defined who I was.

It’s been a challenge for me to write screenplays in plain English after spending half a lifetime in VSCode/Cursor. I always feel the pull to open an IDE…so I can feel immediately “useful” again. To feel that sense of control from running a program, it not working, fixing it, then running again.

What’s helped me is realizing that in film, the same architectural rigor exists, except in a different way. Instead of software architecture & abstractions, you have story architecture & dramatic beats. The architecture is emotional and event-based rather than in code.

#9 — Learning Curve

Beyond a formal film education being very helpful, much of film work can only be learned through hands-on experience. Working with documentary subjects from behind the camera or narrative actors on a set. Trying—seeing what works and what doesn’t in the edit. What wasn’t there that we want? What was there? Over and over again.

Technology workers do not have this required experience. You have to do the thing to get it. But you won’t be doing the thing if your career is not in…the thing.

Since February, I’ve grappled to understand story structure. In March, I thought I had it, but I didn’t. In May, I thought I had it, but I didn’t. In November…

Only 9 months later, after extensive study and painful learning experiences, do I feel I have a strong, intuitive grasp of dramatic structure (at least at a 2nd-3rd year film school student level). I can look for it and see it in any narrative work. I can sense emotional texture—where it sits, where it should sit.

This intuitive nature to film takes a decade+ to develop a sense for at a subconscious level.

#10 — Identity

To switch from a technical career to a creative career, you have to undergo an ego death. You have to shift both your internal perception of who you are, as well as your external brand of who others believe you to be professionally. This can take years.

Technical peers look at creative pursuits with skepticism. There is a “waiting to see if they make it” dynamic always at play.

I’ve watched my previous network of technical peers stop texting me, stop reaching out for coffee chats. There is just not much of an overlap anymore to maintain constant communication. Our paths do not intersect to an extent where value exchange can go both ways.

What I’ve found is that the network just turns into creative people. Writers, editors, cinematographers, other directors. But that transition is a painful and lonely one.

This happens simultaneously to the period of time when you aren’t yet great at the thing (and financially struggling). To become great, you have to publish publicly while still an amateur. You have to embarrass yourself to progress.

A software engineer always has to be aware of the tools they have to solve a particular problem. New tools come out every month, you try them on the weekend, tinker, test, and tweet about it. It’s how you progress in your career, stay relevant, and survive.

I had to become OK with not being on that cutting-edge anymore. Not knowing the newest database or agent framework. This was especially difficult during ’23-’25 when so much was happening in software. I can still understand concepts—I’m just not embedded in the world day-to-day. It is impractical and impossible to do so as a filmmaker. You have 100 years of cinema to catch up on…

This is something I’m still not totally over.

#11 — Builder Culture

Until 23/’24, the culture of San Francisco was a “shut up and build” culture. If you were talking, if you were public with your image, you must not be building anything that important. Being visible and perceived as vain meant career death.

With AI making the building of many mid-tier consumer software products accessible, the challenge shifted to distribution. Getting people to know about the thing. Why the thing matters. Storytelling.

A whole industry realized a massive surface area of attention they were ignoring (X & LinkedIn). 1m10s podcast clips were getting 300k+ impressions on X (300 * $15 CPM → $4,500 in raw attention sitting on the table if you take 20 minutes to edit out a clip).

This happened in parallel to Jason Carman ripping a documentary a week. People in technology began to see the value in seeing your culture accurately represented in media. The value of scaling image.

Every cultural movement is a time that comes and passes. Sometimes it just isn’t the right time for a certain idea, or movement, or work of art. This was Silicon Valley pre-2023.

A cultural window is opening.

#12 — Industry Awareness

I’ve never been to LA. Until February of this year, I wasn’t even exactly sure what Hollywood did. I watched movies as a kid, but I never contemplated further to seeing myself having a career in film.

2 major industries, the entertainment and technology industry, the former valued at $2-3T, the latter $5-8T (thanks Claude), are almost completely unaware of how the other works.

And they aren’t really interested in figuring it out, either. For creative work, humans are at the heart of the work. For technical work, technology and systems are at the heart of the work. The best creatives are (usually) the least technical, the best technical talent is (usually) not the most creative.

Silicon Valley is insular in culture. Founders talk to each other in “group chats” (I’m apparently not in any of them..), 1 Twitter post can take over the narrative of a whole industry for the day, <100-1,000 accounts set the agenda for 500k-1M+ onlookers. We don’t watch the Super Bowl.

We’ve traditionally been culturally tone-deaf and closed off.

On the other hand, entertainment is all about culture. Its creation, transformation, death, rebirth.

#13 — Access

Filmmaking runs on access and trust. You can only tell stories with people whom you have access to, and who trust you enough to step in front of your camera.

When someone wants to do content work, they will ping their network and go with the closest, safest bet. Other engineers, founders, investors—who have an interest in media. But not filmmakers. Filmmakers are not in the network. Local agencies and freelancers can do the work, but they’re not technologists.

Because of this, some of the best visual artists in the world are unable to access Silicon Valley’s stories. It’s both a network and communication problem—visual artists aren’t in-network, and most don’t have the chops to go back-and-forth on both strategy and artistic concerns.

The best artists live in a world of feeling that isn’t grounded in pushing metrics or business results. You can’t put up a job listing for “storyteller” and expect the applicants to be good; you have to identify and mobilize visual arts talent behind a fresh narrative they own.

To tell Silicon Valley’s stories, you have to become an insider.

#14 — Running a Production Cycle

To create a culturally significant media operation, you need to create a production machine. This can look like (1) timeline takeover, posting multiple times a day: clips, longform, live (like TBPN), or (2) infrequent & value-dense, releasing 1 magazine a quarter (like Arena Magazine).

In either case, you must be consistent in the release cadence your medium requires of you (while maintaining tone and quality), and the production operation must be designed to run forever (while adapting to culture).

Prospective media operators in SF do not know how to run a rigorous and logistically intense production operation. This talent exists in LA and NYC.

This creates the grand irony of TBPN being based in LA instead of San Francisco.

#15 — Motivating Creative Talent

Motivating creative talent is different from motivating engineering talent. (The best) creatives care about:

• constantly going deeper on their craft

• doing culturally important work

• seeing alignment between the director/founder and the story

• feeling the director/founder understands the technicalities of their craft

• fair pay, a respectable living

• loving who they work with

• surprising themselves (and others) with what they make

Engineers care about similar things, but creatives are much more focused on skill & process as output, over clean-cut salary, stock package, and bonus compensation. World-class creatives don’t really care about that; they’re obsessively craft-focused.

An engineer can not care about the company, make $500k TC, and go home feeling fulfilled. No matter how much you pay a creative, if they aren’t connected to the work, pushing the boundaries of their skills, pushing their understanding of their medium, great work won’t happen.

I always make it a point to study all of a creative’s work before meeting with them, so I can articulate back to them exactly what I find interesting or applicable to a project. If it’s a cinematographer, I’m watching every second of footage on their Instagram and website, looking for colors they like, themes, moods they gravitate to, etc.

#16 — Unemployment

Creative work is project-based. Very talented people can go weeks without work—not because they’re struggling, but because that’s simply how opportunities flow. Dry spells are built into the career.

In technology, unemployment puts you at zero status. You’re overlooked after introductions at parties, your friend group stays casually aware of it, companies become reluctant to hire the person someone else hasn’t hired. (unless, of course, you were just building your own startup all along…)

When you can attach your identity to a powerful company, you become someone in the ecosystem. You have leverage: relationships & information. You can get someone else a job, share hard-earned lessons from work on Twitter, angel invest in others with your income. You can give people something.

A creative between projects finds no social place in tech. They don’t have a “prev. OpenAI, Stripe, Uber” to connect in conversations. Their skills get overlooked, they can’t find many peers to lean on, and the housing gets too expensive.

So they leave.

Initiatives like “The Intersection of Art and Technology”, TIAT are safe 3rd places for creatives in tech. We need more of them.

#17 — Drama

Story comes from drama. Characters collide with reality and make critical choices that define who they are. The quality of a film is directly tied to how vulnerable, real, and open people can make themselves.

There are varying degrees of drama, from bickering with your roommate over food they stole from your side of the fridge—to a sprawling downtown shootout in LA involving dozens of cops, people falling left and right.

In either case, the drama is rooted in true events and not fabricated. Fabricated drama creates false meaning; it creates nothing.

Many founders and investors are not ready for the level of vulnerability the best storytelling requires. Revealing struggle publicly makes you weak—a target. When in reality, vulnerability and relatability are core to the empathy mechanism that makes film work—letting others see themselves in you and your story.

So some of the best stories still stay hidden, and much emotional landscape has been totally unexplored.

#18 — Dreams

The dream people move to San Francisco with: build a high-growth startup that gets millions of users, earn respect among industry peers, and make a lot of money.

The dream people move to Hollywood with: direct, produce, star in a feature film that has widespread cultural influence and impact. Maybe get famous. Also make a lot of money.

Filmmakers want to make films. Founders want to build a product.

A film studio is not a startup. Startups grow fast. Film studios, unless they take on capital or quickly create original IP, do not. They are service businesses.

Top startup talent is not interested in creating a service business. They want to swing for the fences. Top creative talent is not interested in a high-growth venture; they want to produce the best art.

Two opposite motivations.

#19 — Platform Gap

The world’s greatest visual storytelling talent is on Instagram or YouTube. Sometimes, with a presence on X. Some aren’t even online in any capacity.

The technology industry is on X.

A composer friend who scored 2 of my (very early) documentaries—I met at San Francisco’s first 222.place event doesn’t have any social media except LinkedIn (which he doesn’t maintain).

Talent doesn’t even get to see the opportunities for storytelling here—because they just aren’t on X.

#20 — Cinema vs. Socials

The “laws of gravity” for attention in cinema vs. Instagram vs. Twitter are completely different. In cinema, you have your audience captive for 2h10m. You get them through the door with a compelling premise, but once they’re in their seat, they have 1 choice for what they can watch.

In a theater, you can spend time building long character arcs, write drawn-out dialogue exchanges, build suspense with score and silence. The watcher is glued to their chair; nothing is competing for their attention in a totally dark room.

Twitter and Instagram have completely different rules. We’ll focus on Twitter:

• Default Muted: Videos are muted by default; you have to lean into visual storytelling in the first 4 seconds to get someone to stay & unmute. Clearly seeing the face of someone talking to you vs. abstract visuals will outperform.

• Contrast: Quick cutting visuals, movement, & color contrast stand out on the timeline. Many Twitter users in tech watch in dark mode. Posting a white GIF or video in a bright white room increases the contrast of your video against other content competing for attention.

• 24s-1m30s Video Length: Drop-off on a Twitter video can be 60%+ of watchers after 20s. Clips under 1m30s with a story arc see the most engagement. Information-driven content can budget 30s-1m for maximum engagement. Anything longer and shareability drops off steeply. Viewers will need more context or pre-existing trust that the content is worth their time.

• Emotion Above Everything: Making watchers feel something is all that matters. High production value is downstream of triggering a feeling of presence and immediacy. “This must be important.” Changing emotional charge every 10-15s. A video recorded on an iPhone can outperform a $100k+ commercial if its emotional architecture is stronger.

Many classically educated visual storytellers won’t have hands-on experience with platform idiosyncrasies.

The best tweets are immediate to comprehend. One morning I saw a podcast from Anthropic and decided to scribble a few slides breaking down the cinematography. It took me maybe 20 minutes. The post went (semi-)viral. It was well-timed, 1 sentence, 4 skimmable photos, and information-dense. You could comprehend the message in <10s. 221k impressions * $15 CPM = ~$3,300 in earned attention, just from 1 sentence and 4 scribbled slides.

You can spend 2 weeks, and thousands of dollars, making a documentary and it could only hit 5k in reach on Twitter. Socials don’t reward cinematic craft, they reward engagement. The craft has to be re-engineered to fit the physics of the platform.

#21 — Speed

Film and technology move at 2 different paces. News on X changes daily. Anyone can launch any day and reach everyone. Competitive landscapes are reshaped every week. You have to move fast.

Traditional film moves as fast as it can, but not at startup speed. A director may research a feature film for 2-3 years. Some scripts stay in development hell for a decade. The average time for production of a feature film from concept to screen is 2-5 years. Crews need to be booked 2-3 weeks in advance, sometimes months.

Truly great stories transcend time. They explore deeper, human-centered themes with no expiration date. The films that stick are replayed in theaters forever.

Some anniversaries I watched this year in theaters:
— 50th: One Flew Over the Cuckoo’s Nest
— 50th: Jaws
— 41st: This Is Spinal Tap
— 40th: Back to the Future
— 40th: The Breakfast Club
— 30th: Apollo 13
— 15th: Black Swan

Startups have an expiration date; it’s always around the corner. They’re default dead.

They can’t afford a 3-month production cycle for a brand documentary. So the types of productions that make sense for startups are more limited, and need to stay lean and move 2x as fast.

The medium rewards craft, perfecting every detail in the image (costume, production design, lighting). This is how artists move forward in their careers and get respect from their peers.

Startups don’t have time for that.

#22 — Showing vs Telling

Cinematic storytelling is all about showing, not telling. Technology Twitter rewards long, cerebral rants (like this Substack post) that dive deep into a train of thought through written language. The visual medium rewards you when you deliver images and create watchable results.

A great example of visual storytelling is the first 2 minutes from Back To The Future. I’ll just let Claude explain:

You get big on Tech Twitter by constantly having something new to say. Pushing for an intellectual edge, knowing what comes next in 6 months. This intellectual mind-dumping actually hurts you in writing for the screen.

#23 — Cinematic Opportunity

Software engineering is not very cinematic. The process does not have many visual elements that you can anchor an effective visual story against. You have:

• environment

  • an office (hopefully it looks good)

• individual

  • wide: someone sitting, typing at a laptop

  • close-up: hands typing on keyboard

  • extreme close-up: eyes looking around the screen

  • medium close-up, over-the-shoulder: typing at a laptop

  • many more ways to say—typing at a laptop

• team

  • wide: collaborating at a whiteboard

A very limited shot grammar. You can do sitting or standing interviews, have people walk through the space—then you’re out of cinematic options (so you add motion graphics to spice things up..) So your remaining tools are to focus on character arcs, interpersonal conflict, and situation-based drama.

Deeptech has many visual anchors: bio labs, warehouses, robots, rocket engines, satellites, the open ocean, defense testing in the desert. The whole physical world is your oyster. Things explode, fall down, break, make noise, talk.

Opening shots from S3 Episode 41 with Extropic:

Pure software development happens in 1 place, an office. And the product is invisible.

This does not inspire great visual stories, so they do not get produced.

#24 — Physicality

Film work is physical. Going to locations, up and down flights of steps, doing scouts, meeting people, moving heavy gear, setting up lights, tearing everything down, offloading SD cards at home. It is hands-on work, physical work. Only becoming digital later in post-production & distribution.

While I was an engineer, I would be at my laptop 8-10 hours a day, otherwise in meetings or commuting. There is no physical, mobile aspect to software development. Some engineers are very comfortable with that controlled, sedentary aspect to the work. Relaxed, plenty of room to think.

Film production doesn’t offer that luxury. You’re constantly going from writing in solitude to filming on location, and back again. Constant context switching between very offline and very online. In traditional film, production days can be 12hrs+ long.

You can’t write an automation from home for the physical work. It’s structured chaos every time.

Career Engineers Who Became Filmmakers (Few)

To my knowledge, in film history, there hasn’t ever been a mainstream director who came from a sustained career in software (>2-3+ years) and successfully transitioned to film.

There are only 2 prominent cases that come somewhat close.

Mike Judge (creator of “Silicon Valley”) worked in tech for 3 months before leaving to become a bass player:

Shane Carruth, a software engineer for 2 years developed flight-simulation software, made 2 films before ultimately exiting the industry:

Otherwise, it’s never happened.

Just a Few Engineers

In September, I was talking to Donald Jewkes at Cursor’s studio in North Beach. The room was quiet, tables overturned, lights off. Sunlight from the windows poured in, bouncing off the floor.

I explained my apprehensions about leaving engineering for film. I was already 8 months in, the decision was effectively made. But I was struggling—for all the reasons I’ve just listed.

Ben: “I’m worried if I keep at this career, I won’t earn as much as I would in engineering.”

Donald: “I mean…you can make the same amount doing film.”

Ben: “…but there aren’t really any software engineers who are filmmakers here. I think it’s just us.”

Donald: “…”

Donald: “Yeah, I think it might just be us.”

Those here—those who came here as career engineers—are up against incredible odds to become storytellers.

An entire industry must form. Podcasts are not enough. Corporate video is not enough. 1m10s clips are not enough. 2 film studios are not enough.

We need a talent pool of professional actors, writers, directors, cinematographers, producers. All tech-aware. Living here. Many productions running. Forward career mobility.

Silicon Valley must begin to produce great storytellers.