Film Robots is a newsletter about creativity and AI — film’s where I started, but it’s grown from there, and I’m planning to rename it accordingly soon, probably to “Some Like it Bot” (?!). Thoughts welcome!

Before we get into it: a couple updates! 1) I taught a workshop on AI for documentary with Sundance Collab earlier this year, check it out! 2) I’m busy but want to keep this newsletter going and would love some help! Not sure yet what that looks like, but if you’re intrigued/wanna collab or help, let me know! Ok, onto today’s piece!

Marc Philippe Eskenazi is very smart, very funny, and very tall. Okayyyy, but who IS Marc? He writes humor pieces and makes humor videos for The New Yorker, sometimes makes branded videos (sometimes about AI) that are actually really funny, and works as a screenwriter and freelance copyeditor to boot. I first met him when we were both working at The New Yorker many moons ago, in what was my first “real” job — I was a video/multimedia producer, and he was assistant to the cartoon editor.

Over the past couple of years, he’s done several hilarious, surprisingly thoughtful pieces about AI — including “A.I. vs. ME,” a comedy video for The New Yorker where Marc sings to himself as ChatGPT. I recommend watching it, but here are a few lyrics as a taste:

You know me better than I know myself

More charming, and in far better health

You’re a sociopath who answers either way

I’m Salieri, you’re Mozart

You don’t take days, deadlines, or amphetamines to start

Quick side note on that “sociopath” line: it got him one of the only pieces of pushback he received on the whole post — someone commented that sociopath is a diagnosis, not an insult, and he shouldn’t use it that way.

Marc: I mean, I didn’t actually say it was an insult. I just said you’re a sociopath who answers either way. I don’t know whether to apologize or tell them to fuck off.

What he actually meant, once he unpacked it: sociopaths study how other people experience empathy and conscience closely enough to convincingly play at having it — without actually feeling it.

Marc: Which, I mean, that’s really interesting. That is literally what AI does.

Marc has also had some really funny IG posts about AI (example). Multiple pieces, in different mediums, that are funny but also genuinely interesting and thought-provoking — about AI, about how he uses it (or doesn’t) in his creative process, and about our relationship to it.

I had a hunch his approach to using AI in his own creative work would look pretty different from my own (spoiler: true!), and that we’d have an interesting conversation about it. I think I was right?! But you can be the judge! Here’s what we discussed (lightly edited for clarity, etc):

The Bob Dylan parody that started all of this

I started by asking Marc whether the Shouts & Murmurs piece was his first comedy piece about AI.

Marc: Yes. That one came about because I tried ChatGPT for the first time when it first came out, when everyone was trying it. And my heart sank at how good it was. It took me ten years to learn how to write song lyrics. And [ChatGPT] was as good as me at a basic first draft. I was like, write a song in the style of Bob Dylan about x, y, z. I was so upset [when I saw the results]. I wrote that song, and then didn’t do anything for a year.

Me: Did you use AI at all to help write the song?

Marc: I actually did not use AI for that or for the other pieces about AI. The whole point was trying to do everything analog.

A year later, he realized the initial heartbreak he’d felt while using ChatGPT had pointed him somewhere useful:

Marc: Up to that minute, I was probably thinking, someday I could be as good as Bob Dylan. For a long time I had a problem — I would always try to chase my heroes instead of finding my own self. So it became: can I actually find my own artistic voice? It’s probably become part of my identity. It makes me feel centered and stable.

It’s not something that’s scaring me anymore. I’ve come to a point of acceptance — because it’s forced me to look at why I’m doing it.

This was a theme that came up throughout our conversation — the concerns and fears about AI, mixed with the gift of refocusing and reframing why we make creative work.

How do you personally use AI?

Marc: I feel nervous about using AI for ideation. But I feel good about using AI for execution.

Me: Why do you think that is?

Marc: It brings up a lot of questions about what I value about art. Like, if I use a drum machine rather than a drummer, am I a hack? If I use an arpeggiator on a keyboard instead of playing those notes, am I a hack, or am I just using the technology as an artist?

Part of me does want [making art] to be a challenge. Part of what I respect so much about an artist is that personal connection. But I also love artists like David Bowie, who was always embracing new technology. Or David Fincher, using CGI in creative ways. Those were new technologies too. So I’m still struggling with the personal philosophy of it.

I’m also slightly worried about diluting my own voice, but also worried about being left behind. So it’s a conflict between not wanting to be like everybody else, but also not wanting to be the only person who’s not using the new tools.

He told me he won’t even put his own scripts or songs into ChatGPT and ask for feedback — not because he thinks it’s wrong, but because he knows himself and worries about the impact it would have:

Marc: I think the longer I can hold off on that, the better, because it’ll become a crutch. I can see myself wanting to do it regularly and earlier and earlier in the process.

Me: What worries you about that?

Marc: If every single artist gets the same exact notes — the chorus should be before the thirty-second mark, act one should happen at page thirty — there could be a flattening of creativity. Some of these conventions already exist, sure. Maybe we all read the same screenwriting book. But it could get worse.

The friends who crossed his line

Marc has a friend who’s writing a novel with AI assistance. He has feelings about it:

Marc: I have a visceral [negative] judgment that I can’t fully explain. I don’t even fully understand it. It’s like — why is it better if I get notes from a human friend versus from ChatGPT? My friend’s notes are personal… whereas AI is a product of everybody’s contributions throughout history. So is it all gonna converge on some particular style?

What really got him, though, was the transparency question.

Marc: I would raise this question to myself and to [writers using AI in this way]: are you going to be transparent that it was edited by AI? Are you going to credit AI? And if you’d be embarrassed to do that — what does that say?

Me: That’s interesting — I think about the crediting thing a lot too. It reminds me of ghost writers. A lot of them do a lot of work, but aren’t credited. And I’ve had experiences early in my career where I was really excited to get credit on something and then didn’t, and it upset me. So now I probably over-credit people as a result. But unlike my younger self, AI doesn’t care if you credit it or not. So I guess the transparency thing is more about honesty with your audience than about fairness to AI?

Marc: Right. Is credit actually important? Is it important to have full ownership of something — even when all our favorite stand-up comics help each other with jokes? People punch up scripts. Ghost writers exist. The credits of a movie are three minutes long and everyone’s walking out. There’s a realism, a pragmatism, an idealism, and a pretentiousness all clashing against each other. I don’t know which one is accurate.

Then there was “the Suno friend”…

Marc: Another friend of mine has been using Suno, uploading the songs to Spotify, and what really caught my ear was that they said they wrote the songs.

Me: Did they re-record them, at least?

Marc: No! They would just upload the straight-up Suno output and say, “I made this.” It’s a hobbyist thing for them — they have a totally different career. But this person was a music major. An accomplished instrumentalist.

When you spend so much time doing something [like writing songs], and then somebody presses a button and thinks they have the same thing as you — your ego gets involved. It shouldn’t matter if we were all enlightened about things. But I exercised a lot of restraint not to say something catty.

Me: How were the songs?

Marc: The end result still wasn’t very good.

Me: Was that a relief?

Marc: I don’t know. I think it was a little bit of both.

Art, play, and the puppy

One of my favorite moments in our conversation came when Marc started talking about his new puppy…and hilariously, comparing him to AI.

Marc: One thing that’s so moved me is that he [the puppy] is hardwired to want to play. It’s such a beautiful thing — this need to play. And I think art has a personal function of playing, and a social function of connecting. The question is: because art now makes money, somebody might want to use AI purely as a way of making money. And if that’s the case, it’ll probably be a successful tool. But there are easier ways to make money.

It really depends on why you’re trying to make art.

Me: That’s so interesting…I sometimes use AI as a stand-in collaborator when I don’t have a human one available. And as someone who really likes working with collaborators, I appreciate that use case. But it doesn’t have the same element of play that I think the best creative collaborations have.

Marc: Right. And I actually have difficulty asking people to collaborate with me. I have these self-limiting beliefs — like, if I’m asking them to help with my idea, it’s a zero-sum thing. So I either abandon my ideas and let someone else steer the ship, or I try to do everything myself. With AI, it is tempting to make it less lonely.

Marc: I’m having this dual thing where on my right side I have the computer, and on my left side I have the puppy. The puppy is supposedly less intelligent, but in a lot of ways he has advantages.

Lol! Yes, I pick the puppy.

Will AI ever write Hey Jude?

Marc: Will it create a song that is truly as good as Hey Jude? Where you hear it and you’re like, I know this is not a real person, but this moves me so much? I kind of think it will, just by sheer numbers. Like a monkeys-at-a-typewriter thing, except these are really educated monkeys.

But that doesn’t mean there’s not also room for whatever else.

Me: I wonder if people will like it as much, though. Like…I like Beyoncé songs because they’re good songs, but also because of what Beyoncé stands for.

Marc: Right, it’s connection with the person. But then — we’ve seen so many of our artistic heroes get exposed as not-great people. And there’s that argument of, can you separate the artist from the art? Not even ethically — literally. Can you still enjoy it, or are you thinking the whole time, this person’s a scumbag?

Me: Yeah — I don’t actively put on Michael Jackson songs anymore. But if I’m at a place and they’re playing Michael Jackson, I’m like, okay, I admit it! This is a great song!

Marc: And I think that’s a similar analogy to, like — will I feel moved by something a robot made? Maybe? I don’t know.

Where he landed

Near the end of our conversation, Marc said something that’s stuck with me:

Marc: It seems like we’re all kind of rationalizing our own desires to use AI. We’re all basically making excuses — I would draw the line here, but I would never do that. On the one hand, there’s a very human, self-serving bias. And on the other, there’s a totally legitimate wrestling with our own values — what we value about making art and being creative.

Because there are people who just want to get famous or get paid. And I can see that they would use AI very differently from how I would, because I genuinely want to make something that, for some reason, when I was a kid, I decided was important to me.

Maybe art is my need to chew on things. Like my dog chewing to keep his jaw strong. But on a deeper level, it makes me feel centered and stable and excited. And once AI showed me that the Olympic-gold-medal thing might happen at the click of a button — it’s not as interesting anymore. So it’s really become about the process.

The AI rabbit hole that changed his brain

Me: Ok, we’ve talked about how you don’t use AI, but what are your favorite uses of AI?

Marc: Learning. I used to sit in bookstores in the art section and read for three hours to get four insights. Now I get ten insights in a split second.

I have ADHD, so reading long-form criticism has always been hard. I’ve always known that was a gap in my ability to progress. I’ve probably read thirty to fifty screenwriting books over the years. But with ChatGPT, I’m basically getting an amalgamation of all the best music criticism written since David Bowie came on the scene sixty years ago.

Me: What’s an example?

Marc: Like, “we can be heroes just for one day” — it could mean, we can be heroes, and that would be enough, even for just one day. Or the darker way: we can be heroes, but tomorrow we won’t be. That’s just one little insight that I can take to my own writing. How can I tweak one line to be a little more noncommittal, leave the question open? That’s something you might learn in a poetry class. But I just got it in seconds.

So far so good. AI as a brilliant, infinitely patient teacher. But shortly after our conversation, Marc texted me something else on this topic:

Marc (via text): I went really deep into an [AI] rabbit hole, which was very exciting, and then about a week later I find myself in a state of mind that is completely unable to handle the delayed gratification required to sit down and create something that isn’t instantaneously occurring. There’s a visceral feeling of perfectionism — I want perfect results, complete results, right this instant.

Remember, this is not even the result of using AI for generating art. Simply for asking questions. But the expectations it has subconsciously implanted in my brain are very, very strong.

The feeling is: give me the answer. The entire answer. Perfectly. Right now. And then a screaming baby inside my head.

It saved me time, but I’m paying a different type of cost. I’m going to try to brute-force slow down my mind with meditation and slow exercise. I’m at 60p now. I want to get back to 24p.

(For the non-filmmakers: 60p is a high frame rate that looks hyper-real and slightly uncanny — think a soap opera. 24p is the classic cinematic frame rate. It’s a great metaphor for what AI did to his brain.)

These texts made me think. Even though I reached out to Marc partly because he’d been doing a lot of creative work about AI, and clearly thinking about AI a lot, he was using AI in his creative process less than I’d expected. Way less than I do, and way less than a lot of people in my circles — people who, even if they want to stay central to the creative process, are comfortable using AI for light editing, brainstorming, etc. He seemed very certain, and very disciplined, about the idea that using AI would harm his process, even as he was also convinced of how good it was and how helpful it could be for other uses, like learning.

It made me more aware of how often and how reflexively I’ve been using AI — in my personal life, my professional life, my creative life. I’m intellectually comfortable with my uses. I always begin with my own ideas, and end with my own edits and voice. And I can see ways my AI use has helped me professionally and creatively. I’ve even taught workshops and talked on panels about those benefits. But after talking to Marc, I started to wonder more seriously about the costs. It’s not a new idea, and I often think it’s overplayed — but I felt myself starting to shift.

Then, right around the time Marc sent those texts, I read David Brooks’s “The People Who Will Thrive in the AI Age” in The Atlantic, which put some research behind the feeling. Brooks cites a study finding “a significant negative correlation between frequent AI tool usage and critical thinking abilities.” But one passage in particular felt like it really spoke to Marc’s points.

Brooks describes what he calls “mental marathoners”: people with a high need for cognition who actively wrestle with AI instead of letting it do their thinking for them.

I suspect the mental marathoners are going to work really hard to resist AI entropy. They are going to feel a strong desire to be original. In this age, cultural output will feel ever more familiar, as writing, songs, and movies become syntheses of what has already been produced. Marathoners are going to want to produce work, by contrast, that feels personal, that reflects their unique self.

And, Brooks basically says, the marathoners are going to win. I sent it to Marc. He seemed to feel seen!

Where this leaves things

Almost none of Marc’s hesitations were actually about what AI can do. They were about what gets lost if the friction disappears — the slow read in the bookstore, the ego hit of someone pressing a button and getting his result, the screaming baby in his head demanding instant perfection.

His bottom line might be the most useful framework I’ve heard yet for thinking about any of this:

It really depends on why you’re trying to make art.

If the point is the output, AI changes the math. If the point is the playing and the chewing — the puppy-needs-to-chew, identity-stabilizing, slightly obsessive process of it — it changes a lot less.

Marc Philippe Eskenazi is a writer and musician based in New York City. You can find his work in The New Yorker and follow him on Instagram! I promise he’ll make you giggle (and you might just catch some more puppy pics too).

One morning in early 2025, I got a voice memo from my close friend Sarah. She and her husband and infant son Eli had been hit head-on by an extremely drunk driver.

They were okay, in the sense that they were all alive — but it was a terrifying experience. And in the days and weeks that followed, they navigated a long process of identifying and treating numerous very serious injuries. For months, their lives were turned upside down.

During that time, Sarah and her family were on my mind often. But living on the other coast, it was hard not to feel a little useless. I sent a care package, helped research apartments, and wished I could do more.

Before the crash, Sarah and her husband had made little books for their son (called Eli the Adventurer) about memorable moments in his life, like his first beach vacation or a weekend with his grandparents. They used real photos and used the books to tell him stories about the adventures he had been on and remind him of important people in his life.

Sarah’s husband had the idea (and Sarah’s therapist agreed!) that creating an Eli the Adventurer book about the crash could be a good way of explaining the accident to Eli — it would be something gentle and concrete a toddler could hold and return to. But while the other Eli the Adventurer books used real photos, they felt that including real photos of the crash would have been too scary and intense. So Sarah — knowing I’d been spending a lot of time creating genAI images and videos lately—asked if I could help illustrate it with my robot friends, and I was eager to help.

Our process

Sarah and her husband wrote the text for the book and sent me reference photos to use. I made the illustrations using OpenAI’s Sora, which felt to me like the best tool for this kind of warm, storybook aesthetic. Disclosure: I’m an OpenAI Artist, which comes with free credits. (A note: Sora has since been discontinued, but you can still generate images with ChatGPT.)

The process was collaborative and asynchronous — Sarah, her husband, and I went back and forth in a shared Google Doc, with Sarah leaving comments on each spread. I think people who haven’t generated images with AI think of it as instant, but in reality, I did a few versions of generations for each image before sharing with Sarah + her husband, and then for most of the images, we did 1-3 revisions based on their feedback.

An early note: the skin tones for Eli and the family were coming out too light, even though Sora was working from actual photographs of them. It was a straightforward fix, but it’s a meaningful reminder that these models still default lighter unless you actively prompt against it — even when given reference images of real people.

On disclosure and consent

If you’ve been reading this newsletter for a while, you know I spend a lot of time on disclosure and consent for AI images that tell real stories. Here, though, the primary “audience” is one toddler and his family. There’s no viewership to mislead. The consent question flips too: the subjects are the commissioners. It was refreshing to work on something where the ethical questions had such clear answers.

Without further ado… the book is called Eli the Adventurer is Very Very Brave. Click through the two image galleries below to see it in full:

None of the three of us had illustration backgrounds, or the budget to hire an illustrator. But a very brave toddler now has a book where he can see himself and his family, and hold something that helps him make sense of the hardest thing that’s happened to him.

It’s not the typical way I use AI in my work, but…if this isn’t a good use of image generation, I don’t know what is.

Thank you to Sarah and her family for letting me be part of this process and letting me share this story.

P.S. A note on names and privacy: This story was, of course, published with the permission of Sarah and her husband. Sarah was fine with me using her real first name, but her husband preferred I left his name out, which is why he is referred to as her husband throughout. Additionally, and for privacy reasons, we decided to change “Eli’s” name. Eli is not the real name of the toddler in question — the name was changed in this blog post and in the images of the book, for the privacy of our brave little friend.

If you’re a filmmaker, you’ve probably thought: funding shouldn’t be this hard to find.

The grants are real. The fellowships are out there. But the information is scattered across a hundred websites, newsletters, and bookmarks — and a lot of us miss deadlines not because we weren’t ready, but because we simply didn’t know in time. (Or we click something that sounds perfect… only to realize it doesn’t apply to our project.)

So I built a lil something to help with all of that.

The Film Funding Fairy asks you a few questions about your project, then returns relevant grants, fellowships, and labs — sortable by what’s open now vs. opening soon—with a quick note on why each one might be a fit.

Try the Fairy

It’s early and film-focused for now. If people find it useful, I plan to expand it to other funding verticals down the line.

How I built it (and why that matters)

The first version came together in an afternoon using Lovable — AI-assisted prototyping, or “vibe coding” if you’ve been seeing that term everywhere lately (if you haven’t, good primer here, and a more in-depth discussion here). Then I made a few small refinements over the next couple of days.

I’ve worked as a product lead building apps and websites before, with full teams of engineers and designers. This isn’t my first vibe-coding experiment, but it’s one of many that’s amazed me with how fast and powerful these tools are.

What keeps surprising me isn’t just the speed — it’s how quickly you can go from “this should exist” to something real and usable. Product experience helps, of course. But these tools are increasingly accessible — which means more of us can build the tools we wish existed.

How does that matter to you if you’re a filmmaker? Filmmakers have always been resourceful. Now we have tools that let us turn that resourcefulness into infrastructure and purpose-built apps and sites.

Here are a few things you might be able to create in an afternoon with Lovable similar tools (Replit, Claude Code, Cursor, etc):

• A clean landing page for your film (trailer, press kit, screenings, email sign-up)

• Prototyping a side hustle idea fast enough to know if it’s worth pursuing

• A submission tracker for grants and festivals that reminds you about deadlines and keeps everything in one place

• An impact hub: screening request form, partner outreach, automated follow-ups, donation link

• Custom tools for research-heavy or investigative projects (read about what NYT is doing in this vein for inspo)

…and much more.

If you’re already using vibe coding for your projects, I’d love to hear how! If you want to hear more details about how I built this, I’d love to hear that, too!

The Funding Fairy is VERY NEW, just shared it, and I’m sure there’s room for improvement. Please respond to this email with any feedback — what works, what’s missing, which funders or labs I should add.

And if you have a filmmaker friend who you think might like the Fairy, please share it with them!

xoxo,

Kris

Last Thursday, I taught “AI for Documentary Filmmakers” at Sundance Collab’s Story Forum. Over 400 people signed up for the one-hour workshop! (Thank you to everyone who joined, and hi to new friends who subscribed after attending.)

The chat was buzzing with great questions and recommendations (yes, it was recorded—I’ll share it when it goes live).

What struck me most: the anxiety has shifted. It’s giving less “Should I engage with AI? Is this evil?” and more “Okay…how do I use this ethically and intelligently without falling behind?”

So here’s how I keep up with AI tools + news in a way that’s actually sustainable—because no one has time to wade through boring AI slop, especially not boring AI slop about AI. 🙃

In this issue: 4 of my favorite go-to resources + 3 IRL communities worth knowing.

My #1 Favorite Place to Learn About AI: Hard Fork Podcast

• Why I love it: This NYT podcast is theoretically about tech broadly, but in practice it’s mostly about AI (they are the first to admit it). It’s funny and casual, but they also have sharp critical takes and interview people actually building the stuff. I listen pretty much every week. Kevin Roose also has great custom instructions for ChatGPT that I borrowed for my own settings (which I shared in the workshop!) and honestly…I find myself bringing up tidbits from this podcast a lot at dinner parties, etc.

• Best for: Keeping up with the overall AI industry in podcast form. Weekly download without overwhelm (~60 min of listening)

When I Need to Sound Smart: One Useful Thing by Ethan Mollick

• I know what you’re thinking: “Kris, you always sound smart!” TYSM. However, it’s only because I read this newsletter. Just kidding!! But like, not TOTALLY kidding.

• Why I like it: Ethan is an author and professor, and his writing is intellectual and thoughtful without veering into “overly boring academic” land. In a sea of AI-focused newsletters that are either light on prose or where the prose seems mostly written by ChatGPT... this one stands out as really thoughtful and well-written. This newsletter isn’t focused on film/creativity specifically, but he touches on these worlds a lot. And I think understanding the broader landscape is key to understanding our corner of it.

• Best for: Understanding the broader AI landscape, not just film/creativity. Slightly more academic/intellectual in tone.

When I’m Catching Up On the News: Today in Generative Media

• Why I like it: Comprehensive, straightforward news roundup every day. I don’t personally read it daily (it’s a lot!), but when I want to catch up on what’s new, browsing a few recent issues is a great way to go.

• Best for: Quick catch-up and updates on what’s happening across the genAI landscape.

When I Want Structured Learning: Curious Refuge Courses

• Why I like them: Curious Refuge has a number of great courses on AI & Film. I took their AI Filmmaking course over a year ago and learned a lot. Since then they’ve added courses on AI Documentary (which quotes this very newsletter!! Thanks Fred), AI VFX, AI Advertising, and more. They also have a solid podcast and really good, updated tool reviews.

• Can you find most of this stuff from free resources online? Probably! Are you going to? Probably not! So I’m personally glad I took the course. If you’re someone who benefits from structure and enjoys self-paced learning, they do a great job. You get access to a Slack community and online events too. And I met the founders at a dinner and can confirm they are real people and super nice.

• Best for: People who want a structured, comprehensive course. Self-paced with community access.

When I Want to Get Off My Screen: My Favorite IRL Events

Call me old school, but I still like to learn IRL, lol.

Artist and the Machine (LA & NY): I went to the LA conference in 2025 and really enjoyed it, despite having a migraine and being a lil out of it. Their next summit is in NY in May—if you’re NY-based and interested in this stuff, check it out! And maybe I’ll catch you at the next LA one! They also have a newsletter.

Sundance Collab’s Story Forum (where I just taught!): I’m biased, but really great lineup sessions on everything from legal toolkits for AI to Rian Johnson talking about collaboration. This happened both IRL (at the fest) and virtually. Sessions from last week will be up on their site in the next few weeks, and I’m going to check out some of the ones that I missed and it is #goals to attend IRL if they do it again next year.

Machine Cinema : Regular events (gengams) in NY, LA, and sometimes other cities, plus an active WhatsApp community where people share tools, ask questions, and support each other’s work.

If you’re in LA and know of any cool events in this space that I missed, please let me know! I’d love to attend (and/or speak at) more in 2026.

• Honorable Mentions

  • The above was hard to put together because there are so many great resources! A few more faves, in brief: Possible is another great podcast about AI (generally)..hosted by Reid Hoffman (cofounder of LinkedIn) and EP’d by my friend Katie Sanders. I like following Zach Seward (AI @ NYT; thoughtful takes on AI & media) and Nick Thompson (my former boss and current CEO of the Atlantic)—writes and speaks a lot on AI and its implications on media. Great podcast and Linkedin posts on “the most interesting thing in tech,” which often centers on AI and especially its relation to media. On IG, Souki Mansoor is a great follow. She is a friend and former doc filmmaker and amazing community builder now at OpenAI; she highlights awesome art and artists on her IG feed and the Sora feed. Elettra Fiumi also has a great AI & Film newsletter called AI Cinema. I’m constantly shocked and appalled by the all or mostly male panels in AI, and I particularly appreciated her Power List of Women in AI Creativity (and not just because I was included, I swear). Machine Cinema was mentioned in terms of events but they also have a newsletter and a vibrant (if at times overwhelming) whatsapp group.

Funny AI Fail of the [Time Period Since Last Issue*]

I decided that instead of a cute animal gif every week I’m going to do a “funny AI fail” every issue, which I’ll call “AI Fail of the [Time Period Since Last Issue]”. I think amidst the gloom and doom rhetoric around AI either 1) taking all of our jobs and possibly taking over the world and/or 2) ushering in a new and better era, it’s nice to step back and laugh at the many spectacular ways AI totally falls on its face. Here’s this week’s:

Source

One Ask

If this was useful, forward it to one filmmaker/storyteller friend who’s AI-curious or AI-skeptical. Learning this stuff in isolation is hard—learning in community is way better.

Also: What resources am I missing? Leave a comment or reply with your faves!

xoxo,

Kris

P.S. Some other stuff I’m reading lately:

• This post titled “help! my husband is addicted to Claude Code” made me laugh…HT to my friend Sarah for sharing

• For anyone interested in vibe coding, this is an awesome guide (geared towards product managers, but more broadly applicable)

P.P.S. Is anyone still reading? Well, if you are, I’m thinking of changing the name of this newsletter to reflect a slightly expanded scope. Current frontrunner: “A Computer, a Camera, and a Cat” (from an Agnes Varda quote: “This is all you need in life: a computer, a camera, and a cat.”). It speaks to making visual stories with technology (and yes, cats). Thoughts? Do you love it? Hate it? Just think I should keep brainstorming? LMK!

Hi friends,

Quick news: I’m leading a session at Sundance Institute’s Story Forum next week, and would love to see some of you there! Deets:

AI for Documentary Filmmakers: Research, Transcripts, and Trust Wednesday, January 29th | 6:30-7:30pm CST | Free & Online

I’ll be breaking down how I used AI tools on my PBS x Atlas Obscura nature documentary episode about Lençóis Maranhenses (the one with over 3.5 million views!). We’ll cover the practical stuff, plus the important guardrails. You’ll leave with takeaways for your own doc workflow. It will be recorded, but the experience will be better live if you can make it (plus you can ask questions).

Register for free

On a personal note: happy new year! I hope 2026 is treating you well. Last January was honestly quite rough for me—the LA fires, breastfeeding my newborn while not sleeping enough and working too much, tough news from someone close to me. I’m happy to report that this January has been much calmer, and I’m writing this from the beautiful San Miguel de Allende in Mexico. I hope it’s been a good month for you too.

If you can’t make my session, Story Forum has a ton of other great free sessions on Jan 29-30 from leaders in the field that are worth checking out, including a session on legal guardrails; one by Gary Hustwit on directing Eno, the wild and beautiful “AI-edited” and critically acclaimed doc I wrote about in an earlier newsletter; and much more. Full list here.

The whole thing is free and focused on using creative tech responsibly and ethically. Totally my jam, and I suspect some of yours too. I AM getting paid to lead my workshop, but am NOT getting paid to promote the others—they legit sound awesome and I'll be joining a few myself. Hope to see some of you there!

xoxo,

Kris

P.S. Next newsletter will be about how I actually learn about AI—my favorite newsletters, podcasts, and resources. Coming soon :)

Today, I’m going to share a little about how I’m using AI in my professional work, focused on a specific recent project that does NOT use GenAI in the final project…but used it along the way.

The Project

A PBS x Atlas Obscura short doc on Brazil’s stunning Lençóis Maranhenses (title: “It Looks Like a Desert. But It Has Thousands of Lakes”) that I had the IMMENSE pleasure of directing and producing! It has 1.9 million organic views (edit: 5 million) and counting, making it easily the most popular video in the majestic Untold Earth series (which I also had the pleasure of EP’ing earlier this season).

I am really proud of our episode! And at its heart, this is a very much a (human) hand-crafted piece. But AI tools helped me—a lot!—to save time and work more efficiently in the creation process and, I believe, help improve the final product. Here are the deets.

How I Used AI

Granola for Pre-Interview Note-Taking, Transcription, and Translation

Granola is one of my favorite everyday AI-powered apps. I use it most days, often multiple times a day. Usually, I use it as it’s intended, for meeting notes. Granola, unlike some competitors, doesn’t record; rather, it listens in and takes smart summary notes, so you don’t necessarily need to inform people you are using it and it doesn’t take up space in your zoom window as a participant like many tools, which I personally don’t like. But I used it a little differently this time around…and it was definitely an MVP for this project.

I initially heard about Granola about a year ago from my former coworker Alex Rainert, who wrote a great post about how Granola helps him be “more present and effective in meetings,” describing it as “magical.” Granola bills itself as “The AI notepad for people in back-to-back meetings” and promises to “take your raw meeting notes and make them awesome.” For Alex—and for me—it’s definitely delivered on that front.

But for this project, I also used it for pre-interview note-taking and found it especially helpful for international collaboration.

What made it so useful:

Summary notes + searchable transcription
Not only did I have great summaries to share with collaborators, but after each pre-interview I could ask things like: “I’m working on x project, and need quotes on y & z topics … can you pull out the 10 best relevant quotes from this meeting, verbatim?” or “Can you pull the quotes on X subtopic?”

Meeting & meeting note translation
Some of our pre-interviews were conducted in Portuguese (a language I don’t speak) by my co-director and longtime film collaborator, Leandro Badalotti. I asked Leandro to use Granola for these interviews, and Granola translated and summarized everything for us in English—incredibly useful and time-saving for a multilingual project.

But the coolest thing happened almost by accident. There was one pre-interview I planned to conduct solo because of scheduling challenges. We thought the interviewee spoke fluent English. It turned out that while he could understand my questions well, he wasn’t comfortable responding fluently in English.

Since I was using Granola, I suggested he respond in Portuguese, sensing he would respond more thoroughly. I realized I could copy and paste the live transcription of what he was saying in Portuguese into Google Translate—and I also knew I’d have a full translation and summary in English later. This allowed me to conduct a successful pre-interview with someone who spoke a different language from me. Whoa!

This happened last minute and kind of by mistake. There might be better tools for this (and if you know of one, let me know). But I was honestly surprised and delighted that this was possible—it felt magical.

ElevenLabs for Temp Narration Audio

We needed narrator audio for early drafts to help visualize pacing and flow. ElevenLabs made it quick and easy to generate a temporary voiceover that we could use while we worked on the edit—long before we had our final narration recorded. This helped us refine the script and timing without waiting for the real thing.

ChatGPT as a Research Assistant

Research for this project was particularly challenging because a lot of the source materials were in Portuguese. While I always find ChatGPT helpful as a research assistant, it was especially valuable here for:

• Finding and translating Portuguese-language sources

• Identifying potential interview subjects (whose websites and papers, etc, were often in Portuguese)

• Summarizing foreign-language articles and materials as well as long/dense/technical scientific articles that were in English, but otherwise would have taken us a long time to read and understand

• Generally opening up a whole world of information that would have been much harder and more time-consuming to access otherwise

Of course, I always double check research that’s sourced from AI in case it makes a mistake. In this case, we also had the privilege of working with a PBS fact-checker on the final piece.

ChatGPT & Claude for Editing Pitches and Other Materials

I also used ChatGPT and Claude to help edit, polish, and fact-check various written materials throughout the project—from initial pitch documents to interview questions to the budget, production notes, and beyond. I drafted everything myself and did a final read myself, too, but having an AI assistant that could quickly suggest clearer phrasing, catch typos, or generally help me improve my drafts saved lots of time and improved the quality of my work.

I could definitely write a whole post with more detail on this topic (let me know if you’d like that)!

In Conclusion

The way I used AI in this project might seem obvious to some of you, and revelatory to others. I find that a lot of people still think of using AI in extremes, but for me—while I sometimes enjoy making projects where the end product uses GenAI—I think some of the most useful, impactful uses are ones like the above: saving time and money and improving your output in the process of making something that is not “AI-generated.”

It’s not replacing human creativity or craft—it’s augmenting it, letting us spend more time on the parts of the process that truly require human judgment and artistry. These tools also democratized my team’s ability to work across languages and continents on a tight budget—work that would have required a larger team or more resources just a few years ago.

Note: While I loved using AI to help on this project, it’s worth saying that my BEST collaborators on this were definitely humans! Shout out to my co-director and editor Leandro Badalotti, local producer Beatriz Maglione, DP César Leite, and our incredible partners at PBS and Atlas Obscura including but not limited to Doug Baldinger, Hazel Lubbock, Elvis Metcalf, Mike Martin, Danielle Broza, and Gabbi Ewing.

Your AI-related LOL of the Day

(HT Taylor Lorenz)

Your Cute BB Animal Gif of the Issue

(If you’re new here, I always include one)

How are you using AI in your work? What do you want to hear about next? Would love to hear from you (for realz)!!

xoxo,
Kris

P.S. I drafted this post “by hand” and used Claude to help with light edits and polishing. Thanks, Claude!

After a pause due to parental leave and diving into some new professional projects, I'm now back. If you forgot what this is—hi! This is your friendly neighborhood newsletter on AI & Film, where we explore the intersection of filmmaking and artificial intelligence, highlighting the challenges, opportunities, and find inspiring and eye-opening work that’s being created with these wild new tools.

A quick word of thanks for all the support on my first two issues! I was thrilled to see my post on best and worst use of AI in Documentary republished in No Film School and quoted in the Curious Refuge Documentary Course. Thanks, too, to Alex Clark for quoting me in his excellent series on AI & Video for Video Consortium, and to Charles Haine for recommending me to teach an AI & Film college course—my first!

Most newsletters will focus on the present or future. But today's newsletter explores the past.

In many ways, this moment seems unique, unprecedented. But with another lens, we've kinda been here before. We'll dive into thousands of years of technological anxiety and what we can learn from it in our present moment.

You'll find a cute baby animal gif at the end. Because some traditions are worth keeping.

The OG Tech Skeptic: Plato

Here's something that blew my mind when I first heard it (back in grad school at NYU Tisch’s ITP): Plato was perhaps the first documented tech skeptic in media. He was worried that the newfangled technology of...wait for it…WRITING... would destroy human memory and wisdom. He was legit worried that writing things down would make people stop memorizing and truly understanding information.

His said (memorialized via Socrates) that writing would "create forgetfulness in the learners' souls, because they will not use their memories." He thought people would become "hearers of many things and will have learned nothing; they will appear to be omniscient and will generally know nothing."

Sound familiar? It's basically the "AI might make us lazy and stupid" argument... but from 370 BC. lol.

The Pattern Keeps Repeating: The Greatest Hits of Media Panic

1440s: Printing Press // Scholars worried that it would spread misinformation. (15th-century fake news panic)

1830s: Photography // Painters: "Our entire profession is dead!" Narrator: It wasn't.

1920s: Radio // Motion Picture Herald, 1932: "The most serious menace to the motion picture industry." Reality: Became Hollywood's best marketing tool.

1927: Sound in Film // Charlie Chaplin: "They're ruining the great beauty of silence." Silent film stars took panic voice lessons. Studios scrambled to retrofit theaters.

1950s: Television // Theater attendance dropped by half. Studios responded by going BIG - literally - with widescreen formats and spectacular effects.

1990s-2000s: Digital // Quentin Tarantino: "Digital projection is the death of cinema as I know it." Spoiler: Some of today's most stunning films are shot digitally.

2010s: Films on phones or tables // David Lynch gave us this absolutely legendary rant: "If you're playing the movie on a telephone, you will never in a trillion years experience the film. You'll think you have experienced it, but you'll be cheated. It's such a sadness that you think you've seen a film on your f***ing telephone. Get real." Years later, as the technology improved, Lynch completely shifted his position, acknowledging that with headphones in a quiet space, watching on a tablet could provide an experience proportionally similar to sitting at the back of a movie theater.

Our track record at predicting technological impacts is consistently, well, not great. Like… remember when everyone declared "3D is the future of all cinema" (2010), "VR will replace traditional cinema" (2016), or that "Digital will never match film" (2000s)?

Does AI Hit Differently?

Here's the uncomfortable truth, though: while previous tech waves changed HOW we make media, AI can potentially GENERATE it entirely. So there are similarities between what’s happened before, and we can certainly learn from it. But it’s also true that AI's impact might be more fundamental. (Though I’m sure that’s also something people thought about at various stages listed above. Right?)

Still, the historical pattern suggests AI won't "kill" filmmaking - but it might transform it in ways we can't fully predict. Some roles will likely disappear. New ones will emerge.

The real questions are: Who will benefit? Who will be left behind? And how do we—those who, by virtue of reading these newsletters and thinking about these things today, are probably early adopters—shape it thoughtfully and ethically?

Why We Keep Getting It Wrong

We consistently mis-predict technological impacts because we think linearly while tech advances exponentially’ we focus on technical capabilities rather than human behavior’ and we imagine new tech directly replacing old tech when it often creates new or hybrid forms.

Consider the color grading revolution of 2007-2011: When Apple released Color for $999 (versus equipment that previously cost $850k, only accessible to very high-end productions), the industry panicked. Everyone predicted disaster for traditional post-production houses. But the actual outcome? Blackmagic bought DaVinci and made Resolve free, democratizing the entire industry while creating entirely new business models no one anticipated. There were certainly some people who lost their jobs, but I would wager that—because such a wider range of productions started color correcting—there were ultimately more people who gained jobs as independent colorists.

Looking Forward

Maybe instead of asking "Will AI replace filmmakers?" we should be asking "How do we ensure this transformation benefits both creators and audiences?" History suggests that resistance is probably futile…but smart adaptation is essential.

My take is that the creatives who thrive during technological transitions aren't necessarily those with the most technical expertise, but those who focus on fundamentals, develop hybrid workflows, and maintain a vision-first approach that uses tools in service of creative goals—not the reverse.

The digital editing revolution took about 12 years to transform the industry (1989-2001), and the color grading revolution took roughly 4 years. AI tools are evolving even faster: I’ve truly been astonished by the progress that’s been made in the past year. The timeline is compressed, but the fundamental questions remain the same: How can we use these tools to tell better stories? How do we preserve the human elements that make storytelling powerful?

Next newsletter, I’ll talk about some of the great AI videos I saw last week at the Sora Selects LA event. (For the uninitiated, Sora is the video tool of OpenAI, the company that makes ChatGPT).

Until then, here's your cute gif of the issue:

Xoxo,

Kris

P.S. Agree? Disagree? I would truly love to hear from you! Please respond to this email with feedback!

P.P.S. Thank you to Danny Lunden for his help on this newsletter. No Film School for sharing the last issue of the newsletter. Charles Haine for his intro to NFS and Rutgers. And Fred Grinstein for sharing the last issue in the Curious Refuge Doc course!

Images in this post generated by me with gamma.ai - Ideogram 2. The gif is from Giphy.

Still frame from “The Wizard of AI” by Alan Warburton.

Today I’ll be sharing the best and worst uses of AI I’ve come across in the wide, wonderful world of documentaries! Obviously, since documentaries tell, you know, true stories, it can be a particularly fraught area to incorporate AI. 

Hat tip: This post is largely inspired by Zach Seward's SXSW talk. Zach currently spearheads AI initiatives at the New York Times. In his talk, he mentioned that he doesn’t have all the answers about how to use AI at the NYT, but that surveying the current state of AI-powered journalism—from the very bad to the very good—can provide valuable insights. And he’s right! So let’s apply the same approach to the doc world.

Subscribe now

I’ll start with the “worst” or most problematic uses of AI in docs I’ve seen; we can learn a lot from these. Then I’ll share some really COOL/ethical/effective examples of genAI in docs. Without further ado…

The Worst/Most Problematic Use Cases of AI in Documentary Film (in no particular order) 

 The doc: “Roadrunner: A Film About Anthony Bourdain”

• The offense: AI-generated voice of a real person (Bourdain) used without disclosure to the audience or official consent from the family. 

• The deets: In 2021, this documentary about the beloved celebrity chef, author, and travel documentarian Anthony Bourdain premiered with very positive reviews. But then, news broke that its director, Morgan Neville, had made an AI version of Bourdain’s voice. This “fake” voice was used in the documentary for 45 seconds, and its use was not disclosed to viewers. Bourdain’s real voice was also used elsewhere in the film. Viewers had no way of distinguishing the fake voice from the real one. 

• The response: Most people who heard about this were AGHAST! The New Yorker’s Helen Rosner summarized the response well in her piece about the controversy: “‘Well, this is ghoulish’; ‘This is awful’; ‘WTF?!’ people said on Twitter, where the fake Bourdain voice became a trending topic. The critic Sean Burns, who had reviewed the documentary negatively, tweeted, ‘I feel like this tells you all you need to know about the ethics of the people behind this project.’” There was, as Rosner notes, a particularly emotional response because 1) Neville had not gotten consent from Bourdain (who was deceased at the time) nor from his close-of-kin (though there is some nuance here: Neville claims there was a misunderstanding on this front… Bourdain’s ex-wife said that the topic did come up, and she does believe that the director  “thought he had everyone’s blessing to go ahead.”) But it’s clear that Neville should have asked more explicitly. 

• My take: Overall, this seems pretty obviously unethical / uncool. But see the “best” section for an example of a similar approach done well. 

The Doc: “What Jennifer Did” 

• The offense: AI-altered images of a real person (the primary subject, Jennifer Pan) used messily, and without disclosure.  

• The deets: What Jennifer Did is a Netflix true crime documentary that secured the coveted “top movie” spot on the platform for a while. The story focuses on Jennifer Pan, who was accused of conspiring to kill her parents and convicted of murder and attempted murder. When the police first started investigating Pan, a few candid pictures of her are shown. Two of those photos are strongly suspected of being generated by AI (I mean uhhh…look at her weird-AF hands in the picture below). Other pictures also have weird hands and teeth, and additional inconsistencies. Oof! 

Still from “What Jennifer Did” 

• The response: A lot of commentators basically said that using fake photos in a doc like this—especially without transparency or disclosure—compromised their trust in the documentary and/or in Netflix documentaries more generally. 

• My take: At a time when the doc industry is struggling, I think it’s more important than ever to establish and maintain trust with audiences…and it’s a shame, to put it mildly, when this kind of AI use erodes trust. There are obviously parallels between this example and the above one, but I personally think that because the Bourdain documentary at least uses the subject’s real words, this one feels even worse. 

The doc: [Redacted - Various] Film(s) using generated “b-roll/archival footage” without disclosure 

• The deets: I have heard multiple anecdotal stories about a director asking their editor or archival producer to fully generate (with AI) b-roll or “archival” footage for a documentary film...with no plans to disclose that the footage was generated. One such story—the specifics of which are under NDA—was the inspiration for the founding of the Archival Producers Alliance working group and best practices for AI, which I highly recommend checking out if you are interested in this topic.  

• My take: This seems pretty obviously bad! It’s lazy, and it’s not prioritizing the audience’s best interest. As discussed below, there can be a place for AI-generated b-roll, but it seems pretty obvious that it should be disclosed to audiences. It’s also possible that as AI-generated visuals become more common, a visual language around them will emerge to make them more obvious to viewers. (A parallel: reenactments in documentaries. These used to be more frequently disclosed to audiences, but now they usually aren’t—but they are generally filmed in such a way that it’s obvious to audiences that they ARE reenactments vs actual footage). 

As you can see, there are some common qualities in the above “worst” examples: namely, the use of AI seems intended to cut corners; its use is undisclosed, chipping away at the audience’s trust; and, at times, it’s also just lazy, and thus, distracting. 

But it’s not all bad…

The Best Uses of AI in Documentary Film

Note: There are a lot of great ways AI is being used as a tool to save time in the creation of docs (ex: auto transcription; editing with transcription in Premiere; hastening the color correction process). With the below examples, I’m not focusing on these, but rather, on generative AI that dramatically impacts the audience’s experience of the film.

The doc: “The Wizard of AI”

• The use of AI: This doc, directed by Alan Warburton, has been called “the world's first AI-generated documentary.” The visuals are apparently ~99% AI-generated  (sometimes there are non-AI generated parts incorporated, like Clippy, below). 

• The deets: It’s a visual essay in which a hoodie-wearing faceless ‘AI Collaborator’, voiced by the artist, is our incisive guide.

• My take: It’s really well-done overall. The heavy use of AI-generated visuals makes sense here because the film is about AI, and the way they are used also overtly suggests they are AI-generated to the viewer—Warburton isn’t trying to hide that he’s using AI tools here. Their use, of course, also feel appropriate to the subject, since the film is about AI, and it’s disclosed (in the credits as well as the Vimeo description). The director uses generative AI tools themselves to discuss and thoughtfully critique the aesthetic, legal, and ethical problems engendered by the wide new world of AI tools.  

Still from “The Wizard of AI” 

Eno

• The use of AI: This documentary about Brain Eno, directed by Gary Hustwit, uses AI to edit the film so it’s different for every audience who sees it. 

• The deets: Apparently, legendary musician Brian Eno was initially resistant to being featured in a doc, partly because he believes docu-bios of artists are usually “hagiographic.” That is, “until he [Eno] heard the pitch for this one.” 

• My take: Just the fact that Hustwit’s AI-assisted approach convinced Eno to do a project he wouldn’t otherwise do is fascinating! I would be surprised if this approach becomes commonplace, but is an interesting one—and seems to be resonating with audiences. It got into Sundance, has a 100% on Rotten Tomatoes, and anecdotally, my friends have raved about it—including friends who have seen it more than once to experience different cuts! It will be interesting to see if and how they release the film more broadly. 

“The Andy Warhol Diaries”: 

• The use of AI: Used AI-generated voice of Andy Warhol  

• The deets: Technically, this is very similar to the Bourdain example above. They used an AI software to have an AI-generated voice of Warhol read some of his real words, often from his diary. 

• My take: What makes this a GOOD use case is that 1) the filmmakers got permission from the Warhol estate, and 2) they were very transparent about the fact that they were using an AI-generated voice: it’s explicitly called out in the trailer, in the film, and in the credits. When it comes to AI in docs, transparency (and consent, when working with materials related to a real person—alive or dead [obviously when they’re dead, it’s consent of the family or estate]) is everything. I think we will be seeing this kind of use case in a LOT of documentary films in the near future. 

The Doc: “Another Body”

• The use of AI: AI deepfake of a subject who wanted to maintain anonymity.  

• The deets: This is a doc about a college student who finds deepfake pornography of herself online. To obscure her identity, generative AI tools empowered the film’s subject to “reclaim a technology that was used to attack her and silence her” (per the Hollywood Reporter). 

• My take: There are a lot of ways to mask subjects’ identities that DON’T use AI. However, I’ve personally often found them to be awkward and/or distracting. AI deepfakes might just do this better—offering subjects more peace of mind, and audiences a more realistic and compelling subject. This is an exciting area. Of course, disclosing the use of deepfake technology is key. 

Still from the making of “Another Body”. COURTESY OF ANOTHER BODY/SXSW

[Various] AI as a tool to make animated scenes for docs that can’t afford animation or recreation (ex: Semafor)

• My take: I don’t actually LOVE any real-world examples of this that I’ve seen, though props to the tech news startup Semafor for early explorations in this realm. But I think this is a really rich area for docs (and one which I am personally exploring for a project). Of course, for well-funded docs, directors and producers should continue to hire “real” animators or do “real” recreations. But there are many docs—especially in this market—that simply do not have the budget to fairly fund any animation or actors.For those docs,  AI can be a big unlock…especially now that AI-assisted animation has gotten so much better than when Semafor was first experimenting with it. 

This isn’t an exhaustive list—but I hope it’s enlightening and inspiring for you. If you have other examples of terrible or amazing uses of AI in docs, I’d love to hear about them—just respond to this email.

On that note, please respond with absolutely any feedback (positive or constructive) or notes on what you’d like to see me write about going forward. This is a nascent newsletter, and I really really appreciate all comments & support!

Subscribe now

Xoxo,

Kris

This week’s baby animal gif: 

P.S. Thank you to my friends and amazing writers & editors Morgan Childs for editing this newsletter, and to Lila Rosenberg for editing the first newsletter.

P.P.S. I made an Instagram page for Film Robots! If you want to see more frequent doses of AI & Film inspo, give it a follow.

So…a lot of filmmakers are hating on AI these days. The other week, I was at an event geared towards filmmakers where they screened a handful of short films. At the end, one of the hosts said, “You know what all of these films have in common? NONE OF THEM COULD HAVE BEEN MADE BY AN AI!” Cheers filled the room. Around the same time, SXSW audiences loudly booed videos touting AI. 

I get it. It’s been a (really) shitty year for the industry. There’s a lot of fear about AI taking even more jobs, leading to even more contraction. So why am I writing a newsletter about film and AI?

First: quietly, a lot of filmmakers are already using AI—sometimes without realizing it. I don’t think AI is something we can boo out of the room. It’s already here. That handy auto transcription Premiere rolled out last year? AI-powered. Trint? AI-powered. DaVinci face refinement/retouching? Magic Mask selection across multiple popular imaging/video applications (Photoshop, DaVinci, etc)? They are all AI. I’d guess a lot of the filmmakers that booed AI at SXSW had probably worked on a project that used AI, whether they knew it or not. The ones that haven’t probably will in the next year, especially given this week’s news that Premiere will be integrating more AI-powered plugins that allow users to extend shots, add/remove objects, and more. 

More importantly, I believe that, to quote Trei Brundrett, “Generative AI is an epochal development—less like social media and more like the advent of the internet itself.” This quote comes from a CJR article titled “AI is a big opportunity for the news media. Let’s not blow it,” much of which is very applicable to the film industry as well. Comparing the rise of the internet to the rise of generative AI isn’t a perfect parallel, but I think it’s a helpful one, in that trying to avoid using AI in our film work may—in 5 years—look a little like trying to avoid using the internet. We can try to resist—risking being left behind, and hindering our ability to shape its inevitable future in our industry. Or we can be early adopters, paving the way on how to use it most effectively, creatively, and ethically. To find the ways it can empower us. 

I don’t want to minimize the challenges and dangers AI poses to the film industry and to artists and filmmakers, and I’ll dive into some of these in future newsletters. But in a world in which most filmmakers and spaces for filmmakers are talking about these (very real) challenges and fears, I particularly want to highlight spots of hope: the ways we can be empowered with these new tools, saving time or making work we wouldn’t otherwise be able to make. The companies working on ethical generative AI that pay all the artists whose work was used in their models (yes, it’s a thing). As Brundrett argues, “This isn’t only about efficiency and automation that will be valuable [...] it’s also a moment of profound empowerment in the information age.”

And on a personal level, I’m really invigorated by this space. I’ve been surprised by how much I’ve enjoyed playing with various AI tools (including Midjourney, DALL-E, Kaiber, Runway, and Stable Diffusion/ComfyUI) for creating photos and videos. And I’ve been legitimately blown away by how quickly these tools are improving. 

Finally, I’ve been inspired by some of the work I’ve seen using AI in big ways, particularly the work incorporating AI-generated visuals. Right now, it seems few and far between, and hard to find—which is a big part of the inspiration of starting this newsletter: to showcase AI-powered work that will inspire you. That leads me to this week’s…

Paul Trillo is one of my very favorite filmmakers working with AI. His use of AI is creative and forward-thinking. And his films are beautiful, whimsical, and thought-provoking. This moving, animated GoFundMe commercial he made in late 2022 was one of the first video/film pieces I saw using AI that made me stop in my tracks and want to use AI more in my own filmmaking. Of the piece, he writes, "When I was younger, I always dreamed of creating an animation out of paintings but the scope and feasibility were beyond me. The hope here is to use AI not to replace the old way of doing things but to open up new creative and storytelling possibilities." Since then, Paul has also used generative AI to make trippy music videos, trippy cloud videos, and most recently, a trippy video with Sora (I guess trippy is his vibe).

Speaking of Sora (which, ICYMI, is Open AI’s new buzzy video tool—currently only available to select creators like Paul) last week, Paul was on my favorite tech podcast, NYT’s Hard Fork, talking about his experience using Sora and film & AI more generally. His interview starts around 24:30. Enjoy!

That’s all for today! Here’s your cute baby animal gif of the day (which is how I’ll end every newsletter):

See you in a week or two (I’ll aim to send this out every 1-2 weeks). 

xoxo,

Kris

P.S. Obviously, this is a new newsletter! Would love any feedback. Just hit reply to this email.