aiEDU Studios: Emmy-Winning Producer Gavin Purcell
The former showrunner for The Tonight Show Starring Jimmy Fallon and co-host of "AI for Humans" weighs in on the ways AI is changing the entertainment industry
When we began putting together the initial list of folks to interview for aiEDU Studios, I knew we had to talk to Gavin Purcell.
Gavin is a veteran of what insiders refer to as “linear TV,” most notably for the 8+ years he spent working for Jimmy Fallon (first as producer of Late Night and then producer of The Tonight Show). He’s since made the unlikely pivot from the epicenter of the traditional entertainment industry to the cutting edge of AI, synthetic media, and the creator economy. He’s also a good friend and mentor, whom I often turned to for insights about what’s happening on the bleeding edge of AI.
Two years ago, Gavin decided to bring those insights to the world and launched AI For Humans, a YouTube channel that has quickly become one of the most popular sources of day-to-day updates about the latest developments in AI.
Our episode with Gavin hits precisely what we envisioned for aiEDU Studios — not an interview with canned questions and rehearsed responses, but a free flowing conversation between kindred spirits (nerds) with a lot to say about the wild, wonderful, and scary world of AI.
I’m still experimenting with how to use Substack in connection with aiEDU Studios. On the one hand, we want folks to actually watch the interview and don’t want to simply regurgitate the conversation into written form. But I also don’t want to treat Substack purely as a promotional vehicle for the show. So, here’s my latest attempt to walk that tightrope, with some of the most interesting broad strokes of our convo and links to the time stamps in case you want to skip directly to that section of the video.
The Moment Everything Changed
Gavin began his rapid descent into the AI rabbit hole when he encountered GPT-2 and GPT-3 papers, around 2020-2021.
“It was the first time it could write paragraphs which were cohesive and coherent,” he recalls. “I was like whoa, that’s weird. I’ve made my living writing stuff and now the machine can do this thing.”
Unlike the recent tech fads that Silicon Valley had been churning out to mixed results; AI felt fundamentally different. Having lived through the early internet and personal computers, Gavin recognized something unprecedented. "You know, I got a PC when I was, like eight or 10... lived through early internet, lived through web two. This really does feel as significant as getting that first computer.”
What keeps him engaged isn't just the hype—it's the deepening complexity. "My problem as a person is I tend to jump from thing to thing to thing that interests me and like I've been able to hold on to this AI thing now for a couple of years because it continues to get deeper, more mature and, I think, honestly more transformative than almost anything else I've lived through."
From The Tonight Show to the Wild West of YouTube
After years as showrunner for The Tonight Show Starring Jimmy Fallon, managing a team of 200+ people, Purcell made a counterintuitive move: He started over. Along with co-host Kevin Pereira, he launched "AI for Humans," a YouTube channel that now has 20,000 subscribers and dives deep into the latest AI developments each week.
The transition wasn't just about changing mediums—it was about changing the entire production model. "I'm looking at right now, I have a $200 camera that gives me a really clean shot. That is a straightforward, really good thing, whereas that would have cost like $1,500 before," Purcell notes. "You can buy a microphone that is like $150 that would have been like $600 before."
But the real revelation was the creative freedom: "I really wanted to see what would it look like at my age, you know, middle-aged guy to try to create something entirely from scratch with a partner," he says. “The tools that once required massive teams are now accessible to anyone willing to learn them.”
The $20,000 Question
Our episode was filmed just as news broke about OpenAI’s rumored $20,000-per-month AI agents. While that price tag might shock some, Gavin sees the economics differently: "If that AI agent is a PhD-level AI researcher... $20,000 may sound like a lot to a person, but to an AI engineer, that is not very much money for a year-long salary."
The math becomes even more compelling when you factor in productivity. "Not only do you get that AI engineer if it's as good as an AI engineer that you're going to have on staff but also conceivably, based on compute time, if you can get it, it could work 24 hours a day, seven days a week," Purcell explains.
The real disruption isn't the cost—it's the capability. These agents could work 24/7, and as Purcell puts it: "What kind of is this a human that's going to replace? And like what the value of that is."
Hollywood's Reckoning
One of my goals for the episode was hearing Gavin’s take on the implications for AI on the entertainment industry, which has already been hit by waves of disruption and change, from the rise of Netflix, and then YouTube, and then short-form scrolling video.
The entertainment industry's response to AI has been, in Gavin’s words, "confused." He's heard Jeffrey Katzenberg, DreamWorks co-founder, claim that "80% of the people on an animated movie won't be needed anymore," a quote that generated significant backlash.
Yet the industry's own structure makes it vulnerable. "When you think about a Marvel movie, there is the director, there are the writers, there are the actors, but then there are like 500 graphic artists who are working within Autodesk Maya and all these other software tools," Gavin observes. "If you look at the credits of a Marvel movie it’s never ending. You see the big credits: production, designer, writer, director, the actors, and then you have this list of literally hundreds of people and a lot of those folks are doing the effects work.”
Gavin sees a lot of changes in store. He points to creators like Neural Viz, a single editor who's built an entire universe (or “Monoverse,” as Neural Viz calls it) of AI-generated characters with over 100,000 YouTube subscribers. "He's created aliens, he's created characters that make you laugh. All of these can talk, they can tell a story... that's where the creative side of this comes in." The barrier to entry has gone from mountains to molehills overnight, and we’re just beginning to see the fruits of a burgeoning creator economy that is moving faster than incumbent industry players to experiment and adopt AI tools.
The 'Microstudio' Revolution
This all begs the question of what the future of Hollywood looks like. Gavin believes it will all come down to the democratization of creative tools and what he calls “microstudios”—small teams of 1-5 creative people who can produce content that previously required hundreds of employees.
"You suddenly have these tools where... distribution was easy but production was kind of hard, still right? What I think we're gonna get at now is a place where you can make the thing much easier than you could before.”
The comparison to early YouTube is telling. "If you go to early YouTube, one out of every 100 videos was legitimately interesting. Today, a lot more of it is watchable,” Gavin said. AI generated content will probably follow the same path: “the vast majority of it will be kind of crappy, but there will be good stuff that comes out of that.”
He compares it to the streaming explosion that gave us everything from Netflix originals to niche YouTube creators with millions of subscribers. The difference now is that production barriers are collapsing alongside distribution barriers.
The Parent Trap
As a father, Gavin grapples with AI's implications for the next generation. Both his daughters use ChatGPT regularly for homework help—not to get answers, but to understand concepts better. His younger daughter has gone from struggling in math to straight A's in advanced math.
"She'll take a picture of her homework, and I've watched her do this, she's not trying to get the answers from it, but she uses it to explain to her like how to better do this and like this is somebody who wasn't like a math expert but is now straight A's in advanced math," he explains.
But like me, Gavin worries about the broader implications. "AI is going to probably disrupt a lot—I think most people are underestimating the disruption of work," he says. His advice is both practical and philosophical: "Whatever they're curious about and interested in, that feels like it's going to be something that really kind of resonates with them internally. That's all I care about at this point."
We talked about what advice to share with young people, and we both landed on a common refrain: stay curious, keep learning, and recognize that whether you're enthusiastic or skeptical, this technology demands attention.
Watch the full conversation on aiEDU Studios' YouTube channel, or wherever you enjoy your podcasts.