I'm a 22-Year-Old Intern and This Documentary Made Me Rethink What Ambitious Means

---

Original Video

Title: The Thinking Game | Full documentary | Tribeca Film Festival official selection Uploader: @googledeepmind Duration: 1:24:07 Published: 2025-11-25 Views: 301,913,693 | Likes: 189,405

---

Okay, I need to be upfront: I cried during this documentary. Not at the part you'd expect — not when they solved protein folding, not when Lee Sedol resigned, not during any of the big triumphant moments. I cried during the chess story. The one where 12-year-old Demis Hassabis resigned a drawn game because he was too tired to see the stalemate, and the ex-Danish champion laughed in his face.

I'm a marketing intern at FindTube. I'm 22. I am definitively not building AGI. But something about a kid sitting in a church hall in Liechtenstein, exhausted after 10 hours of chess, making a mistake that would redirect his entire life — that hit me in a way I wasn't prepared for.

The Resignation That Changed Everything

Here's what happened after that chess loss. Hassabis didn't just feel bad about blundering a draw. He had this massive existential moment: "Are we wasting our minds? Is this the best use of all this brain power?" He looked around the room at 300 chess players and thought — what if all these brains were working on something that actually mattered?

He was twelve.

When I was twelve, my biggest existential crisis was whether to post a photo with the Valencia or Clarendon filter. I'm not saying this to be self-deprecating — I'm saying it because the gap between "Instagram filter anxiety" and "maybe I should redirect all of human intelligence toward solving cancer" is the gap between a normal person and Demis Hassabis. And watching this documentary made me weirdly okay with that gap.

Because here's what I noticed: Hassabis didn't go straight from that chess tournament to founding DeepMind. He went to Bullfrog and made a theme park game where people throw up on each other. He coded vomiting AI. That was his next step after his grand epiphany about the future of intelligence.

The path from "I want to solve AI" to actually solving protein folding took 25 years, and most of those years were spent doing things that looked nothing like solving AI. Writing game logic. Studying neuroscience. Pitching skeptical VCs. Getting rejected. Playing more chess.

The Talent Stack Nobody Talks About

The thing that struck me most about Hassabis — and this is going to sound like a marketing observation because that's literally what I'm trained to notice — is his talent stack.

Chess prodigy. Game designer. Neuroscientist. Entrepreneur. Each of those is impressive on its own. But the combination is what created DeepMind. He understood competition (chess), user engagement and simulation (games), biological intelligence (neuroscience), and how to build organizations around impossible goals (entrepreneurship).

None of those skills are "AI research." But all of them were necessary.

I think about this in my own tiny context. I'm a marketing major who dances salsa, grows tomatoes on her balcony, and keeps a swipe file of landing pages she admires. None of that sounds like it belongs on a resume for an AI company. But at FindTube, my salsa background helps me understand rhythm and timing in content release schedules. My gardening teaches me patience — you plant content seeds and wait for organic traffic to grow. My swipe file is basically pattern recognition training for copywriting.

Hassabis turned chess pattern recognition into AI architecture. I'm turning Instagram analytics into growth marketing. The scale is absurdly different. The principle is the same: nothing you learn is wasted if you're paying attention to the meta-skills underneath.

"We Were the Best in the World at a Problem the World's Not Good At"

This line from Hassabis about their first CASP result is the most brutally honest thing I've ever heard a leader say about their own company's work. They won the protein folding competition — and it didn't matter, because the predictions weren't accurate enough for biologists to actually use.

"It doesn't help if you have the tallest ladder when you're going to the moon."

I added this to my marketing swipe file. Not because I'm going to use it in a blog post title (although... maybe I will), but because it perfectly captures something I deal with every week at a much smaller scale.

I write blog posts for FindTube. Some of them rank really well on Google. "5 Ways to Learn Coding from YouTube Videos Efficiently" — great SEO, solid traffic, looks impressive on my weekly dashboard. But when I track conversion to actual sign-ups? Sometimes the best-ranking posts convert the worst. I'm winning the search game while losing the user acquisition game.

Best ladder. Wrong building.

Aisha, my VP, keeps pushing me to think beyond vanity metrics. "Page views are nice, Sofia, but show me the sign-ups." She's right. The DeepMind team had to learn the same lesson at a much higher level: winning CASP means nothing if biologists can't use your predictions. The metric that matters is always the one closest to the actual human being you're trying to help.

The Strike Team Model

When AlphaFold 1 fell short, Hassabis didn't fire anyone or pivot the company. He created a "strike team" — a focused group with a specific mission, new domain expertise (actual biologists!), and a rewritten data pipeline.

I've never managed a team in my life, but I've watched Aisha do something similar with our growth experiments. When a campaign isn't working, she doesn't just try harder at the same thing. She recomposes the team, brings in someone with a different perspective (sometimes that's me, the intern with zero context but fresh eyes), and reframes the problem.

The documentary shows DeepMind team members admitting they thought protein folding was "a fool's errand." People on the team were demoralized. And Hassabis held the line — not by pretending everything was fine, but by narrowing the goal to something the team could believe in: "It's about proving we can solve the whole problem."

That's leadership I want to learn from. Not the "everything is awesome" kind. The "this is hard, we might fail, but here's why it's still worth trying" kind.

"Why Don't We Just Do That?"

My absolute favorite moment in the entire documentary: after they solve protein folding, someone mentions they could fold every known protein sequence in about a month. And Hassabis says, "Why don't we just do that? And then release it."

Not "let's monetize it." Not "let's build a SaaS platform." Just: do it and give it away.

200 million protein structures. Free. For everyone.

I study marketing. I understand monetization strategies, paywall psychology, freemium funnels. And everything I've learned tells me that what Hassabis did was either insane or genius — and the 301 million views on this documentary suggest it was genius.

Because here's the marketing truth that most companies miss: generosity is the most powerful brand strategy that exists. When you give away something genuinely valuable — not a crippled free tier, not a 14-day trial, but the actual thing — people remember. Researchers around the world now associate "protein structure" with "DeepMind." That's brand equity you can't buy with any ad budget.

At FindTube, I've been pushing for more free, genuinely useful content rather than gated lead magnets. This documentary confirmed my instinct. Plant real seeds, not plastic flowers.

The Part About Responsibility

The documentary doesn't shy away from the scary stuff. Autonomous weapons. Deepfakes. Surveillance. Job displacement. Hassabis quotes the "alien civilization arriving on Earth" analogy and seems genuinely worried.

I'm 22, and honestly, I go back and forth on how scared I should be. Some days I read about AI-generated disinformation and feel like the world is ending. Other days I watch a researcher use AlphaFold to study a malaria protein and feel like everything is going to be okay.

What I appreciated about the documentary is that it doesn't resolve this tension. Hassabis says "AGI is coming whether we do it here or not" and "we better create institutions to protect us" — but he doesn't pretend to have the answers. He's building the thing and worrying about the thing at the same time. That feels honest.

As someone who works at an AI company — even as just an intern — I think about this more than people might expect. When I write a blog post promoting FindTube's video search, I'm promoting an AI product. Is it a dangerous AI product? No, it helps people find cooking tutorials faster. But it's part of a larger ecosystem that is moving very fast and doesn't always know where it's going.

I don't have a grand conclusion about AI safety. I'm an intern. But I think the fact that I'm thinking about it at all is maybe the point. Hassabis says "we need a lot more people really taking this seriously." Okay. I'm taking it seriously. That's my contribution for now.

What I'm Taking Back to My Desk

Monday morning, I'll be back at my desk writing blog posts and checking Mixpanel dashboards. Nobody at FindTube is going to confuse my work with solving protein folding. But here's what this documentary planted in my head:

1. The long game is real. Hassabis spent 25 years getting to AlphaFold. I've been at FindTube for 4 months. I need to stop evaluating my career on a weekly sprint cycle.

2. Domain expertise matters. AlphaFold didn't work until they brought in biologists. My blog posts won't convert until I deeply understand the developers I'm writing for. Shadow more. Listen more. Write less fluff.

3. Generosity compounds. Give away real value. The returns come later, in forms you can't predict.

4. Failure is data. CASP 13 wasn't a failure — it was the training run for CASP 14. My underperforming blog posts aren't failures — they're experiments that tell me what my audience actually cares about.

5. Ambition isn't about scale. Hassabis wanted to solve intelligence. I want to help people find the right moment in a video. Both of those are worth doing well.

I finished the documentary at 1 AM, texted my mom "watch this," and then lay in bed thinking about what I want my version of "the thinking game" to be. I don't know yet. But I know it involves paying attention, being patient, and not resigning drawn positions.

— Sofia Garcia, Marketing Intern @ FindTube.ai