Let’s talk about something that’s been on my mind lately. I came across an interview with a founder — let’s call him Lobster — who runs a company in the middle of the AI boom. Most founders these days talk about efficiency, automation, and moats. Lobster talked about something else: fun.
Not “user engagement.” Not “gamification.” Just fun. Like, the kind of fun you had when you were nine years old, building something for no reason other than the sheer joy of watching it work.
I took seven notes from that conversation. They might change how you think about your relationship with AI.
Note #1: Serious people build serious tools. Curious people build the future.
One of the first things Lobster said was: “The people who are most afraid of being replaced by AI are the ones who treat their work like a checklist.” That hit hard. When you treat your craft as a set of repeatable tasks, you become a task-doer. And task-doers are exactly what AI replaces best.
The people who will thrive, he argued, are the ones who approach AI the way a child approaches a box of LEGOs — not knowing what they’re going to build, but excited to find out.
Note #2: The best prompt engineering is also the worst prompt engineering.
There’s a whole industry now teaching people “prompt engineering.” But Lobster had a contrarian take. He said the most interesting outputs he’s gotten from AI came from terrible prompts. “I once typed: ‘Write a breakup letter from a toaster.’ The result was surprisingly profound.”
The point isn’t that you shouldn’t refine your prompts. It’s that the most interesting discoveries happen when you give yourself permission to be weird. The algorithm rewards clarity. But humans reward surprise.
Note #3: Precision kills creativity. Permission gives birth to it.
In a corporate environment, every piece of content needs a use case, a KPI, and a sign-off. Lobster thinks that’s exactly the wrong mindset for the AI era. “If you’re only using AI to make your existing work 10% faster, you’re using it wrong. The real gain is the 10x possibility that comes from asking stupid questions.”
This is hard for most of us to accept. We’re trained to optimize. But optimization is a hill-climbing algorithm — it gets you to the local maximum, not the global one. The global maximum is found by wandering.
Note #4: The best AI users have a “low shame threshold.”
Here’s a pattern Lobster noticed: the people who get the most value from AI are the ones who are comfortable looking foolish. They’ll ask the same question five different ways. They’ll say “I don’t understand, explain it like I’m a goldfish.” They’ll generate 50 variations of a logo just because it’s fun to see what happens.
In a culture that worships competence, this is a superpower. Competence is the enemy of exploration. If you’re always trying to look smart, you’ll never try the dumb thing that turns out to be brilliant.
Note #5: AI creates a new form of “flow.”
Flow state is when you’re so absorbed in what you’re doing that time disappears. We usually think of flow as something that happens when you’re an expert — a musician improvising, a programmer deep in code. Lobster argues that AI can give non-experts a taste of that flow.
“I can’t code,” he said. “But with AI, I can prototype an idea in 20 minutes. That feeling of ‘I made something’ — it’s addictive. And it’s not just about productivity. It’s about the joy of creation.”
The implication is profound: AI doesn’t just make us faster. It makes us more playful. And play is where original thinking comes from.
Note #6: The biggest risk isn’t that AI will replace you. It’s that you’ll stop doing things that are worth doing.
Lobster told a story about a friend who used AI to automate his entire content creation pipeline. He was producing 50 articles a week, all high-quality, all optimized for SEO. And then he stopped. Because what was the point? The machine was doing the work, but he wasn’t invested in any of it.
The lesson: if you outsource the fun parts of your work, you lose the reason you started working in the first place. AI should be your collaborator, not your replacement. And the best collaboration happens when you retain the part that feels like play.
Note #7: In the long run, the only sustainable competitive advantage is taste.
Here’s the final note, and it’s my favorite. Lobster said: “Everyone will have access to the same AI. The only difference will be who decides what to make with it. That decision is driven by taste.”
Taste isn’t something you can scale or automate. It’s built from thousands of hours of paying attention to what delights you, what surprises you, what makes you want to share it with someone else. Taste is the residue of play.
So if you’re worried about AI taking your job, don’t try to be more efficient. Try to be more interesting. Ask better questions. Dare to be silly. Build something that makes you grin.
Because in the age of AI, the most serious thing you can do is have fun.