Surprise! There is no building today, just reflecting.

Since Monday, with the help of AI, I have conceptualized and written a 6,000-word “treatment” of a pilot episode for a faux-sitcom about my life. I have learned some basic best practices in strong sitcom-style writing, and I have fleshed out full character cards with a 12-episode narrative arc for season one. I created mockups for marketing branding for that site (both stylistic design choices as cover art, and a musical theme song). And then I put all of this online in a clean website that I built in minutes.

This is an astonishing level of output. And as a reminder: I did this all in 3.5 hours.

Creating a faux-sitcom about my life is an example of the kind of project that a person idly daydreams about, the thing your friends tell you that you should do over cushy lunch catch-ups or over a pint at the bar. Before AI, the barrier to entry to actually just put pen to paper and see what happens would have been far too high. Today, it’s anybody’s game. 

Is this a perfect, ready-for-submission piece to the Netflix reviewing team? Absolutely not. But it didn’t have to be. To me, this was less about me wanting to pursue a closet career as a movie-maker, and more about me wanting to demonstrate creative mastery over a suite of AI tools. To show what’s possible with a little determination, a lot of creativity, and a willingness to think outside the box.

When we talk about how teachers can assign work that leans into the AI age, this is the kind of thing I’m talking about. 

Yes, I conceptualized a comedic take on my very real life. But if imagine the starter prompt wasn’t to write a sitcom about your own life, but to instead transform a great work of literature, history, or science? To come up with a 6-episode docuseries that accurately explains concepts like climate change, or the timeline of events chronicling the Industrial Revolution?

As I’ve been learning, subject matter mastery is often a hidden benefit that comes after the motivation to build. The more time you spend immersing yourself in an area of work in a highly kinesthetic state (ie: actually going through the motions to write a script, and consider some of the marketing elements), the more you learn about what’s actually involved to do the work.

Of course, this is the crux of project-based learning. Get deep in a project in an intense way over a very short period of time, then step back and reflect on what you learned — both the skills you picked up, and the industry knowledge you acquired.

If the skills are versatile enough, you can transfer them to another industry without a lot of fuss. And this is where it all comes back to AI fluency. Being able to manipulate AI for a wide variety of creative and practical outcomes is quickly becoming the currency of demand for the future of work.

If you learn how to vibe code a marketing splash page for your summer internship at a tech startup, you can almost certainly apply that skill in your next job. If you learn how to look at the work you do as a series of steps that each add up to longer workflows — and you can quickly spot places where AI can expedite a part of that step — you will become invaluable on any operations team. If you look at a stack of data or a tome of old blog posts and can imagine it repurposed as infographics or a TikTok series, you will bring creative energy and a fresh look to the work you do.

That’s why I believe the best thing schools and employers can do today is get people building. Widen the aperture of achievement from singular, siloed specialists into a lens that reveals much more cross-functional and intersecting possibilities.

When I considered the direction I could take my faux-sitcom project next, a lot of possibilities emerged. Did I have more content to create? Maybe I could transform part of the script into an AI-generated movie, a teaser reel, or even reimagined as a comic book or zine. Or maybe it was time to lean into the business case. I’m sure I could just as easily generate a budget for filming, flesh out casting ideas for characters, and even a list of producers and agencies to pitch. Or would I take the exaggeration game even further, setting up a custom ChatGPT critic to “review” the pilot episode, or even set up a mock “Rotten Tomatoes” style review site of my show? I can go on and on…

The point is this: When you can build with AI, the possibilities are endless. And that’s both an incredible unlock and also a major problem. Without constraints, it’s hard to know when to stop. When you can do anything, how do you know where to draw the line?

I’ve found that time and money are the strongest constraint-based motivators for me. Without enough time, you can’t get anything done. With too much time, you languish in possibility. (I imagine the same is true for too much money, but I’ve not personally experienced that side of the problem.)

That’s why I time-boxed myself to 60 minutes a day on this project. That time constraint gave me a bias to build. And once that fixed window ended and the buzzer sounded, I can look back on what I’ve started and ask the more thoughtful question: Is this thing I created worth further investment of my time, or should we leave this one on the shelf for now?

Evaluation and quality assessment are also really important questions to learn. Over time, we learn how to fine-tune our human radar for high-quality content. And now, we must retune our radar for what feels like “good” content in the AI age. This is important for anyone who works with a computer today, but it’s particularly important for the AI-native generation, who will be the ones we need to listen to about what makes “good” vs. “bad” AI content.

That’s why it’s more important than ever that we all get at-bats and building with AI, to tune that radar, build that muscle, and establish not just surface-level awareness, but linguistic-level fluency. 

As anyone who has studied a foreign languages has likely learned, you can take all the French classes in the world, but it’s really damn hard to call yourself fluent until you’ve immersed yourself in a french-speaking culture and spoken only that language, day in and day out for weeks or months on end. Over time, even if you might not have the words for to explain it, you tune your ear to “listen” for the thing that just sounds a little bit off in someone’s accent, or the idiom that doesn’t quite land as it should. These subtle nuances are tough to explain unless you’ve lived in it. 

Working with AI is the same way. It takes fluency to establish taste. I’m building out my own fluency through daily practice, “AI Power Hour” build sessions, and by committing to experimenting with a new way of using AI every single week. What about you?