Adam Wathan's surprising take on AI


Overall, I've been pretty unimpressed with many of the promises coming from the "AI code bois." I got tired of watching video demos that hyped things up, only to try it myself and be disappointed.

But after this conversation with Adam Wathan (creator of Tailwind CSS) and Brian Casel (Instrumental.dev), I'm starting to see AI development differently:

video preview

I was surprised that Adam is this bullish on AI dev tools:

"I used to be more skeptical. But then I forced myself to download Cursor and build an entire project without typing any code, making the computer type the code for me. I had to throw myself in the deep end."

Adam and Brian break down their AI workflow, and we also discuss:

  • Why "vibe coding" is mostly bullshit (but still useful for some projects).
  • 90% of programming is grunt work, and AI is great at it.
  • Why experienced developers have a huge advantage in this "AI era."
  • Two different modes of using AI: learning vs. building
"AI is like keyboard shortcuts on steroids for the things I am an expert in."

Adam also got vulnerable about how AI is negatively affecting the Tailwind UI business. "I'm working on my next act," he said. (He also shares what he's working on next.)

They both recommended I try the new Claude Code and Cursor. (I haven't tried either.)

Really curious what you think about Adam and Brian's take. Reply and let me know here (or leave a comment on the YouTube video).

Cheers,
Justin Jackson

PS: This video is just one of the segments from today's episode of The Panel podcast. Listen to the whole thing!

We're recording episodes live now. Join us in chat next week, June 12.

Justin's newsletter

I'm the co-founder of Transistor.fm (podcast hosting and analytics). I write about SaaS marketing, bootstrapping startups, pursuing a good life, building calm companies, business ethics, and creating a better society.

Read more from Justin's newsletter
Reasonably unreasonable

"Never tell me the odds." – Han Solo, Star Wars. What are the odds? Yesterday, I watched my favorite hockey team (the Edmonton Oilers) lose to the Florida Panthers in the Stanley Cup finals. Every season, all 32 NHL teams have the same goal: "Win the Cup." But only one team can win. Even a strong team (like the Oilers) has only a 12% chance of winning. Put another way: they had an 88% chance of not winning. What are your odds? A founder building a new startup also faces daunting odds. As...

Sketchy door

When I got to the address, the door looked like this: sketchy However, once inside, I was surprised to see an incredible podcast recording studio: Justin and Nathan at Kit studios On April 1st, Nathan Barry (founder of Kit) asked me to join him in an industrial area of Chicago to record a podcast. This was for the opening of the new Kit Studios in Chicago. Even cooler: anyone with a Kit account can use it. (I'm told the outside is much nicer now. 😉) Nathan and I sat down, and he interviewed...

Hey,

I want to hear from you. This week, I'm sharing 3+1 bite-sized thoughts, ideas, and prompts. Hit reply here, or on Bluesky, with your thoughts! ------- 📺 🗡️ Will YouTube kill the podcast industry? I've been thinking and writing a lot about YouTube, podcasting, and open protocols (RSS) vs big centralized platforms. Read my post | Reply on Bluesky (If you're of a "certain generation" and you're thinking, "who uses YouTube for podcasts?" you might want to read this draft I'm working on) -------...