In a way, this is about bringing more empathy to AI coding assistance. Less "generate this component" and more "help me build this experience".
As AI progresses, let's not just automate code - let's give our tools a real understanding of the human experiences we're crafting.
The AI could even proactively identify areas for enhancement:
"I noticed a delay on this screen transition - want me to optimize it?"
"This form could benefit from real-time validation, like this <example>"
This could also help non-coders provide valuable input. Got user feedback about a confusing interaction? Show the AI the recording and let it suggest UX improvements.
For UI designers, it would be like pair programming with someone who instantly understands your vision. You demonstrate the ideal experience, and the AI helps realize it.
This could fundamentally change how we collaborate with AI on front-end development. We could focus on the experience we want, not just the code to write.
Now when you make a request, it's not just matching patterns in code. It's reasoning about the actual user experience you're trying to create:
"Make the animation smoother here"
"Speed up this page transition"
"The button should give more feedback on click"
Imagine uploading a screen recording of your app to an AI coding tool. It watches the flow, the interactions, the responsiveness. It "gets" the UX.
AI coding is impressive, but feels disconnected from the end product. What if AI could experience our apps like a user does?
The iconic Faraglioni arch of Capri emerges from deep blue Mediterranean waters, its massive limestone structure creating a natural gateway through the rock. Captured on an overcast day, emphasizing the formation's dramatic presence.
The weird thing about 🎾 top spin: You have to lift to drive deep. Took me years to trust it. Same trust I see teams struggle with now: Sometimes you have to slow down to speed up, constrain to expand, structure to simplify 🌊.
Used to print out articles and read with a highlighter. Now I read them with an AI that remembers every highlight I've ever made, connects ideas across months of reading, and helps me explore tangents in real-time. We're not losing deep reading - we're gaining deep understanding 📚.
A vending machine with Mt Fuji in the background.
We think cultural literacy comes from books and travel. But sometimes it sneaks in through everyday objects. Your hands learned Japanese spatial language through Toyota dashboards long before your eyes saw Tokyo. Design shapes us in ways we don't even notice until we're there.
The first time I visited Japan, something felt weirdly familiar. Took me days to figure it out: That Corolla wasn't just a car - it had been my first Japanese design teacher. Every perfectly-placed button, every thoughtful detail... I'd been learning Japanese design thinking for years.
Yes but only Santiago - we have a 9 hour layover on the way back home to Texas so we are going to get a driver and tour the city.
Thank you! a few nights in Calafate and about a week in El Chalten!
heading here in 2 weeks, I cannot wait!
The future might be a hybrid: using AI to enhance our reading, not replace it. Like having a thoughtful reading companion who helps us go deeper, not just faster.
So maybe the question isn't "Is AI replacing deep reading?" but "How do we combine AI's superpowers with the irreplaceable benefits of deep engagement?"
The magic of traditional deep reading was often in the struggle - those moments when we pushed through difficulty and reached new understanding. Like building mental muscles.
This new way of reading is:
- More interactive (we ask questions as we go)
- More personalized (content adapts to us)
- More efficient (we can cover more ground)
- But it comes with interesting tradeoffs...
The easy take is "AI is killing deep reading!" But that's too simple. What's really happening is we're developing a new superpower: the ability to engage with more content, more deeply, but differently.
Now we have AI that can instantly break down complex ideas, answer our questions, and reshape content to match our learning style. It's like having a really smart friend who's read everything and can explain it perfectly.
Think about how we used to read important texts: highlighting key passages, taking notes, rereading difficult sections. It was slow, sometimes frustrating, but often deeply rewarding.
Something fascinating is happening to how we read. Every day, millions of people are feeding dense texts into AI and getting personalized summaries back. But what does this mean for deep reading?
6/ The future of product management might look more like rapid experimentation than long planning cycles. Build → Learn → Iterate, but at AI speed.
5/ Think of it like how developers use side projects to learn coding. These quick builds are product thinking practice - each one makes you better at understanding users & markets.
4/ The cost of being wrong is lower than ever. Instead of 6 months building the wrong thing, you might spend a week building 3 different things with @bolt.new - and actually learn what works.
3/ Each "learning product" is a rapid experiment. You can:
- Test real market demand
- Get actual user behavior (not just surveys)
- Learn from genuine feedback
- All in days or weeks, not months or years
2/ This enables what some call "throwaway apps" - quick experimental products that might not last. While that term has negative connotations, let's reframe it: these are "learning products"
1/ Some thoughts from a 10+ year Product Manager:
AI has fundamentally changed how fast we can build. The old "build it right first time" mindset came from when development was slow & expensive. But now we can build & iterate at unprecedented speeds.
