Moving Fast & Slow with AI
Psychology splits how we make decisions into two systems: System 1 and System 2.
At work and with AI, I see most things fall into this framework, and into a clash between the two.
System 1 is the brain on autopilot: quick, easy, top down. System 2 works harder and thinks through something difficult, bottom up.

System 1 handles easy tasks, and System 2 handles hard ones. Friction usually pulls us into System 2. I think most people's AI use breaks that balance.
Hard things should be hard and easy things should be easy. That's why we have two systems. But because AI draws on so much public knowledge, most things, maybe eventually all of them, will be easy to reach and done for us. The work of making something happen has been replaced by a shortcut for everything, and it's changing how we live. Students who used to write papers now hand the AI a prompt and let it write the whole thing. Engineers write code while walking in the park. A lot of everyday critical thinking no longer takes any effort.
I make a lot of art and work with a lot of engineers, and many of us take pride in the hard work it takes to bring something to life. The process is what builds our attachment to the result. As you make something, it changes with new ideas and influences along the way. When I make something with AI, it's usually one or two prompts to get an MVP, compared with the hours it takes to make something usable by hand. That drop in effort feels like it's snowballing, and incentives from the top only push it further.
We've already seen easy access to content hurt kids. Think of Baby Shark on loop. Why wouldn't the same apply to adults? Our brains are still plastic and can be reshaped, for better or worse. The brain works like a muscle: it gets stronger through hard work, breakdown, rest and rebuilding. Studies already compare people who write papers with AI or search engines against people who write them alone, and brain connectivity is much higher in the people writing on their own.
It may be conjecture, but I think AI is making us dumber, because we no longer need to retain information. Why learn and remember something if I can outsource recall to a second, AI brain? At work that can be useful, but I can see this easy access working against us as we lose the ability to train our own brains. The average person is more educated than the generation before, and we're slowly making ourselves dumber than that generation.
That's probably why I've been adding friction back into my life, so I don't wear down my own critical thinking. You lose what you don't use, so right now I'm retraining that muscle instead of handing it to a system I don't understand. It's part of why I try not to reach for AI first, whether I'm making art or writing something like this. It's very tempting to let AI write everything or run my main tasks, especially when I'm rewarded for it. But I'm focused on getting better wherever I can, and if that means a sharper brain, it's worth it.
I don't know the full solution beyond making myself think and do the work. Sometimes I just want an answer, like for a math problem. But when a task needs real thinking, it's better to push through than hand it off. Many colleagues add friction too, sketching ideas or writing a full draft before bringing AI in to edit. I still think there's a balance to find.
If you take one thing from this, notice what you reach for when you have a question or a task. Being deliberate about how we use AI is what will protect our brains.
Inspirations and references:
www.scientificamerican.com/article/kahneman-excerpt-thinking-fast-and-slow
en.wikipedia.org/wiki/Thinking,_Fast_and_Slow (Wikipedia article on the book)
