Don't get too attached to your code

One hidden benefit of using AI to write code is that it can actually reduce your emotional bias.

When I write something myself, my brain gets emotionally invested in it. I start to develop confirmation bias: every little positive signal becomes “proof” that my work is good, that my course is well written, that my idea is solid. I want to believe it’s great because I put in the effort.

A similar bias shows up when you’re building new products. Historically, the cost of building a product was so high—time, money, energy—that founders became emotionally attached. They felt compelled to see it through, even when the signs weren’t great, simply because they had invested so much.

With AI, that dynamic changes.

The cost of running an experiment is now incredibly low, and you’re not the one doing most of the work. Because you’re less emotionally invested in the output, it’s easier to be objective. If an AI-generated project or experiment fails, you can just move on and try another one. If it works, great—you can put it to use.

In other words, AI doesn’t just speed up building; it also makes it easier to let go.