I like reading this, because I love focusing and getting stuff done and vibing with AI in the past has left me feeling exhausted and dirty, like I don't deserve the thing I build. (Btw I'm also an expert at procrastinating! I really need to get into the flow and that is hard but rewarding.)
But recently I have been building a Django app that is all about the data model and with complicated RBAC, and I've been discussing a lot, and I mean a lot, like pages and pages of text back and forth, with Claude code (yes code because I do ask it for the occasional edit, most is done by me).
Claude walks me through things, shows me ups and downs of my plans, points at my flaws in thinking, points at where the huge preliminary text in a docstring does not match what I implemented. And I enjoy it! I'm learning, the picture is getting clearer and I feel good about what I'm building. I feel confident that it is correct and good!
In other words, what will differentiate people is not how smart they are but their relationship to mental effort.
Yeah, so smarter people? A smart person will quickly whip up a solution with AI, and use AI optimally, as I have observed online. It's already smart people with math degrees who are using AI to create proofs. Ordinary, non-math people are not doing this. I think it's more like AI magnifies the talents of already talented people.
So did typesetting. But now nobody wants me to ink pages with my bins of moveable type, so I have to go be smart somewhere else, in another field, where everyone is leagues ahead of me in all the ways that count.
Signed,
a fictitious 18th century typesetter, perhaps the best and smartest there ever was
I like reading this, because I love focusing and getting stuff done and vibing with AI in the past has left me feeling exhausted and dirty, like I don't deserve the thing I build. (Btw I'm also an expert at procrastinating! I really need to get into the flow and that is hard but rewarding.)
But recently I have been building a Django app that is all about the data model and with complicated RBAC, and I've been discussing a lot, and I mean a lot, like pages and pages of text back and forth, with Claude code (yes code because I do ask it for the occasional edit, most is done by me).
Claude walks me through things, shows me ups and downs of my plans, points at my flaws in thinking, points at where the huge preliminary text in a docstring does not match what I implemented. And I enjoy it! I'm learning, the picture is getting clearer and I feel good about what I'm building. I feel confident that it is correct and good!
Related are these two comments I made earlier: https://news.ycombinator.com/item?id=48743028
This is my favorite article on the different ways to use AI.
https://danielmiessler.com/blog/keep-the-robots-out-of-the-g...
https://archive.is/4slNQ
In other words, what will differentiate people is not how smart they are but their relationship to mental effort.
Yeah, so smarter people? A smart person will quickly whip up a solution with AI, and use AI optimally, as I have observed online. It's already smart people with math degrees who are using AI to create proofs. Ordinary, non-math people are not doing this. I think it's more like AI magnifies the talents of already talented people.
So did typesetting. But now nobody wants me to ink pages with my bins of moveable type, so I have to go be smart somewhere else, in another field, where everyone is leagues ahead of me in all the ways that count.
Signed,
a fictitious 18th century typesetter, perhaps the best and smartest there ever was