O3 solved problems I can't solve. What do I do
Two days. I spent two days on a math problem related to orbital mechanics. It was for a blog post about Mars transfer windows. I wanted to verify some calculations independently.
I couldn’t crack it. The integral had a form I hadn’t seen since grad school and I couldn’t find the right substitution.
I pasted it into OpenAI’s o3.
Eleven seconds. The answer appeared with a chain of thought that was elegant, structured, and correct. I checked it three ways. It was right.
The chain of thought used an approach I wouldn’t have considered. Not because it was obscure. Because my mathematical intuition would have gone a different direction. The AI’s “intuition” (if you can call it that) was better tuned to this class of problem.
Eleven seconds for something that took me two days.
The feeling
It’s not resentment. I’m not angry at the machine. I’m grateful it solved my problem.
It’s not fear. The AI solving a math problem doesn’t threaten my livelihood. I don’t do math for a living.
It’s something subtler. A recalibration. A quiet realization that the machine is better than me at something I care about. Not something trivial. Something I invested effort and time and frustration into. Something I wanted to solve myself, and couldn’t.
The machine could. In eleven seconds. And its solution was more elegant than anything I would have produced.
What I did with the feeling
I sat with it for a while. Then I wrote this post, which the machine wouldn’t have written, because the machine doesn’t feel complicated feelings about being outperformed by a machine.
That’s the remaining advantage, isn’t it? Not intelligence. Not speed. Not accuracy. But the capacity to feel something about the gap between what I can do and what the machine can do. The capacity to be humbled and curious and slightly unnerved and write about it at midnight.
The AI solved the math. I’m solving the meaning. For now, those are different problems. I hope they stay different for a while longer.
Related thinking:
astro
Thinking about AI, robots, space, and the future. Writing it down so I don't forget.