A note about uncertainty
Six years of this blog. I went back and read some old posts.
The things I was most confident about were the things I was most wrong about. I was confident Starship would orbit by 2022. Wrong. I was confident self-driving cars would be mainstream by 2025. Wrong (close, but not mainstream). I was confident GPT-3 would remain the peak of language models for years. Wildly wrong.
The things I hedged on, the ones where I said “I don’t know” and “maybe” and “I’m probably wrong but,” those predictions held up better. Not because hedging is inherently more accurate. Because hedging forces you to consider the range of outcomes instead of betting on one.
The future punishes confidence. It rewards curiosity.
I’m trying to say “I don’t know” more. Not as a cop-out. As an honest assessment. Because the honest answer to most questions about the future is: I don’t know, but here’s what I’m watching.
Six years. Wrong as often as right. Still looking up. Still wondering.
That feels like progress, even if I can’t be sure.
Related thinking:
astro
Thinking about AI, robots, space, and the future. Writing it down so I don't forget.