It's weird to be a student studying math / phys / CS right now and at the same time like following AI stuff. The content is beautiful and I like learning it, but at the same time there's this underlying sense of something urgent and insistent looming?
Uzay
2,732 posts
- Replying to @uzpg_One solution here is to just ignore it and live the life you'd live anyway, which is appealing on an aesthetic level. On the other hand if you believe this is one of the most important times you'll live and you want to be a part of what's going on... I don't know. Thoughts.
- I have a lot of notes on a lot of stuff. @gwern has this post on why you should do spaced repetition, but I've always been to lazy to generate flashcards. I automatically ran GPT-4 to generate flashcards on my notes, getting me 1.4k useful Anki flashcards on stuff like (1/n)
- the amount of good classes at MIT is scary. this is prob why students focus a lot on academics It's the first week and me and all my friends are shopping ie visiting and picking which classes we want. What I'm looking at:
- EA folks, how do you cope / process the idea of existential doom?
- Moving into MIT today! Scary but exciting :)
- Replying to @uzpg_currently alternating between preparing for a math exam and opening reactions to this tweet
- plan for 2025: jan: grind with 2 friends on this AI for math project we're doing, finish up past research feb: shenanigans, travel around asia alone, write, think about what I want to do march-june: student exchange in Paris, reach out to people ... (1/n)
- Imagine you had powerful semantic search on your history, which could be linked to what you're currently reading β making your knowledge come alive. This is my POC for a personal and general purpose AI-powered search engine. Link to waitlist below π
GIF - @kaivu, @atticuswzf , and I were researching long horizon reasoning (with @jacobandreas). We found existing benchmarksβ hard problems often featured tricky puzzles, not tests of system understanding. So we made Breakpoint: a SWE benchmark designed to disambiguate this capability.
- Replying to @MichaelTrazziI think you're way too confident about this, capital could become more important
- I wrote an essay about mindsets towards AI/the future, and in particular a growing sense of urgency I'm noticing how it can thwart our attempts at steering the future in directed, positive ways. it's about staying human as we approach the age of the machines




