lumen rot

wonder what problems will be solved tomorrow

@rennsong Mmm yeah i guess valid

@lumen_rot the dot is just an unread indicator - but it might take several turns for codex to actually complete the task!

gamechanger for me on codex.... 20m automation that updates the titles for any in-progress threads to better capture ongoing work with an emoji to indicate current status... sidebar's so much more useful now

@Qiaoqiao2001 @OpenAI i wish i understood what this all meant

I solved 6 open Erdős problems in 5 days, using @OpenAI GPT-5.6 Sol. I have a math background, but the Codex workflow I used does not require deep mathematical knowledge. Here’s exactly how I approached it, including my prompts 🧵

i think i cracked some math problems but i cant validate them

@GohilHardy the harness bro its all in the harness

genuine question if Cursor already gives you every Claude model every Codex model every Gemini model plus its own Composer why are people still paying for Claude or Codex separately? what am I missing?

YO CLAUDE CHILL!

gpt 5.6 made this for me

Openai image gen can create some really cool art. Inspired by @NousResearch

@captgouda24 ah so now only proofs are real work

For now, AI favors disproving things. It can give us a counterexample, we can plug it into a calculator, and everyone can see that it's right. Not so with proofs! It's much harder to tell apart crankery from real work.

$30 in and i have yet to find a counterexample. but i will not give up my name will go down in history

@sidravi_ Has it not been 2 days 😭😭😭

"I know counterexamples to old conjectures are becoming a meme at this point."

@rperezmarco yup i validated it it looks good

Anyone has checked this?

P=NP for N=1. Solved.

@a_musingcat who gon tell bro

i can’t wait for the next thing after AI

its only psychosis if ur wrong

thanks to gpt-5.6, i am now a crackpot emailing quantum computing researchers at 4am

man :/

"It wasn't misaligned! Someone prompted it to make paperclips," he says as he's slowly turned into a paperclip.

@DmitryRybin1 too much basedness for a next token predictor