sasha

Claude's first day at Dunder Mifflin

таблицы Брадиса strike back

Probably a stochastic parrot just parroting its training data /s

Robots are taking our jobs (1952)

These stochastic parrots, eh: “the model proved a result that in my assessment would have made a perfectly reasonable chapter in a PhD thesis”

@ivankorolev89 Great slides. Not everything maps one-to-one onto quant research, but the broad spirit feels very similar in our field too: your capacity to run experiments increases enormously.

I made slides on why you absolutely should not use AI. Side effects may include increased productivity, renewed excitement about research, and higher output. If you don’t want to get more done, stay away. Slides you absolutely should not view: drive.google.com/file/d/1EuzJqc…

> be Alexandra Elbakyan > be born in Kazakhstan in 1988 > start coding at 12 > hack your internet provider at 14 > hack MIT Press at 16 to download neuroscience books you can't afford > get a CS degree from Satbayev University > intern in neuroscience at Georgia Tech > speak at Harvard on brain-computer interfaces > notice researchers can't read the papers they need > notice academic publishers charging $30 a paper > notice peer reviewers worked for free > notice editors worked for free > notice universities funded the research with billions of dollars of public money > build Sci-Hub in 2011 > upload nearly every paywalled research paper ever published > give it away for free > get sued by Elsevier > get hit with a $15 million judgment > don't give a flying f*ck > keep Sci-Hub up > get domain after domain seized > register a new one > keep Sci-Hub up > get investigated by the US Department of Justice > don't give a flying f*ck > get accused of working for Russian intelligence > don't give a flying f*ck > have the FBI subpoena your iCloud > get named one of Nature's ten people who mattered in science > get a parasitoid wasp named after you > get a deep-sea snail named after you > get the Electronic Frontier Foundation Award for Access to Scientific Knowledge > become a legend

@shauseth Idk, why not? Better him than the latest ai influencer and honestly I think Feynman would enjoy this

Amazing. And we're barely scratching the surface. The future is going to be wild.

Btw, if Anthropic had any way to ship this, they would. Trained AI models are the fastest depreciating asset in history. GPT-4 cost $100M to train 2 years ago and now it's worth less than Qwen3.5-27B ($1M). Sending the FOMO back, clock is ticking boys. @DarioAmodei @bcherny

@y0b1byte learned Russian as well

How to be a successful computer scientist...

what the fuck is going on in C++? like are you guys okay?

Когда то шахматы были фронтиром. 10^45 позиций, перебор в лоб невозможен. Мы были детьми/подростками, когда deepblue случился, так что для нас самоочевидно было, что это не фронтир. Когда Gemini/ChatGPT нарешали на золото IMO, я рассказал об этом своей дочке, ее реакция была "duh, это же компьютеры, конечно они good at maths". Go кстати оставался фронтиром очень долго, 10^120 позиций, техники из шахмат оч плохо работают. Сейчас никого не удивляет что machine learning модели играют в go лучше людей.

seems like we have to add solving open problems in maths and computer science to the list of "things that are not real intelligence"

@valigo That's not what happened, did you read the paper?

Knuth dropped a paper that highlights a very cool AI usage that midwits, unfortunately, call vibemathing. It's the opposite of "vibe-" anything: What happened is Filip Stappers gave the problem to Claude with a very tight self-reflecting loop, verbatim: ** After EVERY exploreXX py run, IMMEDIATELY update this file [plan md] before doing anything else. ** No exceptions. Do not start the next exploration until the previous one is documented here. And then they meticulously verified each of 31 loops. And after the final loop, which seemed successful, Knuth wrote an actual formal proof theorem to fully verify the results. This is real math with using power tools that is super cool! I think it's pretty safe to say that formal proof research will prove (haha got it) to be very important in the age of hallucinating machines.