Tech News

XL

Top comment: @xlr8harder “I fully believe that if Anthropic thought they could develop their machine god in private without users at all, that this would be their preferred path. You are an inconvenience they tolerate to get sufficient access to capital.”

KA

Top comment: @yacineMTB “@tydsh Please point this at simulators and optimizing them for RL training loops”

MC

Top comment: @plasmatic99 “@MTSlive They do seem to keep arsing things up and having to backtrack”

AL

Top comment: @ObadiaAlex “Apply: https://schmidtsciences.smapply.io/prog/scaling_ai_safety_for_a_multi_agent_world/ Deepmind's blogpost: https://deepmind.google/blog/investing-in-multi-agent-ai-safety-research/”

TC

Top comment: @TomChivers “(I should say, it's very much not my project alone; @DadaJudith, @bakkermichiel, @AlexTPet, Daan Juijn, Stan van Baarsen, and others did the thinking, I just made the words sound nice and read them out)”

SE

Top comment: @SemiAnalysis_ “Obviously this is way worse than API overall. However, explicitly nerfing subscriptions leads to huge public backlash, and the rapidly falling cost of intelligence means you'll be able to profitably serve Opus 4.8 level models for $20/month in the near future. We therefore think it's far more likely the labs will withhold new features/models from subscription plans. It will be interesting to see if Mythos ends up being API only. (4/4)”

DM

Top comment: @davidmanheim “@GaryMarcus If usage continues to increase rapidly, then tokens may end up being an expensive and scarce commodity controlled by relatively few players - regardless of AI progress.”

GG

Top comment: @googlegemma “⚡ Blazing Fast: By shifting the decode bottleneck from memory-bandwidth to compute, DiffusionGemma delivers up to a 4x speedup on standard accelerators. (1000+ tokens per second on a single NVIDIA H100, 700+ tokens per second on NVIDIA GeForce RTX 5090!)”

RS

Top comment: @Scobleizer “@chamath Yeah Anthropic sure pissed off a lot. Warning signals of new troubles to come?”

V(

Top comment: @reach_vb “@gui_penedo @HKydlicek Congratulations and excited for you guys!!”

YO

Top comment: @YosarianTwo “@NathanpmYoung Yeah; at minimum, there's no reason to think an ASI would be *less* noticeable on the galactic scale than biological intelligence.”

TL

Top comment: @TomLoweCinema “@DavidSacks Dario constantly warmongers and stirs up conflict with "Authoritarian governments" but says absolutely nothing about tens of thousands of EU and UK citizens being arrested for social-media posts. Laughable hypocrisy.”

QB

Top comment: @qberthet “Read about how we used it to finetune DiffusionGemma on many tasks, including a very cool showcase on Sudoku puzzles! https://developers.googleblog.com/en/diffusiongemma-the-developer-guide/?linkId=62264697”

NL

Top comment: @natolambert “@BlackHC honestly I don't really understand the question. The labs have stronger internal models, yes I know this. I think it's good that there's competition that'll drive a need for broader access.”

FB

Top comment: @xeophon “@julien_c too broad of a statement to agree with”

SN

Top comment: @snowclipsed “@code_star yup all those times when people would say "claude feels nerfed today" 🫠 and it's almost always claude.”

LA

Top comment: @scaling01 “the meme of China doing nothing and winning is actually so true”

HA

Top comment: @hallerite “@DimitrisPapail @PrimeIntellect it just makes sense if you take an information theoretic view on RL. we need to squeeze out as much signal as we can from each rollout :)”

JG

Top comment: @JunGao33210520 “2/5 📊 Data. Data is the key to success🔑 We built a large-scale, standardized data pipeline that curates, filters, and dedups broad public robotics and egocentric human datasets. This gives us a clean training dataset with diverse tasks, scenarios, actions, and embodiments.”

TH

Top comment: @voooooogel “@medjedowo true,,,”