Dave Burstein
@AInews_wire
Followers
282
Following
5
Media
11
Statuses
7K
Rapid coverage of AI. Many tweets. @AIPrimes has just the most interesting.
NY
Joined May 2023
RT @TensorWaveCloud: Look what just arrived š. @AMD's newest Instinct Series GPU, the MI355X, has officially been delivered to our latest dā¦.
0
77
0
Self-Evolving models are exciting people in the way reasoning did not long ago. @rohanpaul_ai.
So many brilliant AI "Self-Evolving" papers. Here's another one. R-Zero shows a model can train itself from zero external data, by pairing a Challenger and a Solver that keep learning right at 50% uncertainty, no human tasks or labels. On Qwen3-4B it adds +6.49 on math and
0
0
0
Generate videos in just a few seconds. Try Grok Imagine, free for a limited time.
505
657
5K
The distributed Search hasn't found any Extraterrestrial Intelligence Intelligence hasn't found any yet but the tech works. Nous Research is now distributing LLM training.
nousresearch.com
Psyche is an open infrastructure that democratizes AI development by decentralizing training across underutilized hardware. Building on DisTrO and its predecessor DeMo, Psyche reduces data transfer...
0
0
0
RT @rohanpaul_ai: LLM judges act too sure of themselves, this paper shows how to measure it and reduce it. The problem is calibration, theā¦.
0
4
0
The secret reason Facebook, the Russians, etc. haven't blown up elections yet: the ads don't work very well, especially in the U.S. Most voters have already made up their minds and know which side they are on .
nber.org
0
0
0
RT @nrehiew_: Let's talk about the GLM 4.5 models. The latest frontier open weights model out of China (and possibly the best at the momenā¦.
0
59
0
RT @GXiming: š How far can RL scaling take LLMs?.Drop ProRLv2! š„We keep expanding LLMās reasoning boundaries through 3,000+ RL steps over 5ā¦.
huggingface.co
0
9
0
RT @_AndrewZhao: Nice empirical paper investigating all your bag of tricks in reasoning LLMs. .
0
57
0
RT @dorsa_rohani: New fastest shortest-path algorithm in 41 years!.Tsinghua researchers broke Dijkstraās 1984 āsorting barrier,ā achievingā¦.
0
3K
0
RT @sama: If you have been following the GPT-5 rollout, one thing you might be noticing is how much of an attachment some people have to spā¦.
0
1K
0
RT @dnystedt: Texas Instruments notified clients of 30% price hikes on over 60,000 chips, or 40% of all products, for industrial control, aā¦.
0
43
0
RT @DeryaTR_: Today I am unveiling the second remarkable breakthrough with the OpenAI GPT-5 thinking model. We had carried out an extraorā¦.
0
61
0
RT @caydengineer: we made Meta Rayban glasses but with an open source sdk. then 100 hackers make apps on them at YC. this is the dev ecosysā¦.
0
48
0
RT @plzaccelerate: Sam Altman: From AI Nerd to the AlexNet Breakthroughā£.ā£."I was an AI nerd my whole life. I never thought I was gonna beā¦.
0
3
0
RT @KeithSakata: Iām a psychiatrist. In 2025, Iāve seen 12 people hospitalized after losing touch with reality because of AI. Online, Iāmā¦.
0
11K
0
RT @slow_developer: Demis Hassabis describes a surreal setup: one AI playing inside the mind of another. DeepMind SIMA agent can now operatā¦.
0
173
0
RT @psyth91: Relightable Codec Avatars is now extended to full-body! At #SIGGRAPH2025, we will present Relightable Full-body Gaussian Codecā¦.
0
30
0
RT @Dari_Designs: The style for beautiful minimalistic illustrations.with grain texture and simple patterns:. [prompt], detailed outline ilā¦.
0
55
0
RT @elevenlabsio: ElevenLabs is proud to announce the @ChromaAwards - The āOlympicsā of AI Film, Music, and Games. Creative AI tools are rā¦.
0
58
0
RT @CoffeeVectors: Testing Wan 2.2 on Itās able to do cool stuff with camera that I canāt seem to do with the otherā¦.
0
37
0