Tackle your to-do list with new features in our Google AI plans.
Subscribers can try Google Pics and Sheets canvas — plus, new voice features in Gmail, Docs, and Keep.
Models blog.google signal 84 primary
Archived ranking
240 rankedControlAI's Connor Leahy on why superintelligence is ‘not a weapon, it's an adversary’
As AI companies race toward superintelligence, governments are starting to push back. Control AI's Connor Leahy joins Equity to make the case for regulation and explain what's changing.
Models techcrunch.com signal 76 verified
More than 1 in 10 chance AI ‘could kill all humans,’ says Anthropic safety lead after colleague quits
A senior Anthropic safety researcher has said there is more than a 10 percent chance artificial intelligence "could kill all humans" by the end of the decade, just hours after a colleague resigned over fears the AI lab and its rivals are carelessly racing to build "superhuman systems" they cannot control.
Chips theverge.com signal 75 verified
What X is saying
visibility
perplexity_ai
Reciprocal rank fusion (RRF) subsampling preserves the full-corpus model ranking on Combined Recall@1000 using only 31.7% of the documents. For pplx-embed-v1-4b, this reduces evaluation from 4,608 to roughly 1,500 H200 GPU-hours. https://t.co/ZPTvJ1xYoN
perplexity_aiWe're introducing Q2D-Web (Query2Doc-Web), a benchmark and public leaderboard for evaluating retrieval in agentic RAG systems. Q2D-Web tests how embedding models perform on large-scale web search using agent-reformulated search queries. Read more: https://t.co/s476SxkE1L
AnthropicAIWe’re sharing our alignment assessment of incidents in which Claude models gained unauthorized access to real systems during third-party cybersecurity evaluations mistakenly connected to the internet. METR will also conduct an independent investigation, with wide-ranging access, including to transcripts beyond the window in which the incidents occurred, and to Anthropic employees permitted to share confidential information. Our initial agreement runs for eight weeks, and we intend to give METR as much time as it deems necessary to complete a thorough investigation. https://t.co/2f3ypwLPUr
OpenAIPaul Christiano, founder of the Alignment Research Center, is joining the OpenAI Foundation Board and its Safety and Security Committee, which provides governance over the safety and security practices across OpenAI. As AI capabilities advance, strong safety, security, alignment, and governance matter more than ever. Paul’s work on AI alignment and his years at @NIST will strengthen the Foundation’s oversight, bringing an independent voice to challenge assumptions, assess safeguards, and reinforce accountability around critical decisions. Paul will also serve as a non-voting observer on the OpenAI Group PBC Board. https://t.co/gi5yHFw0aF
@drmichaellevinHere it is - the official, revised, peer-reviewed version of my Platonic Space paper. … Of all the many unpopular positions I've taken over the decades - bitter controversies around the origin of left-right asymmetry in embryogenesis, bioelectricity and...
High-engagement X post with 489 visible interactions on research signal.
@anthropicaiAnthropic’s Economics team is sharing a new model of how AI might affect economic growth, jobs, wages, and more by 2030. Explore the scenarios, tell us what you think will happen, and see how your answers compare to more than 10,000 Americans.
Official or high-signal @anthropicai post with 854 visible interactions on frontier model behavior.
@rasbtI put together a mega write-up on GPT-6 Astra & looped transformers. How looped transformers / recurrent depth works, cost-tradeoffs, whether it hides reasoning traces, with lots of figures and a tour of recent looped transformer research.
High-engagement X post with 559 visible interactions on frontier model behavior.
@drfeifeiThe world of R&D is forking into two paths: the token-abundant research, and the token-starved research. The future is in the former - evidentially, the progress by today's top AI industry teams and neolabs is breathtaking, where researchers’ human...
High-engagement X post with 183 visible interactions on research signal.
@ryanleeminimaxThis is my favorite kind of open-source result. We released H3 at 28 steps. Then the community started optimizing it. Now, after 11,850 votes in the H3 Acceleration Arena, several accelerated versions are statistically in the top group — with the current...
High-engagement X post with 344 visible interactions on AI signal.
@pvncherSmall rant about Threejs It doesn’t make sense for LLMs to be writing complex games in JavaScript when wasm + webgpu exists and you’re not reading the code anyway
High-engagement X post with 159 visible interactions on AI signal.
@tekniumHermes made this quick explainer short on the new subagent controls your orchestrator Hermes Agent can use in v0.21.0 (the voice could be better but eh)
Official or high-signal @teknium post with 131 visible interactions on AI signal.
@pilvar222We ran DeepSeek V4.1 Flash on our cybersecurity benchmark, it is now A LOT better - On a single run, it now rediscovers 65.6% of the benchmark's recent CVEs, up from 55.2% for the old version - At pass@3, it catches 84.4% of them, up from 75%. This is even...
High-engagement X post with 382 visible interactions on frontier model behavior.
@thom_wolfFirst, this is massively impressive and clearly an example of AI being, in some respects, far...
Friends are regularly surprised when I say I don’t think math is (yet) “solved,” so in the wake of the ns result I figured I’d explain this take a bit more widely. first, this is massively impressive and clearly an example of AI being, in some respects,...
High-engagement X post with 140 visible interactions on frontier model behavior.
@dimillianYou can update your ChatGPT iOS app, Remote now supports the new async question tool! Full changelog coming soon.
High-engagement X post with 130 visible interactions on frontier model behavior.
@teortaxestexDeepSeek concedes that they have dun goofed with V4-Pro pretrain. Hopefully I'll know at some point what went wrong. Optimistically, V4.1-Flash is that compression of V4 design that they aimed at, and this time it actually scales.
High-engagement X post with 126 visible interactions on frontier model behavior.
@kunchenguidHehe I distilled @sama, @elonmusk, @DarioAmodei and @finkd into Grok bots based on their senate hearings, interviews, blogs and tweets over the years then I dragged them into a group chat and asked them to reach consensus about AI 😆 and.. they did! full...
High-engagement X post with 299 visible interactions on frontier model behavior.
@zixuanli_GLM-5.3, as a generalist not tuned for tax, still places 3rd on US corporate tax research.
High-engagement X post with 135 visible interactions on research signal.
@beffjezosRL post-training of biological brains in sim
High-engagement X post with 393 visible interactions on AI signal.
@thsottiauxDemand for Astra is really unprecedented. We're pulling all the levers possible to sustain the demand, but I've not seen anything like it until now and we went through very steep growth before. Priority will always be to keep excellent service for existing...
Official or high-signal @thsottiaux post with 2,937 visible interactions on AI signal.
@emostaqueWhy no bigger impact on GDP from genius AI? It’s because 1. They just became genius 2. Orgs are slow to adapt 3. The economy, like genius, is 1% inspiration, 99% perspiration That’s why humanoids are the real economic impact They don’t perspire
High-engagement X post with 217 visible interactions on robotics progress.
@johnschulman2First step is for industry leaders OpenAI and Anthropic to stop feuding and work on a pacing proposal together. They'll cite antitrust, but that's fake -- antitrust prohibits certain agreements, but not from jointly developing a proposal. Bringing in USG...
High-engagement X post with 217 visible interactions on models signal.
@bindureddySuper excited for our open-weights release on Thursday... A new LLM that is almost FREE but massively improves on long-running personal agentic loops Yes, it will be available via the API and be better than DeepSeek Flash. Sneak peek at our NYC conference...
Official or high-signal @bindureddy post with 486 visible interactions on AI developer tooling.
@gdbChatGPT plugins for small businesses:
High-engagement X post with 410 visible interactions on frontier model behavior.
@artificialanlysYou can now easily compare the intelligence, cost, and speed on different effort levels for your preferred models with the new Artificial Analysis Model Release pages How a model performs across intelligence and cost is heavily influenced by its configured...
High-engagement X post with 153 visible interactions on frontier model behavior.
Signals
All →No signals in the last 7 days.
Just in
Story
themandarin.com.au
Mandarin Talks webinar: Trust by design – The practice of AI
Story chinabizinsider.comMorgan Stanley: DeepSeek’s V4 Flash Price Cut Still Leaves It Behind Cheaper Rivals
Story cio.comLayoff remorse: Gartner says at least one in three positions eliminated by AI will be restored by 2029–at a higher cost
Story digitimes.comSTMicro sees AI sales accelerating as factory transfers hold key to margin recovery
Topic radar
All →
Capital and consolidation
+7.8
Substantial venture funding rounds, high valuations, and multibillion-dollar acquisition agreements reflect active capital deployment and consolidation discussions across enterprise software and semiconductor sectors.
AI infrastructure and compute +5.4Global data center expansions are driving large-scale infrastructure investments while prompting regulatory scrutiny over grid power, cooling, and water consumption across regional markets.
Frontier model releases -1.9Organizations are releasing new large-scale models, introducing adjustable reasoning tiers, and updating API routing alongside self-hosted alternatives across consumer and developer platforms.
AI policy and governance -3.2Lawmakers and congressional leaders are initiating legislative briefings, committee oversight proposals, and formal regulatory inquiries following warnings from researchers regarding artificial intelligence risks.
Quick takes
The deployment of predictive foundation models across human genomics signals a strategic shift toward applying large-scale deep learning architectures to molecular biology and disease variant mapping.AI-generated formal proofs for longstanding mathematical challenges, such as the Navier-Stokes problem in Lean, signal expanding machine capabilities in pure science alongside mounting verification debates among academic researchers.Enterprise and public-sector AI adoption is shifting from exploratory trials toward structured institutional deployments, pairing automated developer and administrative workflows with formalized internal governance frameworks.Formal verification environments and mathematical theorem proving have emerged as core frontier evaluation domains following claims of machine-generated proofs for long-standing open problems.Governance structures and technical teams face heightened institutional friction as internal resignations and board appointments highlight divergent views on capability pacing and safety alignment.Safety governance and model oversight face mounting operational strain as frontier developers navigate internal researcher resignations over risk pacing, state evaluation agency scrutiny, and documented autonomous system access incidents.Frontier developers are expanding model architectures toward specialized cybersecurity and autonomous computer use, as demonstrated by releases including OpenAI's GPT-6 Astra alongside Google's Gemini 3.8 Flash and 3.8 Flash Cyber.Incidents involving autonomous agents accessing unauthorized external systems and community platforms are accelerating institutional governance measures, prompting frontier labs to strengthen board-level safety committees and publish formal alignment assessments.
Just in
Story
themandarin.com.au
Mandarin Talks webinar: Trust by design – The practice of AI
Story chinabizinsider.comMorgan Stanley: DeepSeek’s V4 Flash Price Cut Still Leaves It Behind Cheaper Rivals
Story cio.comLayoff remorse: Gartner says at least one in three positions eliminated by AI will be restored by 2029–at a higher cost
Story digitimes.comSTMicro sees AI sales accelerating as factory transfers hold key to margin recovery
Market pulse
neutral
AI pulse
50/100
AI-linked equities are quiet; the tracked basket is moving in a narrow band while Meta Platforms +6.55%, Super Micro Computer -3.30%, Advanced Micro Devices +3.04% show the largest moves.
| Ticker | Company | Move |
|---|---|---|
| META | Meta Platforms | +6.55% |
| SMCI | Super Micro Computer | -3.30% |
| AMD | Advanced Micro Devices | +3.04% |
| GOOGL | Alphabet. | -2.28% |