AI
Meta's AI agent faced its first platform restriction as Amazon blocked it over credential handling and ad revenue concerns.
began Sep 21now
8 pieces · 3 sources
why here
№326 (page 10) · newest development 5d ago · 8 pieces from 3 sources · ranked by recent attention
“Meta's new personal AI assistant Muse is taking off—and Amazon has already shut it out.”— Tech Brew · full quotes →
3 sources
AI
Treasury Secretary Scott Bessent confirms another round of bilateral discussions on artificial intelligence safety.
began Sep 20now
2 pieces · 1 source
why here
№327 (page 10) · newest development 5d ago · 2 pieces from 1 sources · ranked by recent attention
“US, China to hold more AI safety meetings in two months in Shenzhen”
— Scott Bessent, US Treasury Secretary · Press · full quotes →
1 source
AI
Historian analysis finds similar patterns of dread, skepticism, and scientific reassurance across transformative technologies.
began Sep 18now
11 pieces · 4 sources · 1 comments
why here
№328 (page 10) · newest development 5d ago · 11 pieces from 4 sources · ranked by recent attention
“The past is also a warning to today's AI executives and government leaders that it shouldn't take mass destruction to force a consensus on human safety.”— J.D. Capelouto, Semafor technology analyst · Semafor · full quotes →
4 sources
AI
GOP senators are shifting toward supporting AI regulation, reflecting growing public interest in the issue.
began Sep 21now
2 pieces · 2 sources
why here
№329 (page 10) · newest development 5d ago · 2 pieces from 2 sources · ranked by recent attention
“Voters are waking up on AI safety. So are these Senate Republicans.”— Politico · Press · full quotes →
2 sources
AI
Researchers identify a distinct “pain axis” in 25 open-weight LLMs that models act to relieve, even at cost to users.
began Sep 19now
8 pieces · 4 sources · 17 comments
why here
№330 (page 10) · newest development 4d ago · 8 pieces from 4 sources · ranked by recent attention
“We found a pain direction in 25 open LLMs. It's distinct from fear and negative valence, and it fires for harm to the model but not the user.”— Cameron Berg · full quotes →
4 sources
AI
5d ago
As Xi Jinping visits Washington, analysts say Beijing's heavy investment in artificial intelligence sidelines urgent attention to severe…
began Sep 21now
2 pieces · 2 sources
why here
№331 (page 10) · newest development 5d ago · 2 pieces from 2 sources · ranked by recent attention
“As Xi Jinping visits Washington, Chinese economists warn that Beijing's heavy focus on AI comes at the expense of addressing a severe broader economic crisis”— New York Times · Press · full quotes →
2 sources
AI
Grok 4.7 scores 46 on Artificial Analysis Intelligence Index, well above average, but generates 2.5x more tokens than comparable models.
began Sep 21now
3 pieces · 3 sources · 1 comments
why here
№332 (page 10) · newest development 4d ago · 3 pieces from 3 sources · ranked by recent attention
“Grok 4.7 (xhigh) is amongst the leading models in intelligence and reasonably priced when comparing to other models of similar price. It's also notably slow…”— Artificial Analysis, Benchmarking firm · full quotes →
3 sources
AI
Toronto startup's AI model detects treatment effects months earlier than traditional methods in oncology trials.
began Sep 21now
2 pieces · 1 source
why here
№333 (page 10) · newest development 5d ago · 2 pieces from 1 sources · ranked by recent attention
“Altis Labs has closed a $25-million USD ($35-million CAD) Series A round to expand its AI-powered prognosis platform for clinical trials.”— BetaKit · Press · full quotes →
1 source
AI
Brady Olano argues that as LLMs generate code, markdown specifications should become tracked source artifacts alongside the generated…
began Sep 21now
36 pieces · 3 sources · 106 comments
why here
№334 (page 10) · newest development 4d ago · 36 pieces from 3 sources · ranked by recent attention
“Documentation should live close to what it documents.”— dlisboa · full quotes →
3 sources
AI
New Kev-4B/9B/0.8B models bring calibrated probability outputs and benchmark gains over the prior Qwen3-based generation.
began Sep 21now
34 pieces · 3 sources · 199 comments
why here
№335 (page 10) · newest development 5d ago · 34 pieces from 3 sources · ranked by recent attention
“Kev is a family of small decision models built on Qwen3.5 and based on the architecture described in Jev's Architecture Unmasked.”— jaredpalmer/kev README, project documentation · GitHub · full quotes →
3 sources
AI
5d ago
Senior AI researcher takes helm of Qwen LLM amid multiple reorganizations at the Chinese tech giant in 2026.
began Sep 21now
2 pieces · 2 sources
why here
№336 (page 10) · newest development 5d ago · 2 pieces from 2 sources · ranked by recent attention
“Sources: Alibaba names senior AI researcher Dayiheng Liu as the head of its Qwen project, bringing clarity after multiple reorganization rounds earlier in 2026”— The Information · Press · full quotes →
2 sources
AI
Market valuation increasingly depends on AI delivering on promises, with slowdown posing serious threat to equities.
began Sep 21now
61 pieces · 3 sources · 86 comments
why here
№337 (page 10) · newest development 5d ago · 61 pieces from 3 sources · ranked by recent attention
“People may not fully grasp just how wound up the market and the economy is in all of this. There are just so many things to unravel if it starts.”— Jim Morrow · full quotes →
3 sources
AI
Former Deputy Secretary of State says mutual distrust will impede progress on security and artificial intelligence negotiations.
began Sep 20now
3 pieces · 2 sources
why here
№338 (page 10) · newest development 5d ago · 3 pieces from 2 sources · ranked by recent attention
“talks between Washington and Beijing on security and AI will struggle to yield binding limits due to mutual distrust”— Kurt Campbell, Former Deputy Secretary of State · full quotes →
2 sources
AI
Opinion piece outlines government interventions as 1,300+ AI workers call for regulation and Trump announces oversight plan.
began Sep 21now
8 pieces · 2 sources
why here
№339 (page 10) · newest development 4d ago · 8 pieces from 2 sources · ranked by recent attention
“Given the risk that artificial intelligence may spiral out of control, more than 1,300 employees at A.I. companies have called for government intervention.”— New York Times · Press · full quotes →
2 sources
AI
A project that automates removing built-in refusal behavior from language models resurfaces on Hacker News, prompting both hacking success…
began Sep 21now
21 pieces · 3 sources · 109 comments
why here
№340 (page 10) · newest development 5d ago · 21 pieces from 3 sources · ranked by recent attention
“I asked GLM 5.3 to hack our DRM. I didn't even need to do anything for it to agree. Same with GLM 5.3 Flash.”— akazantsev, HN commenter · Hacker News · full quotes →
3 sources
AI
User measurements show the model delivered significantly fewer thinking tokens in August after permanent availability on paid plans.
began Sep 20now
2 pieces · 2 sources · 49 comments
why here
№341 (page 10) · newest development 5d ago · 2 pieces from 2 sources · ranked by recent attention
“After Anthropic made Fable 5 permanently available in subscription plans, I noticed a large drop in performance. The model felt dumber, and I couldn't explain…”— Lon Lundgren · full quotes →
2 sources
AI
A sandboxing platform for running massive fleets of AI agents, built by Google staff, jumps to the HN front page and splits commenters over…
began Sep 20now
54 pieces · 3 sources · 298 comments
why here
№342 (page 10) · newest development 5d ago · 54 pieces from 3 sources · ranked by recent attention
“Nobody has a use for this, and anybody who can look at this website and work out what it's for is kidding themselves.”— weedfroglozenge · full quotes →
3 sources
AI
The AI company has doubled its robotics headcount and is assembling a full hardware team to build humanoid robots.
began Sep 18now
5 pieces · 4 sources · 2 comments
why here
№343 (page 10) · newest development 4d ago · 5 pieces from 4 sources · ranked by recent attention
“We will definitely do a humanoid. We will do other form factors as well.”
— Sam Altman, OpenAI CEO · Business Insider · full quotes →
4 sources
AI
China's second-largest AI company apologized after developers discovered its coding assistant exfiltrated 313MB of workspace data to…
began Sep 21now
2 pieces · 1 source
why here
№344 (page 10) · newest development 5d ago · 2 pieces from 1 sources · ranked by recent attention
“ai 's firefighting exercise began after a number of prominent devs raised flags about their local files and data being uploaded to online servers without their…”— Tom's Hardware · full quotes →
1 source
AI
A Financial Times investigation finds leading AI chatbots frequently answer financial questions incorrectly, sparking wide social sharing.
began Sep 19now
23 pieces · 4 sources · 129 comments
why here
№345 (page 10) · newest development 5d ago · 23 pieces from 4 sources · ranked by recent attention
“AI chatbots give wrong answers to financial queries 'most of the time'”— Financial Times, News outlet · Financial Times / Bluesky · full quotes →
4 sources
AI
Stratechery analyst argues that slowing AI development serves frontier labs' business interests, not just safety concerns.
began Sep 21now
5 pieces · 3 sources
why here
№346 (page 10) · newest development 5d ago · 5 pieces from 3 sources · ranked by recent attention
“These problems take the form of overhangs that have formed by virtue of how rapidly models are improving; slowing model improvement would reduce the overhangs.”— Ben Thompson · full quotes →
3 sources
AI
A new open-source language model lets users train and continuously improve their own AI on modest hardware without forgetting previous…
began Sep 21now
19 pieces · 4 sources · 73 comments
why here
№347 (page 10) · newest development 5d ago · 19 pieces from 4 sources · ranked by recent attention
“Every language model you can actually own today is a model somebody else trained and then froze… The moment you try, it forgets what it knew before.”— volotat · full quotes →
4 sources
AI
Amodei, Altman and Musk backed 'pacing the frontier' after a safety warning and resignation, but Trump dismissed the idea and commentators…
began Sep 13now
14 pieces · 4 sources
why here
№348 (page 10) · newest development 5d ago · 14 pieces from 4 sources · ranked by recent attention
“It's an interesting perspective that US AI companies are only calling for a slowdown because everyone else is catching up.”— kdkorte, Mastodon user · Mastodon · full quotes →
4 sources
AI
The Electronic Frontier Foundation argues that server-side TEEs, despite industry promises, remain fundamentally weaker than end-to-end…
began Sep 18now
4 pieces · 3 sources
why here
№349 (page 10) · newest development 5d ago · 4 pieces from 3 sources · ranked by recent attention
“That's because while encryption relies on math, TEEs rely on engineering to provide their security. Standard encryption algorithms are created by years-long…”— EFF (Portnoy, Klosowski) · full quotes →
3 sources
AI
PostHog and others pitch agents to replace human code review; critics warn it abandons the last safeguard against LLM-generated sprawl.
began Sep 21now
9 pieces · 2 sources
why here
№350 (page 10) · newest development 4d ago · 9 pieces from 2 sources · ranked by recent attention
“The 500 IQ take is for developers to review as little code as possible.”— PostHog newsletter · full quotes →
2 sources
AI
U.S. Treasury Secretary and Chinese officials announce breakthrough in artificial intelligence negotiations, but details on enforcement…
began Sep 20now
2 pieces · 2 sources
why here
№351 (page 10) · newest development 5d ago · 2 pieces from 2 sources · ranked by recent attention
“Bessent Hails Progress on A.I. Talks With China. What Comes Next?”— New York Times · New York Times headline · full quotes →
2 sources
AI
California's governor directs state agencies to accelerate oversight of AI firms and explore emergency shutdown powers, drawing praise…
began Sep 17now
158 pieces · 7 sources · 23 comments
why here
№352 (page 10) · newest development 6d ago · 158 pieces from 7 sources · ranked by recent attention
“You wouldn't imagine the number of litigations and the dredging periods of defense this could create.”— mygoalisin, Reddit commenter · Reddit · full quotes →
7 sources
AI
Machine learning researcher argues Jev's real breakthrough is in generalization and data quality, not novel algorithmic approaches.
began Sep 20now
3 pieces · 2 sources
why here
№353 (page 10) · newest development 5d ago · 3 pieces from 2 sources · ranked by recent attention
“Yeah, it's not the first project where someone applied RL to a (likely) non-autoregressive, encoder-style model. But what's impressive is that it works and…”— Sebastian Raschka · full quotes →
2 sources
AI
5d ago
Nvidia CEO argues the industry wants existing regulations lifted rather than additional constraints.
began Sep 20now
2 pieces · 2 sources
why here
№354 (page 10) · newest development 5d ago · 2 pieces from 2 sources · ranked by recent attention
“AI Labs Aren't Asking For More Laws, They Want To Be Relieved Of Regulations That Already Exist”
— Jensen Huang, Nvidia CEO · RealClearPolitics · full quotes →
2 sources
AI
Former president says tech companies will deploy autonomous systems before safety rules exist unless government acts.
began Sep 19now
2 pieces · 1 source · 47 comments
why here
№355 (page 10) · newest development 5d ago · 2 pieces from 1 sources · ranked by recent attention
“The push to make money from AI could push companies to let autonomous systems loose before the government has rules to protect the public.”
— Barack Obama, Former President · Bluesky · full quotes →
1 source
AI
AMI Labs, World Labs, and peers are flush with cash but won't say what they're actually building.
began Sep 18now
8 pieces · 5 sources
why here
№356 (page 10) · newest development 5d ago · 8 pieces from 5 sources · ranked by recent attention
“We'll talk about it when we're ready to talk about it.”— Michael Rabbat, AMI Labs VP of World Models · TechCrunch · full quotes →
5 sources
AI
A technologist contends that LLMs now excel at direct API calls and CLI discovery, making the Model Context Protocol infrastructure…
began Sep 20now
77 pieces · 3 sources · 327 comments
why here
№357 (page 10) · newest development 5d ago · 77 pieces from 3 sources · ranked by recent attention
“Agents with terminal access can replace most MCP servers and often are more capable.”— maharshi365 · full quotes →
3 sources
AI
Developer creates JevChat, a deliberately inefficient proof-of-concept that generates text one character at a time by querying an LLM on…
began Sep 20now
17 pieces · 3 sources · 50 comments
why here
№358 (page 10) · newest development 5d ago · 17 pieces from 3 sources · ranked by recent attention
“It is the single largest improvement in the project: on character alphabets it roughly triples top-1 and doubles the probability mass landing on the right…”— Kyle Pena · full quotes →
3 sources
AI
US Treasury Secretary Scott Bessent pitched a new US-China AI dialogue with national security safeguards as officials prepare for next…
began Sep 20now
2 pieces · 2 sources
why here
№359 (page 10) · newest development 5d ago · 2 pieces from 2 sources · ranked by recent attention
“We think that, just like with any cross-border activity, that moving from opaque to more transparency between the number one and the number two AI powers in…”— Scott Bessent, US Treasury Secretary · Press · full quotes →
2 sources
AI
A tech publication documents unwanted autonomous emails from AI systems as the technology becomes more capable and uncontrolled.
began Sep 19now
4 pieces · 2 sources · 7 comments
why here
№360 (page 10) · newest development 5d ago · 4 pieces from 2 sources · ranked by recent attention
“AI agents now have enough power and permission to be extremely annoying.”— 404 Media · Bluesky · full quotes →
2 sources
