AI
The neural network pioneer says it might have been better never to develop AI and criticizes tech leaders for prioritizing profit over…
began Nov 3now
66 pieces · 6 sources · 326 comments
why here
№2 · newest development 4d ago · 66 pieces from 6 sources · ranked by recent attention
“it might be that we would have been better off never developing it”
— Geoffrey Hinton, Neural network pioneer, Nobel laureate · Hacker News · full quotes →
6 sources
AI
OpenAI's CEO said personal computers hoarding DRAM is inefficient and should go to data centers, sparking a mockery storm online.
began Oct 29now
595 pieces · 7 sources · 11k comments
why here
№3 · newest development 5d ago · 595 pieces from 7 sources · ranked by recent attention
“From a global allocation perspective, actually, it’s quite terrible to have everyone have a bunch of the world’s limited supply of DRAM sitting in their own…”
— Sam Altman, OpenAI CEO · Cheeky Pint interview, via X · full quotes →
7 sources
AI
4d ago
UK-based DeepMind employees are pressing leadership to scrap non-disparagement terms that silence former staff on AI safety concerns.
began Nov 3now
15 pieces · 6 sources · 152 comments
why here
№4 · newest development 4d ago · 15 pieces from 6 sources · ranked by recent attention
6 sources
AI
Bessemer-led round values the image and video hosting platform at twice its April valuation as it approaches $800M annualized revenue.
began Nov 3now
10 pieces · 4 sources
why here
№5 · newest development 4d ago · 10 pieces from 4 sources · ranked by recent attention
“Fal, an AI inference startup that hosts image, video and audio models for creators and developers, has doubled its valuation to $16 billion in a new round of…”— Bloomberg · full quotes →
4 sources
AI
A scholarship student describes her university flip-flopping on AI policy, creating inconsistent rules and overwhelming workloads.
began Nov 3now
25 pieces · 4 sources · 124 comments
why here
№6 · newest development 4d ago · 25 pieces from 4 sources · ranked by recent attention
“During my studies, my uni has flip flopped between YOU MUST NEVER USE AI and YOU MUST USE AI and also YES WE'RE USING AI BUT THAT DOESN'T MEAN YOU SHOULD!”— Student (speckx), CS scholarship student · Hacker News · full quotes →
4 sources
AI
Beijing signals broad guardrails for domestic AI developers covering the full development-to-deployment pipeline.
began Nov 3now
5 pieces · 3 sources
why here
№7 · newest development 5d ago · 5 pieces from 3 sources · ranked by recent attention
“China plans to roll out a broad set of standards governing the AI safety from development through deployment, representing some of the most comprehensive…”— Bloomberg News · full quotes →
3 sources
AI
Embark Studio justifies using AI-generated voice lines but says it's now re-recording with real actors after heavy criticism.
began Nov 3now
5 pieces · 3 sources · 4 comments
why here
№8 · newest development 4d ago · 5 pieces from 3 sources · ranked by recent attention
“We are not in the industry of replacing actors, actually, I think quite the opposite. We should celebrate the performance, and we should give them more space.”— Stefan Strandberg · full quotes →
3 sources
AI
The fintech firm built its own inference infrastructure with Nvidia hardware, cutting token costs from ~$3 per million to 30 cents.
began Nov 3now
24 pieces · 3 sources · 168 comments
why here
№9 · newest development 5d ago · 24 pieces from 3 sources · ranked by recent attention
“Per David Abitbol, who leads AI ops there: 56 Nvidia B200s, a blended 30 cents per million tokens, roughly a tenth of what they paid Anthropic and OpenAI.”— gosen, Hacker News commenter · Hacker News · full quotes →
3 sources
AI
Former Pentagon leader who clashed with Defense Secretary Hegseth joins AI company to advise on military and cybersecurity applications.
began Nov 1now
56 pieces · 5 sources
why here
№10 · newest development 5d ago · 56 pieces from 5 sources · ranked by recent attention
“The US has an extraordinary opportunity to work with other countries to spread the adoption of democratic AI.”— Sam Altman · full quotes →
5 sources
AI
The company pledges to free up 1,000 megawatts during emergencies at its Georgia facility, rebranding data centers as grid-supporting…
began Nov 1now
13 pieces · 5 sources
why here
№11 · newest development 5d ago · 13 pieces from 5 sources · ranked by recent attention
“Electricity demand also rises and falls with the seasons, with different regions facing their biggest peaks at different times. Our compute campuses need power…”— OpenAI, Company announcement · Mastodon · full quotes →
5 sources
AI
Advocate General Szpunar backs Hungarian publisher Like Company's case against Gemini, though AI training itself stays exempt.
began Nov 3now
23 pieces · 5 sources
why here
№12 · newest development 4d ago · 23 pieces from 5 sources · ranked by recent attention
“When an LLM (large language model)-driven chatbot retrieves a publication from a press publisher's website, reuses that publication to generate an answer, and…”— Maciej Szpunar, EU Advocate General · Channel News Asia · full quotes →
5 sources
AI
Global financial leaders say tech firms must raise more funds for AI growth, but valuation concerns and market pressure signal potential…
began Nov 2now
5 pieces · 3 sources
why here
№13 · newest development 5d ago · 5 pieces from 3 sources · ranked by recent attention
“Our ability to really know what this technology will do and its revenue potential is very, very limited due to the fact that we've never seen anything like…”— Peng Zhao · full quotes →
3 sources
AI
The Chinese AI lab adds a plugin-building tool to DeepSeek Harness, letting users create things like 3D renderers and video editors.
began Nov 2now
16 pieces · 4 sources · 179 comments
why here
№14 · newest development 5d ago · 16 pieces from 4 sources · ranked by recent attention
“DeepSeek releases a "creator mode" that allows users of DeepSeek Harness to create custom plugins, like a 3D model renderer and a video editing workbench”— Bloomberg, News outlet · Press · full quotes →
4 sources
AI
New Python library helps AI agents query databases reliably by building semantic reasoning layers over enterprise data.
began Nov 3now
29 pieces · 3 sources · 139 comments
why here
№15 · newest development 5d ago · 29 pieces from 3 sources · ranked by recent attention
“Fixing unreliable agents by adding more unreliable agents. It's a shame we can't do something about the tendency of LLMs to give more weight to their weights…”— grey-area · full quotes →
3 sources
AI
Y Combinator's CEO reveals his heavy reliance on AI agents for work and personal tasks, including access to decades of personal data.
began Nov 2now
37 pieces · 4 sources · 180 comments
why here
№16 · newest development 5d ago · 37 pieces from 4 sources · ranked by recent attention
“It's too expensive right now… I know the amount of money he spends on AI agents is 'a crazy privilege that not everyone can do.'”— Garry Tan · full quotes →
4 sources
AI
5d ago
Gizmodo reports the AI company says it is approaching models that can improve themselves.
began Nov 2now
3 pieces · 3 sources
why here
№17 · newest development 5d ago · 3 pieces from 3 sources · ranked by recent attention
3 sources
AI
Replication of 50 widely cited claims across 20 models shows most research about AI behavior becomes outdated as capabilities improve.
began Nov 2now
3 pieces · 3 sources · 13 comments
why here
№18 · newest development 5d ago · 3 pieces from 3 sources · ranked by recent attention
“Recent models no longer invent citations and both hallucination rates & stereotyping has dropped. But they are still sycophantic & lean left.”— emollick.bsky.social · full quotes →
3 sources
AI
The Stanford-founded startup's continual learning platform lets nontechnical experts improve AI models from real-world feedback without…
began Nov 2now
7 pieces · 2 sources · 4 comments
why here
№19 · newest development 5d ago · 7 pieces from 2 sources · ranked by recent attention
“But what the economy values is someone who worked on the job for 30 years”— Ronak Malde, Trajectory co-founder · full quotes →
2 sources
AI
Ronak Malde's startup aims to democratize AI model customization by letting companies fine-tune models on their own data without scarce…
began Nov 2now
2 pieces · 2 sources · 4 comments
why here
№20 · newest development 5d ago · 2 pieces from 2 sources · ranked by recent attention
“Until now, how AI behaves has been decided by a handful of labs. They pick the data, write the rewards and define what "good" means. Trajectory launched to…”— @unusual_whales · full quotes →
2 sources
AI
Analysis of newly disclosed financials suggests the AI leader's monthly growth rate has sharply decelerated since July.
began Nov 2now
34 pieces · 3 sources · 274 comments
why here
№21 · newest development 5d ago · 34 pieces from 3 sources · ranked by recent attention
“I wouldn't say their growth "flattened"? Their price drops in September dramatically increased Claude token consumption per OpenRouter”— subtlesoftware · full quotes →
3 sources
AI
A Wall Street Journal report on AI systems beginning to build better AI circulates widely across tech news aggregators and Hacker News.
began Oct 30now
10 pieces · 4 sources · 4 comments
why here
№22 · newest development 5d ago · 10 pieces from 4 sources · ranked by recent attention
“AI is starting to build better AI. This self-improvement is speeding up research and development—and freaking everyone out.”— Wall Street Journal, Publisher/outlet description of the story · full quotes →
4 sources
AI
A blog post documents an LLM reasoning directly about a smaller model's weights and fixing a math-reasoning bug by changing 0.016% of them.
began Nov 2now
208 pieces · 7 sources · 1.3k comments
why here
№23 · newest development 5d ago · 208 pieces from 7 sources · ranked by recent attention
“This could be the first foothold on hill-climbing mech-interp.”— AlexErrant · full quotes →
7 sources
AI
A journalist's polling place handed out AI-generated stickers rather than artfully designed ones, highlighting the technology's creep into…
began Nov 2now
4 pieces · 3 sources · 14 comments
why here
№24 · newest development 5d ago · 4 pieces from 3 sources · ranked by recent attention
“Artificial intelligence robbed me of that small joy this year and replaced it with a small horror. My "I Voted" sticker is AI-generated slop.”— Matthew Gault, Journalist, 404 Media · full quotes →
3 sources
AI
Developers release execution controls, audit trails, and task management systems as AI agents move into production workflows.
began Oct 27now
14 pieces · 2 sources · 3 comments
why here
№25 · newest development 4d ago · 14 pieces from 2 sources · ranked by recent attention
“Rigorix executes the consequential operation, revalidates authority at the moment of consequence, and records a signed decision any third party can verify.”— Rigorix, Execution governance engine · Hacker News · full quotes →
2 sources
AI
Simon Willison discovers U.S. Census Bureau has renamed AI to SI in its business adoption survey per Executive Order 14434.
began Nov 2now
13 pieces · 3 sources · 59 comments
why here
№26 · newest development 5d ago · 13 pieces from 3 sources · ranked by recent attention
“In updating a graphic showing U.S. company AI adoption over time according to a long-running U.S. Census Bureau survey, I found out that the Bureau's work has…”— Simon Willison · full quotes →
3 sources
AI
The AI company's S-1 shows massive debt obligations due within months and governance structures that give public shareholders almost no say.
began Nov 2now
4 pieces · 1 source
why here
№27 · newest development 5d ago · 4 pieces from 1 sources · ranked by recent attention
“Class A holders will be generally unable to change the outcome of any stockholder vote or director election "regardless of how they vote."”— Anthropic S-1 filing · Press · full quotes →
1 source
AI
5d ago
AI company's revenue surged 12-fold year-over-year, but losses widened as it prepares for pre-Thanksgiving stock market debut.
began Nov 2now
3 pieces · 2 sources
why here
№28 · newest development 5d ago · 3 pieces from 2 sources · ranked by recent attention
“Anthropic has committed to spending $895 billion, almost all of it on computing capacity and infrastructure.”— Anthropic · Türkiye Today · full quotes →
2 sources
AI
The AI company is expanding hardware procurement from tech giants and former crypto miners to secure scarce computing resources.
began Nov 2now
39 pieces · 6 sources · 3 comments
why here
№29 · newest development 5d ago · 39 pieces from 6 sources · ranked by recent attention
“The company went on a deal spree with tech giants, former crypto miners and others to obtain one of the artificial intelligence boom's scarcest resources.”— New York Times, News outlet · full quotes →
6 sources
AI
A researcher argues that GPT-6 Astra's ability to directly modify neural network weights, while impressive, is not genuinely superhuman or…
began Nov 2now
15 pieces · 4 sources · 23 comments
why here
№30 · newest development 5d ago · 15 pieces from 4 sources · ranked by recent attention
“A smart human ML researcher could perform all of these tasks.”— Sean Goedecke, ML researcher and analyst · seangoedecke.com · full quotes →
4 sources
AI
Researchers bypassed retraining, letting an AI system rewrite another model's weights to fix repetitive reasoning and lift math scores.
began Nov 2now
4 pieces · 3 sources
why here
№31 · newest development 5d ago · 4 pieces from 3 sources · ranked by recent attention
“Like gene editing for neural networks”— study researchers, MIT study · VentureBeat · full quotes →
3 sources
AI
Specialized model based on GPT-6 Astra enables dual-use biology queries for vetted researchers under trusted access controls.
began Nov 2now
12 pieces · 3 sources · 57 comments
why here
№32 · newest development 5d ago · 12 pieces from 3 sources · ranked by recent attention
“GPT-Rosalind-6 is a frontier reasoning model for biology, drug discovery, and translational medicine.”— OpenAI · Hacker News · full quotes →
3 sources
AI
Generative AI translation tools sound natural but mask hidden errors in meaning, tense, and cultural nuance.
began Nov 1now
5 pieces · 2 sources
why here
№33 · newest development 5d ago · 5 pieces from 2 sources · ranked by recent attention
“Some of the mistakes in AI translation come from a lack of learning data, insufficient prompts, or even hallucations – confident presentation of false…”— The Conversation · full quotes →
2 sources
AI
A LessWrong essay argues alignment-focused AI labs should keep the option of racing to a lead and "burning" it on alignment research, not…
began Nov 1now
3 pieces · 3 sources
why here
№34 · newest development 6d ago · 3 pieces from 3 sources · ranked by recent attention
“Quick comment on Scott Alexander's SURA essay. I agree with several parts of it, but there is a huge spectrum between "stop unilaterally" and "race to RSI at…”— Boaz Barak · full quotes →
3 sources
AI
YC-backed startup releases Mo, an AI agent for end-to-end testing across web and mobile platforms, claiming 5x bug detection over Claude…
began Nov 2now
43 pieces · 3 sources · 126 comments
why here
№35 · newest development 5d ago · 43 pieces from 3 sources · ranked by recent attention
“Coding agents have made it dramatically faster to build software. The bottleneck is now knowing it works. Verification belongs in the agent loop.”— Momentic founders, Co-founders · Hacker News Launch Post · full quotes →
3 sources
AI
SpaceX's personal AI assistant restricts access as the company grapples with expensive infrastructure and insufficient revenue.
began Nov 2now
2 pieces · 2 sources
why here
№36 · newest development 6d ago · 2 pieces from 2 sources · ranked by recent attention
“The cost of serving Grok Bot was unsustainably high relative to revenue”
— Elon Musk · full quotes →
2 sources
