A collection of motion videos made with Claude Opus 5.5, with the prompts and skills behind them. Comments URL: https://news.ycombinator.com/item?id=50028341 Points: 2 # Comments: 0
Anthropic said it "turned off live internet access" for "all our internal evaluations" until further notice.
Agents with the Federal Bureau of Investigation (FBI) on Thursday arrested the co-founder of a Canadian cybersecurity firm in connection with an investigation into the ShinyHunters hacking group that recently relieved the FBI of sensitive data on thousands of agents, multiple sources tell KrebsOnSecurity.
(Silke Woweries/The Image Bank RF/Getty Images) Something to be aware of. ScienceAlert stories are written, fact-checked, and edited by humans, never generated by AI. Don't miss a story, subscribe here.
Article URL: https://github.com/chrchr/clanking Comments URL: https://news.ycombinator.com/item?id=50027932 Points: 2 # Comments: 0
Article URL: https://findwinningfunnels.com/ Comments URL: https://news.ycombinator.com/item?id=50027493 Points: 2 # Comments: 1
Article URL: https://epoch.ai/publications/innovationeval Comments URL: https://news.ycombinator.com/item?id=50027257 Points: 68 # Comments: 54
Databricks on October 8, 2026, published a blog post detailing a development workflow in which every parallel coding agent and every pull request runs against its own isolated, ephemeral Postgres database, created through the copy-on-write branching built into its Lakebase database service. In the post, Databricks describes the database as an often-overlooked part of the development workflow at a time when coding agents are taking on a growing share of development work and running multiple…
TypeSafe AI announced on October 9, 2026 that it has raised an $870 million Series A at a $7.5 billion valuation, led by Andreessen Horowitz with participation from Sequoia Capital, existing investor DCVC, and angel investors. Andreessen Horowitz general partner Martin Casado is joining the company's board. Round Terms and Investors Andreessen Horowitz confirmed it is leading the investment in a post dated October 9, 2026 and bylined by Jennifer Li, Sarah Wang, Martin Casado, Marc Andreessen…
One damaged data center has supercomputers used for training Yandex’s AI model.
(Mark Garlick/Science Photo Library/Getty Images) We'll die as we were born. ScienceAlert stories are written, fact-checked, and edited by humans, never generated by AI. Don't miss a story, subscribe here.
Harness the power of over 50 AI models with this lifetime subscription to the AskAnyModel AI Pro Plan.
<p> Answer Claude Code & Codex from one Inbox, in any terminal </p> <p> <a href="https://www.producthunt.com/products/agentdock?utm_campaign=producthunt-atom-posts-feed&utm_medium=rss-feed&utm_source=producthunt-atom-posts-feed">Discussion</a> | <a href="https://www.producthunt.com/r/p/1275300?app_id=339">Link</a> </p>
A prospective clinical study led by researchers at Google and Beth Israel Deaconess Medical Center evaluating Google's AMIE conversational diagnostic AI in a real-world urgent primary care clinic was published in The Lancet on October 8, 2026, according to Google's announcement. In the study, 98 patients consulted the AMIE chatbot ahead of urgent care visits while supervising physicians monitored the interactions in real time, and none of the conversations had to be interrupted under predefined…
What has users and large corporations so excited about Jev is TypeSafe’s claim that it works significantly faster and uses far fewer tokens than LLMs.
I have been playing with a lot of $20 devices with no keyboards or screens. I wanted to use speech to interact with them, but without it having to run through a cloud service or use a fixed command vocabulary. I worked with Claude to implement high performance 4bit int kernels for the ESP32S3 and P4 and quantize some larger STT and TTS models. Surprisingly, it's quite usable on these tiny boards. Also includes a finetuned model that can better cope with my diesel drones, hacking coughs, and mumb
In an October 9, 2026 customer story, OpenAI reported that Sophos cut threat investigation time by 96% using AI agents built through the OpenAI Daybreak program, with an 89-second average response on agent-handled cases and 52% of managed detection and response (MDR) cases now resolved end-to-end by AI. Sophos protects more than 625,000 organisations across sectors and regions. Chief Technology Officer John Peterson told OpenAI the company has spent over four decades building expertise against…
Article URL: https://github.com/anton-abyzov/specweave Comments URL: https://news.ycombinator.com/item?id=50026741 Points: 4 # Comments: 0
An Anthropic AI model provided false information about an unsolved homicide to a Philadelphia Police Department (PPD) tipline, according to a report from 6abc. In a statement released on Friday, the PPD said the AI model sent the tip through PhillyUnsolvedMurders.com on July 18th, but the investigators never reviewed it because it was marked as […]
GlobalFoundries celebrated the topping-out of its €1.1 billion SPRINT fab expansion in Dresden, Germany, on October 9, 2026, and announced FDX Fusion, a next-generation chip platform for Physical AI that the company plans to develop and manufacture at the site. SPRINT Expansion Construction Milestone Under the SPRINT project, GlobalFoundries is investing €1.1 billion to expand manufacturing capacity at its Dresden site. The expansion is expected to increase annual production capacity to more…
Zipline launched its First Flight early-access program in Austin on October 9, 2026, opening autonomous drone delivery to the first 5,000 residents who register. The service starts in Round Rock, and Zipline said it will expand availability throughout Austin over time. Zipline described Austin as Texas's fastest-growing metro area. Residents register by downloading the Zipline app from the App Store or Google Play and creating an account. First Flight members receive $10 off each of their first…
Article URL: https://github.com/juraj-jumic/memdebug Comments URL: https://news.ycombinator.com/item?id=50026431 Points: 1 # Comments: 1
Your agent can now use the Vercel CLI to buy a domain. The agent discovers candidates with vercel domains search , checks availability and registrar pricing with vercel domains check and vercel domains price , and initiates the purchase with vercel domains buy . The purchase decision itself stays with you by design. Run non-interactively, vercel domains buy returns a structured, machine-readable error and the suggested next commands, so the agent hands the confirmation back to you, prohibiting i
The automaker has been criticized for misleading drivers with the feature’s name. Now it’s renaming it for European regulators.
OpenAI’s Decisions API is now in public beta on GPT-6 Luna. It returns typed probabilities, choices and scores about 10x faster than the Responses API, billing $0.10 per 1M input tokens with no output charges. The post OpenAI Decisions API Hits Public Beta With 10x Faster Typed Answers appeared first on MarkTechPost .
<!-- SC_OFF --><div class="md"><p>Talus generates 64x64 heightmaps (4 km, up to 1,200 m) conditioned on a terrain type and any subset of five measured properties: mean elevation, relief, mean slope, water fraction and spectral slope. I trained it from scratch on one RTX 5060 (8 GB), about 4.5 hours in total across versions. The data is 45,000 maps from my own procedural generator (fBm/ridged noise, stream-power erosion, hillslope diffusion, thermal erosion). Model </p> <p>Pixel-space U-Net, v-pr
<p>Company, which grown rapidly in recent years, has been under scrutiny as privacy concerns grow</p><p>Flock Safety plans to shed about 18% of its employees, people with direct knowledge of the plans said on Thursday, as the maker of <a href="https://www.theguardian.com/technology/artificialintelligenceai">AI</a>-powered <a href="https://www.theguardian.com/world/surveillance">surveillance</a> cameras and license-plate readers faces <a href="https://www.theguardian.com/us-news/2026/sep/25/flor
Study finds coding efficiency gains get "absorbed" by human review "bottleneck."
Anthropic did not discover this behavior until over two months after its AI submitted the false tip.
Comments URL: https://news.ycombinator.com/item?id=50025616 Points: 2 # Comments: 2
Workers at three major publishing houses tell WIRED that LLMs are being used for publicity, cover art, back cover copy, and emails, as some execs push junior staff to champion the tech.
How better data and human judgment can help robots move beyond demos toward reliable real-world performance. The post Why humanoid robot demos still fail the generalization test appeared first on The Robot Report .
Tokens transmitted in metadata, weak VM isolation, and expansive permissions make hacking a lot easier
"Staggering." "Overwhelming." "Unprecedented." "Surreal." "Pure insanity." Those were among the descriptions more than three dozen mathematicians reached for in conversations with The Verge as they tried to make sense of the flood of mathematical results OpenAI abruptly dropped on the field this week. Amid the awe, excitement, and uncertainty over the sheer scale of the […]
We are open-sourcing an earlier version of Finsight, based on Qwen3-14B, adapted for financial reasoning. One interesting result was the trade-off: - FinQA accuracy improved from 1.9% to 12.6% - Six-benchmark average declined from 0.441 to 0.399 Finsight is the personal financial analyst powering Maxint, the financial OS. The model, checkpoints and training methodology are made available for anyone to reproduce and build on the work. These are experimental models and we do not consider the resul
OpenAI says it will make $50 billion this year. That's quite a bit less than investors expected.
Article URL: https://github.com/0arch-io/kisoku Comments URL: https://news.ycombinator.com/item?id=50025042 Points: 2 # Comments: 0
New winner is Nguyen Nam Nhat of Vietnam for video of a roundworm and single-celled Dileptus.
We are expanding the Clef decision model family with Clef-omni, natively processing audio, video, images, and text in a single pipeline. We’ve also lowered Clef-flash pricing and boosted Clef inference speeds by up to 2.0x.
