Should we combine protocols for AI Control Research?
In AI control[1] research, we've developed many protocols: trusted monitoring, untrusted monitoring, resampling, etc. They each have a different safety-usefulness tradeoff, and labs might use a…
Full feed
Feed 6,506 posts
Every story across every category, newest first. Each card links to the original publisher; daily-brief posts open as editorial pages.
Deploy a Production-Ready NVIDIA AI-Q Blueprint on Oracle Cloud Infrastructure
AI agents have changed a lot in the last two years. The first could only answer one question at a time. Then came multi-turn chat, where…
There's a lot of sloppy thinking around open models. You can ban them and make it impossible for US companies to use them, but this won't stop A) glob…
There's a lot of sloppy thinking around open models. You can ban them and make it impossible for US companies to use them, but this won't…
Why was the denazification of West Germany so successful? Arguably, it's because it was antidemocratic. If the Western Allies had immediately held fre…
Why was the denazification of West Germany so successful? Arguably, it's because it was antidemocratic.If the Western Allies had immediately held free elections after the war,…
RT UWE BOLL: RAW (Official): Coming 2027 From Event Film, CITIZEN VIGILANTE 2. @realDonaldTrump @elonmusk @BollFILMS @CitizenVMovie #UweBoll #CitizenV…
RT UWE BOLL: RAW (Official)Coming 2027 From Event Film, CITIZEN VIGILANTE 2. @realDonaldTrump @elonmusk @BollFILMS @CitizenVMovie #UweBoll #CitizenVigilante
Not making a strong argument is a relief
When I was in middle school, one of our teachers gave us a “don’t do drugs” talk.Somebody asked him whether he had ever used drugs himself.…
Taps sign
Taps signOpenAI: We believe in broad access and plan to make GPT-5.6 Sol, Terra, and Luna generally available in the coming weeks.For now, at the request…
RT Kushal Byatnal: we're hosting a mango tasting next week at the AI eng conference in SF! 🥭 we're literally importing 100s of mangos from India fo…
RT Kushal Byatnalwe're hosting a mango tasting next week at the AI eng conference in SF! 🥭we're literally importing 100s of mangos from India for this…
b9820
sched : reintroduce less synchronizations during split compute (#20793) CUDA: Improve performance via less synchronizations between token (#17795) Adds CPU-to-CUDA copy capability to ggml_backend_cuda_cpy_tensor_async() Adds function…
RT lily zhang: from Write & Learn (thx @MilksandMatcha @philipkiely @swyx), my most important takeaway, 𝐛𝐮𝐢𝐥𝐝 𝐚 𝐝𝐞𝐧𝐬𝐞…
RT lily zhangfrom Write & Learn (thx @MilksandMatcha @philipkiely @swyx), my most important takeaway, 𝐛𝐮𝐢𝐥𝐝 𝐚 𝐝𝐞𝐧𝐬𝐞 𝐫𝐞𝐰𝐚𝐫𝐝 𝐟𝐨𝐫 𝐲𝐨𝐮𝐫 𝐰𝐫𝐢𝐭𝐢𝐧𝐠.fun fact. most people know @swyx…
What happened after 2,000 people tried to hack my AI assistant
What happened after 2,000 people tried to hack my AI assistant Fernando Irarrázaval ran a challenge on hackmyclaw.com to see if anyone could leak secrets held…
Gossamer: a Rust-flavoured language with real goroutines and pause-free memory
Article URL: https://gossamer-lang.org/ Comments URL: https://news.ycombinator.com/item?id=48690231 Points: 19 # Comments: 4
OpenAI limits GPT-5.6 rollout after government request, says restrictions shouldn’t be the norm
“We don’t believe this kind of government access process should become the long-term default,” says OpenAI. “It keeps the best tools from users, developers, enterprises, cyber…
reviews are in ✍️ use Creator Insights to keep improving, just like Violini🙌 ➡️ access at http://c.ai/creator-insights (log into your http://c….
reviews are in ✍️use Creator Insights to keep improving, just like Violini🙌➡️ access at http://c.ai/creator-insights (log into your http://c.ai account to view your insights!)Character.AI: caiMag creator…
Slisp: Simple Lisp compiler (Linux/amd64)
Article URL: https://github.com/skx/slisp Comments URL: https://news.ycombinator.com/item?id=48690200 Points: 34 # Comments: 1
OpenAI's GPT-5.6 Sol launches to rival Claude Mythos under government access rules it calls unsustainable
OpenAI's new flagship GPT-5.6 Sol beats Anthropic's Claude Mythos 5 in coding benchmarks, but the US government is forcing a restricted rollout. OpenAI isn't happy about…
U.S. government will decide who gets to use latest upgrade to ChatGPT
Article URL: https://www.washingtonpost.com/technology/2026/06/26/openai-says-us-government-will-vet-users-its-latest-ai-model/ Comments URL: https://news.ycombinator.com/item?id=48690101 Points: 44 # Comments: 21
have been testing 5.6 for a while and VERY happy with it. DO NOT view this as just a “cyber” release, it is the new sota workhorse model, completely…
have been testing 5.6 for a while and VERY happy with it. DO NOT view this as just a “cyber” release, it is the new sota…
OpenAI poaches Uber India chief to lead its biggest market outside the US
The hire marks OpenAI's latest push into India, expanding offices, partnerships and hiring.
Pre-Modern Armies for Worldbuilders, Part III: Paying for It
Article URL: https://acoup.blog/2026/06/26/collections-pre-modern-armies-for-worldbuilders-part-iii-paying-for-it/ Comments URL: https://news.ycombinator.com/item?id=48689859 Points: 6 # Comments: 0
Incident Report: CVE-2026-LGTM
Incident Report: CVE-2026-LGTM Spectacular hypothetical incident report by Andrew Nesbitt. Day 2, 16:00 UTC --- Two AI review agents from competing vendors, both attached to a…
RT MTS: SITUATION EXPLAINED: 70% of frontier model queries could run locally for free. @ClementDelangue, co-founder and CEO of @huggingface: "There wa…
RT MTSSITUATION EXPLAINED: 70% of frontier model queries could run locally for free.@ClementDelangue, co-founder and CEO of @huggingface:"There was an interesting study from Stanford published last…
Webradio server – broadcasts audio source to clients
Article URL: https://github.com/tau-org/tau-tower Comments URL: https://news.ycombinator.com/item?id=48689760 Points: 5 # Comments: 0
RT Nick Frosst: now seems like a good day to remind people we have an apache 2.0 coding model you can run with 20 gigs of ram locally for free forever…
RT Nick Frosstnow seems like a good day to remind people we have an apache 2.0 coding model you can run with 20 gigs of ram…