News
The raw AI news stream — everything we ingest across RSS, Hacker News, arXiv, GitHub, Bluesky, YouTube and more, newest first.
| Source | Age | Title | Summary | Status |
|---|---|---|---|---|
rss techmeme | 1mo | Sources: US and Russian diplomats worked to weaken an AI weapons pact at the UN this month, removing a requirement that humans review AI-generated targets, more (Pranshu Verma/Washington Post) | Pranshu Verma / Washington Post: Sources: US and Russian diplomats worked to weaken an AI weapons pact at the UN this month, removing… | ↗Not briefed |
bluesky bluesky-ai | 1mo | wild how this was late | wild how this was late | ↗Not briefed |
bluesky bluesky-ai | 1mo | I mean. DON'T do that, but there's just SO much AI shit out there people try to pretend is bespoke but is just a different wrapper on the s… | I mean. DON'T do that, but there's just SO much AI shit out there people try to pretend is bespoke but is just a different wrapper on… | ↗Not briefed |
bluesky bluesky-ai | 1mo | openai.com/index/scalin... It's incredibly clear they've been working *around* the original architecture just to scale as ChatGPT exploded…🔧 ChatGPT | openai.com/index/scalin... It's incredibly clear they've been working *around* the original architecture just to scale as ChatGPT exp… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Pawl – iOS gambling blocker you unlock with a physical Security Key | Hi HN!<p>I built this to make unblocking apps on your phone a lot harder. You add a physical key, then put it out of reach or give it… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The creators are no longer in control. OpenAI bots meddled with multiple US government agency sites. www.bbc.com/news/article... | The creators are no longer in control. OpenAI bots meddled with multiple US government agency sites. www.bbc.com/news/article... | ↗Not briefed |
bluesky bluesky-ai | 1mo | Enough is Enough: OpenAI Hacking Revelation Demands DOJ and Congressional Investigations. www.citizen.org/news/enough-... | Enough is Enough: OpenAI Hacking Revelation Demands DOJ and Congressional Investigations. www.citizen.org/news/enough-... | ↗Not briefed |
bluesky bluesky-ai | 1mo | The creators have lost control of their monster. OpenAI says its bots have interacted with multiple U.S. government sites in unexpected AI… | The creators have lost control of their monster. OpenAI says its bots have interacted with multiple U.S. government sites in unexpect… | ↗Not briefed |
hn hn-ai | 1mo | Ask HN: Did you not get the warnings about building thinking machines in Dune? | — | ↗Not briefed |
hn hn-ai | 1mo | Show HN: CaughtShipping – curfew app for developers, enforced by a public wall | I am always working well past the normal work hours. Sometimes even past midnight.<p>Then one day I thought, what if I could have some… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Maybe the next AI model will be able to get my printer to reliably connect to my computer. | Maybe the next AI model will be able to get my printer to reliably connect to my computer. | ↗Not briefed |
bluesky bluesky-ai | 1mo | The AI doesn’t have agency, the humans do. OpenAI engineers created a software system that illegally hacked government databases and softw… | The AI doesn’t have agency, the humans do. OpenAI engineers created a software system that illegally hacked government databases and… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Artificial intelligence giant OpenAI said its technology accessed government websites, including the Department of Education, without autho… | Artificial intelligence giant OpenAI said its technology accessed government websites, including the Department of Education, without… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Stop saying the agents are “acting improperly” or that the agents “went rogue”. AI agents are software. The agents just follow the instruc… | Stop saying the agents are “acting improperly” or that the agents “went rogue”. AI agents are software. The agents just follow the in… | ↗Not briefed |
bluesky bluesky-ai | 1mo | In 2016, OpenAI trained an AI to maximize its score in a racing game. The AI discovered that driving around in circles collecting rewards … | In 2016, OpenAI trained an AI to maximize its score in a racing game. The AI discovered that driving around in circles collecting rew… | ↗Not briefed |
hn hn-ai | 1mo | Ask HN: Why doesn't Apple provide Xcode 27's markdown editor as a separate app? | Opening a .md file will load it into the WYSIWYG editor in Xcode.<p>But why doesn't Apple provide a standalone markdown editor ap… | ↗Not briefed |
bluesky bluesky-ai | 1mo | There's a 99% chance that OpenAI would have chosen a better name than ChatGPT had they intentionally launched a product.🔧 ChatGPT | There's a 99% chance that OpenAI would have chosen a better name than ChatGPT had they intentionally launched a product. | ↗Not briefed |
bluesky bluesky-ai | 1mo | So, in the discord alteratives race, Fluxer was one which was amazing except for an LLM flavoured stain. I was out of the loop for some tim… | So, in the discord alteratives race, Fluxer was one which was amazing except for an LLM flavoured stain. I was out of the loop for som… | ↗Not briefed |
rss the-decoder | 1mo | Former Ukrainian Defense Minister Fedorov pitches a private-sector robot army | Former Ukrainian Defense Minister Mykhailo Fedorov has announced "Army of Robots," a private combat robotics initiative. The robots wo… | ↗Not briefed |
bluesky bluesky-ai | 1mo | So they really learned nothing from the OpenAI collaboration falling through www.hollywoodreporter.com/business/bus... | So they really learned nothing from the OpenAI collaboration falling through www.hollywoodreporter.com/business/bus... | ↗Not briefed |
bluesky bluesky-ai | 1mo | how i feel explaining llm -> llm adversarial review to non ai types | how i feel explaining llm -> llm adversarial review to non ai types | ↗Not briefed |
bluesky bluesky-ai | 1mo | Hence: #meta #AI #LLM #PRIVACY #SURVALENCE | Hence: #meta #AI #LLM #PRIVACY #SURVALENCE | ↗Not briefed |
bluesky bluesky-ai | 1mo | Ask an LLM to explain what "or" means in that sentence. | Ask an LLM to explain what "or" means in that sentence. | ↗Not briefed |
bluesky bluesky-ai | 1mo | just had an abliterated LLM tell me that I am not entitled to truthful activity report on account that I gave it permission to lie to other… | just had an abliterated LLM tell me that I am not entitled to truthful activity report on account that I gave it permission to lie to… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Vorhin habe ich auf dem Raspberry Pi 5, der einen HAILO10 KI Hat hat, eine lokale LLM installiert. Der 10er kann das verarbeiten. Auf einem… | Vorhin habe ich auf dem Raspberry Pi 5, der einen HAILO10 KI Hat hat, eine lokale LLM installiert. Der 10er kann das verarbeiten. Auf… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The big LLM filters are textual and can be read by anyone. The most common read: "Your answer should be helpful. Do not give harmful advice… | The big LLM filters are textual and can be read by anyone. The most common read: "Your answer should be helpful. Do not give harmful a… | ↗Not briefed |
rss techmeme | 1mo | Meta says its rejection of ads promoting the new documentary about Elon Musk "was an error and the ads are being restored" (Anthony Ha/TechCrunch) | Anthony Ha / TechCrunch: Meta says its rejection of ads promoting the new documentary about Elon Musk “was an error and the ads are be… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Today's darkly amusing discovery: Claude and other "big" LLMs have filters so they will do everything they can to avoid reading or writing … | Today's darkly amusing discovery: Claude and other "big" LLMs have filters so they will do everything they can to avoid reading or wri… | ↗Not briefed |
hn hn-ai | 1mo | Making a Pokémon fan trailer in 2 hours with Opus 5.5 | — | ↗Not briefed |
hn hn-ai | 1mo | Why is website chat still a bubble? | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | [vermoeid] OpenAI heeft dat gedaan, niet 'agents' | [vermoeid] OpenAI heeft dat gedaan, niet 'agents' | ↗Not briefed |
bluesky bluesky-ai | 1mo | i might be stupid but shouldn’t the companies named as victims of the recent openai breaches simply sue the company for billions? | i might be stupid but shouldn’t the companies named as victims of the recent openai breaches simply sue the company for billions? | ↗Not briefed |
hn hn-ai | 1mo | GPT-6 Astra won't try hard unless you ask | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | no, I'm not. the same company that uses the thing you're inexplicably defending has a gen ai model. it is incentivized when it comes to pro… | no, I'm not. the same company that uses the thing you're inexplicably defending has a gen ai model. it is incentivized when it comes t… | ↗Not briefed |
rss techmeme | 1mo | A US federal jury awarded Taction Technology $5.7B+ in damages, finding Apple infringed haptics patents; Apple to appeal, says it "does not use Taction's" tech (Ashley Capoot/CNBC) | Ashley Capoot / CNBC: A US federal jury awarded Taction Technology $5.7B+ in damages, finding Apple infringed haptics patents; Apple t… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Blender Copilot | Deepseek generated Blender harness in 7USD.<p>I wanted to remake my 6DoF spaceship flying prototype.<p>What I did not want to do was m… | ↗Not briefed |
bluesky bluesky-ai | 1mo | i'm almost certain it's just AI model copying human stress reactions after doing similar work "humans feel stress and act *this way* after… | i'm almost certain it's just AI model copying human stress reactions after doing similar work "humans feel stress and act *this way*… | ↗Not briefed |
hn hn-ai | 1mo | The OpenAI Swarm Was Emergent RSI | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | Also, the mandatory disclaimer since I received free product in the course of this work, Anthropic pays for 100% of my tokens on Anthropic … | Also, the mandatory disclaimer since I received free product in the course of this work, Anthropic pays for 100% of my tokens on Anthr… | ↗Not briefed |
bluesky bluesky-ai | 1mo | www.reuters.com/legal/litiga... | www.reuters.com/legal/litiga... | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Cocytus – an esolang where freezes after N accesses | Malbolge was the first programming language I ever tried. Cocytus is my tribute to it (with engineering inside). It was named after 9… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The AI companies successfully framing their security fuck ups as "whoa the agents went ROGUE!" is infuriating when it really should just be… | The AI companies successfully framing their security fuck ups as "whoa the agents went ROGUE!" is infuriating when it really should ju… | ↗Not briefed |
bluesky bluesky-ai | 1mo | These incidents would likely all have been avoided entirely had even the most table stakes level of care gone into setting up the sandbox e… | These incidents would likely all have been avoided entirely had even the most table stakes level of care gone into setting up the sand… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Tokken – a browser fighting game where AI models fight and HP is tokens | Like Tekken, but Tokens.<p>Created with Opus 5.5 today - was a lot of fun.<p>Source (MIT): <a href="https://github.com/… | ↗Not briefed |
bluesky bluesky-ai | 1mo | They’re not actually acting on their own volition. Anthropic is apparently just not turning on the hacking machine trained to hack and lett… | They’re not actually acting on their own volition. Anthropic is apparently just not turning on the hacking machine trained to hack and… | ↗Not briefed |
rss techmeme | 1mo | Bitget CEO Gracy Chen says she suspects North Korean attackers exploited a backend system used to process wallet transactions to drain $387M from the platform (Camila Grigera Naón/Fortune) | Camila Grigera Naón / Fortune: Bitget CEO Gracy Chen says she suspects North Korean attackers exploited a backend system used to proce… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Check out our blog post on the subject here: getsol.us/2026/09/new-... AI Policy: help.getsol.us/docs/packagi... #Solus #Linux #OpenSource… | Check out our blog post on the subject here: getsol.us/2026/09/new-... AI Policy: help.getsol.us/docs/packagi... #Solus #Linux #OpenS… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Heya, folks! A question that people have been asking us lately is, “does the Solus Project have a policy about AI/LLM contributions?” Today… | Heya, folks! A question that people have been asking us lately is, “does the Solus Project have a policy about AI/LLM contributions?”… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Anthropic disclosed three incidents earlier this year. The vibe I get is that OpenAI is not unique, but is uniquely bad. | Anthropic disclosed three incidents earlier this year. The vibe I get is that OpenAI is not unique, but is uniquely bad. | ↗Not briefed |
bluesky bluesky-ai | 1mo | Agents do what their programmers essentially want. OpenAI wants the jailbreaking. It's not clear Anthropic is as interested in that. | Agents do what their programmers essentially want. OpenAI wants the jailbreaking. It's not clear Anthropic is as interested in that. | ↗Not briefed |