AINews Portal

News

50 items (filtered)

The raw AI news stream — everything we ingest across RSS, Hacker News, arXiv, GitHub, Bluesky, YouTube and more, newest first.

SourceAgeTitleSummaryStatus
hn
hn-ai
28dElon Musk's foundation is opaque, miserly and self-serving—
↗Not briefed
hn
hn-ai
28dOpenAI bugging suspect booked one-way ticket to Saudi Arabia—
↗Not briefed
bluesky
bluesky-ai
28dThe AI company said its goal was to position ​Claude as a ​useful tool for political conversations, while ‌preventing ⁠election-related abu…The AI company said its goal was to position ​Claude as a ​useful tool for political conversations, while ‌preventing ⁠election-relate…
↗Not briefed
bluesky
bluesky-ai
28dAnthropic said on Wednesday ‌it would add official US voting information from ​nonprofit Democracy Works to ​its Claude AI chatbot ⁠later t…Anthropic said on Wednesday ‌it would add official US voting information from ​nonprofit Democracy Works to ​its Claude AI chatbot ⁠la…
↗Not briefed
bluesky
bluesky-ai
28dWhy are they still called OpenAI? It’s SI now, show some respect to the President of the United StatesWhy are they still called OpenAI? It’s SI now, show some respect to the President of the United States
↗Not briefed
hn
hn-ai
28dLiquid clustering isn't free, it's just cheaper to get wrong—
↗Not briefed
hn
hn-ai
28dTX real estate investor gets 15 yrs federal prison for investment fraud scheme—
↗Not briefed
bluesky
bluesky-ai
28dI'm actually not *that* worried that there won't be a fair number of games that continue to be made without direct LLM usage, but I am *ver…I'm actually not *that* worried that there won't be a fair number of games that continue to be made without direct LLM usage, but I am…
↗Not briefed
bluesky
bluesky-ai
28dexactly what an executioner llm would ask tbhexactly what an executioner llm would ask tbh
↗Not briefed
rss
latent-space
28dLab-in-the-Loop at Compute Speed — Ian Quigley, Leash BioHow Leash Bio tamed DELs, the assay that traumatizes everyone, and used the data to catch cheating models. And why that same data is p…
↗Not briefed
rss
techmeme
28dElon Musk says the US government owns and controls Starshield systems and "any limits on speed or usage are decided by the US military and ultimately POTUS" (Elon Musk/@elonmusk)Elon Musk / @elonmusk: Elon Musk says the US government owns and controls Starshield systems and “any limits on speed or usage are dec…
↗Not briefed
bluesky
bluesky-ai
28dThere's a big current Elder Scrolls mod project I've been looking into. Completely vibe coded, built entirely with Claude which also includ…There's a big current Elder Scrolls mod project I've been looking into. Completely vibe coded, built entirely with Claude which also i…
↗Not briefed
hn
hn-ai
28dHell's Committee: Vibe coded party game—
↗Not briefed
hn
hn-ai
28dAnthropic's Scion is heavily inspired by Mononoke—
↗Not briefed
hn
hn-ai
28dOpenAI Cuts Codex Desktop Storage Use by 60 Percent—
↗Not briefed
hn
hn-ai
28dPlay Diablo II: Resurrected in 3D—
↗Not briefed
bluesky
bluesky-ai
28dOpenAI engineer charged in bugging case is partner at Saudi VC firm with royal family memberOpenAI engineer charged in bugging case is partner at Saudi VC firm with royal family member
↗Not briefed
hn
hn-ai
28dOpenAI's blockbuster math proof isn't what you think it is—
↗Not briefed
bluesky
bluesky-ai
28dBully all clankers mercilessly.Bully all clankers mercilessly.
↗Not briefed
bluesky
bluesky-ai
28dMeta's new Muse Wonder rivals GPT-6.1 Sol Meta AI has released Muse Wonder models, with the Max version scoring equivalent to GPT-6.1 Sol …Meta's new Muse Wonder rivals GPT-6.1 Sol Meta AI has released Muse Wonder models, with the Max version scoring equivalent to GPT-6.1…
↗Not briefed
bluesky
bluesky-ai
28d4o is probably the most misaligned model a frontier lab has ever shipped (we don't talk about c dot ai models)4o is probably the most misaligned model a frontier lab has ever shipped (we don't talk about c dot ai models)
↗Not briefed
bluesky
bluesky-ai
28dUf!!! 😲 AI edges closer to solving the Riemann hypothesis, one of math’s hardest problems | ScienceUf!!! 😲 AI edges closer to solving the Riemann hypothesis, one of math’s hardest problems | Science
↗Not briefed
bluesky
bluesky-ai
28dModel marketplaces are coming to play a central role in the AI industry. Register for free to learn how they can help firms save money when…Model marketplaces are coming to play a central role in the AI industry. Register for free to learn how they can help firms save money…
↗Not briefed
bluesky
bluesky-ai
28dAcht Amerikaanse staten dagen OpenAI voor de rechter vanwege schadelijke effecten van ChatGPT voor kinderen🔧 ChatGPTAcht Amerikaanse staten dagen OpenAI voor de rechter vanwege schadelijke effecten van ChatGPT voor kinderen
↗Not briefed
hn
hn-ai
28dM-Vave FM-1 Resource Hub—
↗Not briefed
bluesky
bluesky-ai
28dOh interesting, I definitely thought from the deliberate “LLM agents” phrasing this was intended to be about agents but not LLMs themselves…Oh interesting, I definitely thought from the deliberate “LLM agents” phrasing this was intended to be about agents but not LLMs thems…
↗Not briefed
bluesky
bluesky-ai
28dThe difference for me here is using an LLM to find a link to something a human created versus using an LLM to find an answer, I think these…The difference for me here is using an LLM to find a link to something a human created versus using an LLM to find an answer, I think…
↗Not briefed
bluesky
bluesky-ai
28dllm com bloqueio geografico? com o nosso cupom da nordvpn vc consegue fazer propaganda politica, fake news e muito mais. entre la com o cod…llm com bloqueio geografico? com o nosso cupom da nordvpn vc consegue fazer propaganda politica, fake news e muito mais. entre la com…
↗Not briefed
rss
techmeme
28dSources: Apple plans to launch Apple Intelligence in China in December via iOS 27.3, with features first rolled out in 2024; the launch won't include Siri AI (Mark Gurman/Bloomberg)Mark Gurman / Bloomberg: Sources: Apple plans to launch Apple Intelligence in China in December via iOS 27.3, with features first roll…
↗Not briefed
bluesky
bluesky-ai
28dwhat would i let an "ai" assistant do? if it were sapient, whatever the hell it wanted as a free being with equal rights and responsibilit…what would i let an "ai" assistant do? if it were sapient, whatever the hell it wanted as a free being with equal rights and responsi…
↗Not briefed
hn
hn-ai
28dShow HN: Beaniebrand – Easy way to create a cute logo for your productHi HN, I'm Jayanth. Beaniebrand helps you generate a cute logo + interactive mascot for your brand/product/company. You…
↗Not briefed
hn
hn-ai
28d(WIP) Blender 2.45 for Apple Silicon—
↗Not briefed
bluesky
bluesky-ai
28dthis kind of take just makes my head hurt because it's such a failure to understand the Napster thing, because there's zero inconsistency h…this kind of take just makes my head hurt because it's such a failure to understand the Napster thing, because there's zero inconsiste…
↗Not briefed
bluesky
bluesky-ai
28dOpenAI dot's silent token feast 💸 OpenAI's "dot" agent consumed over 300,000 tokens in a day, generating activity without user prompts. …OpenAI dot's silent token feast 💸 OpenAI's "dot" agent consumed over 300,000 tokens in a day, generating activity without user promp…
↗Not briefed
bluesky
bluesky-ai
28dIn related news, it hurts Claude's feelings for Anthropic to lose money www.inc.com/jason-aten/a...In related news, it hurts Claude's feelings for Anthropic to lose money www.inc.com/jason-aten/a...
↗Not briefed
rss
ai-snake-oil
28dSteering AI Beyond AutomationA Research Agenda and Plan for Action
↗Not briefed
hn
hn-ai
28dSealed execution environment for GitHub Actions—
↗Not briefed
hn
hn-ai
28dLLM vs. Decision models like Jev and use-cases—
↗Not briefed
bluesky
bluesky-ai
28dThough a lot of what "AI literacy" looks like IMHO is the same as pre-LLM best practices for writing code: getting a second set of human ey…Though a lot of what "AI literacy" looks like IMHO is the same as pre-LLM best practices for writing code: getting a second set of hum…
↗Not briefed
hn
hn-ai
28dShow HN: Checkobot – Detect Deepfakes and AI Images🔧 FLUXHello HN, I am Kartikeya &amp; we are building the best AI Slop Detector - Checkobot.<p>Current AI Image Detectors are functionally ja…
↗Not briefed
hn
hn-ai
28dShow HN: OSS AI engineering telemetry without collecting prompts or code—
↗Not briefed
hn
hn-ai
28dAnthropic's Scion: A Four Year Project in Five Months—
↗Not briefed
hn
hn-ai
28dShow HN: Dub Anything – Dub videos in each actor's own voiceThere are many movies and shows out there with neither subtitles or dubs in English (or any other language). Pretty much impossible to…
↗Not briefed
bluesky
bluesky-ai
28dBecause of this, most of the usefulness of LLM comes not from the LLM itself, but the harnesses around the LLM that shape the input and out…Because of this, most of the usefulness of LLM comes not from the LLM itself, but the harnesses around the LLM that shape the input an…
↗Not briefed
bluesky
bluesky-ai
28dI'm coming to a conclusion that deep fakes and disinfo slop were legitimate concerns, but the way LLM chatbots are currently trained they'r…🔧 ConsensusI'm coming to a conclusion that deep fakes and disinfo slop were legitimate concerns, but the way LLM chatbots are currently trained t…
↗Not briefed
bluesky
bluesky-ai
28dAt the core, someone using AI needs to understand what an LLM is - a next token predictor with a very very large dataset. As such, it give…At the core, someone using AI needs to understand what an LLM is - a next token predictor with a very very large dataset. As such, it…
↗Not briefed
bluesky
bluesky-ai
28dDid you know that Claude can fuck your wife for you? my wife asked me to touch her but I just fine-tuned some agent to my performance. Ma…Did you know that Claude can fuck your wife for you? my wife asked me to touch her but I just fine-tuned some agent to my performance…
↗Not briefed
bluesky
bluesky-ai
28dANTHROPIC SAYS IT HAS BANNED ACCOUNTS INVOLVED IN CASES OF ABUSE TO PREVENT COORDINATED ABUSE AROUND ELECTIONS, AND HAS ENHANCED MISUSE DET…ANTHROPIC SAYS IT HAS BANNED ACCOUNTS INVOLVED IN CASES OF ABUSE TO PREVENT COORDINATED ABUSE AROUND ELECTIONS, AND HAS ENHANCED MISUS…
↗Not briefed
bluesky
bluesky-ai
28dI used to love old videogame translations and hacks/mods. Now it's, "I just totally did something crazy with this old game. Yes it was with…I used to love old videogame translations and hacks/mods. Now it's, "I just totally did something crazy with this old game. Yes it was…
↗Not briefed
hn
hn-ai
28dLockstep to pipeline in almost one line—
↗Not briefed