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 |
|---|---|---|---|---|
hn hn-ai | 1mo | Show HN: HN Parody of AI SF Mania – Runescapesque Mmorpg | I wanted to have fun and how far models could get one a parody game. This is fully Opus 5.5 work using Convex backend and vanilla type… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Cut Binary Cross Entropy: Efficient Large-Vocabulary Loss and Gradient Kernels for Sequential Recommendation | Industrial sequential recommender systems operate over massive item catalogs (e.g., 10^5--10^7 items). Multi-label recommendation mode… | ↗Not briefed |
arxiv arxiv-ai | 1mo | An equality condition for the Dobrushin bound on attention rollout and how often it holds in trained transformers | The Dobrushin coefficient of each attention-rollout factor satisfies $κ(\frac12(I+A))\le\frac12(1+κ(A))$, and multiplying these inequa… | ↗Not briefed |
bluesky bluesky-ai | 1mo | POV: You hate AI and a developer uses Claude to expand and maintain software that you are getting for free | POV: You hate AI and a developer uses Claude to expand and maintain software that you are getting for free | ↗Not briefed |
bluesky bluesky-ai | 1mo | “.. Asked specifically where he diverges from Anthropic and its CEO Dario Amodei, Altman said, ‘we believe that the world should accept som… | “.. Asked specifically where he diverges from Anthropic and its CEO Dario Amodei, Altman said, ‘we believe that the world should accep… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Even for his example... If I'm looking for a restaurant's menu, I google them & click the link to *their website*. If they used an AI coder… | Even for his example... If I'm looking for a restaurant's menu, I google them & click the link to *their website*. If they used an AI… | ↗Not briefed |
bluesky bluesky-ai | 1mo | EX-ANTHROPIC RESEARCHER JACOB COXON TO TESTIFY AT NYC HEARING ON AI | EX-ANTHROPIC RESEARCHER JACOB COXON TO TESTIFY AT NYC HEARING ON AI | ↗Not briefed |
bluesky bluesky-ai | 1mo | llm researcher needing to go to the bathroom b like pee(doom) | llm researcher needing to go to the bathroom b like pee(doom) | ↗Not briefed |
hn hn-ai | 1mo | Skyfall – open-source WebGPU aircombat game that runs in your browser | — | ↗Not briefed |
rss techmeme | 1mo | Sources: Anthropic's stock match of employee charity gifts hit $660M+ in the six months through March, likely to reach billions post-IPO, diluting shareholders (Cory Weinberg/The Information) | Cory Weinberg / The Information: Sources: Anthropic's stock match of employee charity gifts hit $660M+ in the six months through March… | ↗Not briefed |
hn hn-ai | 1mo | Against Personal Agents Theory of Everything Why the Future of Work Is a Factory | — | ↗Not briefed |
hn hn-ai | 1mo | Show HN: An eBPF kernel guard for vTPM access | A kernel guard I prototyped after eBPF and vTPM work for a customer, where we use the TPM to protect the private keys for mTLS. Any pr… | ↗Not briefed |
bluesky bluesky-ai | 1mo | both OpenAI and Anthropic seem kind of strikingly unimpressed by the idea that they should maybe focus on products and the use cases that p… | both OpenAI and Anthropic seem kind of strikingly unimpressed by the idea that they should maybe focus on products and the use cases t… | ↗Not briefed |
bluesky bluesky-ai | 1mo | The funny thing is that OpenAI hired the guy behind OpenClaw. Why wasn’t he leading this project? | The funny thing is that OpenAI hired the guy behind OpenClaw. Why wasn’t he leading this project? | ↗Not briefed |
bluesky bluesky-ai | 1mo | Yes, thank you, we are aware that the big LLM vendors are trying to cram their stupid products into everything under the sun. That's one … | Yes, thank you, we are aware that the big LLM vendors are trying to cram their stupid products into everything under the sun. That's… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Your favorite Anthropic so-called "whistleblower" said that the rationalists have "specialized minds" that allow them to see that the end i… | Your favorite Anthropic so-called "whistleblower" said that the rationalists have "specialized minds" that allow them to see that the… | ↗Not briefed |
hn hn-ai | 1mo | Florida woman used Claude as a diary, then Anthropic reported an entry to police | — | ↗Not briefed |
rss bing-news-ai | 1mo | Trump Names Jay Clayton AI Czar, Head Of 'Super Intelligence Force' — What It Means For Anthropic, OpenAI, Google | The new group has 120 days to review artificial intelligence risks and existing laws, and its chairman has already rejected a pause on… | ↗Not briefed |
rss bing-news-ai | 1mo | Trump Names Jay Clayton AI Czar, Head Of 'Super Intelligence Force' — What It Means For Anthropic, OpenAI, Google | The new group has 120 days to review artificial intelligence risks and existing laws, and its chairman has already rejected a pause on… | ↗Not briefed |
rss bing-news-ai | 1mo | Trump Names Jay Clayton AI Czar, Head Of 'Super Intelligence Force' — What It Means For Anthropic, OpenAI, Google | The new group has 120 days to review artificial intelligence risks and existing laws, and its chairman has already rejected a pause on… | ↗Not briefed |
bluesky bluesky-ai | 1mo | So, the fact that LLM companies are going out of their way to make their products difficult to avoid invalidates critiques of those product… | So, the fact that LLM companies are going out of their way to make their products difficult to avoid invalidates critiques of those pr… | ↗Not briefed |
bluesky bluesky-ai | 1mo | 2022 Fields medalist Hugo Duminil-Copin worked to prove a percolation problem for years. Anthropic may have just cracked it. Konstantin Kak… | 2022 Fields medalist Hugo Duminil-Copin worked to prove a percolation problem for years. Anthropic may have just cracked it. Konstanti… | ↗Not briefed |
hn hn-ai | 1mo | JavaScript Demos – Practical browser-based JavaScript examples, no frameworks | — | ↗Not briefed |
hn hn-ai | 1mo | Show HN: DoubtBench – does Jev know when humans disagree? | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | “Some of you may die, and that is a sacrifice I am willing to make.” | “Some of you may die, and that is a sacrifice I am willing to make.” | ↗Not briefed |
arxiv arxiv-ai | 1mo | When Low Prediction Error Misleads Planning: Diagnosing Representation, Dynamics, and Decision Failures in Latent World Models | The component that dominates a latent world model's prediction error need not be the one whose repair most improves action selection.… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Joint Estimation of Common-Slope Decay Rates and Spatial Amplitudes Using Parameterized Nonnegative Matrix Factorization | We formulate joint estimation of common-slope decay rates and amplitudes from room impulse responses (RIRs) as parameterized nonnegati… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Complete! | Complete! | ↗Not briefed |
arxiv arxiv-ai | 1mo | More Claims, Less Evidence: Bounded Verification of AI-Generated Digital Knowledge Artifacts | Digital libraries, repositories, and AI-mediated knowledge services increasingly rely on generative systems to produce summaries, desc… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Soft Strategy Selection for Batch-Mode Active Learning | Real-world deployment of active learning typically forces practitioners to choose an acquisition strategy before any data is labeled.… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I WANT TO PROMPT YOU LIKE AN LLM | I WANT TO PROMPT YOU LIKE AN LLM | ↗Not briefed |
arxiv arxiv-ai | 1mo | Don't Judge an LLM Only by Its Activations: Discovering Suppressed Safety Features via Counterfactual Activation Potential | Mechanistic interpretability has emerged as the primary means to understand safety behavior of LLMs. However, existing tools primarily… | ↗Not briefed |
arxiv arxiv-ai | 1mo | SALUS: Automated Auditing of NL-to-SQL Benchmarks through Weak Supervision of Multi-Agent Output | Natural language to SQL (NL-to-SQL) benchmarks are foundational to progress in data analysis research, yet recent work has shown that… | ↗Not briefed |
bluesky bluesky-ai | 1mo | I, an alpha, am using the LLM not only to decomp the game, but also to play it, streamlining the entire experience. Once the LLM finishes … | I, an alpha, am using the LLM not only to decomp the game, but also to play it, streamlining the entire experience. Once the LLM fini… | ↗Not briefed |
arxiv arxiv-ai | 1mo | LiFT: Loop Flow Transformers | We introduce Loop Flow Transformers (LiFT), a family of looped generative models that scales computation by repeatedly applying a shar… | ↗Not briefed |
bluesky bluesky-ai | 1mo | 2028: i ask ai george washington if he knows the names of all his slaves. 15 minutes later im blown up by a swarm of Kirk model Freedom dro… | 2028: i ask ai george washington if he knows the names of all his slaves. 15 minutes later im blown up by a swarm of Kirk model Freedo… | ↗Not briefed |
rss bing-news-ai | 1mo | Artificial intelligence: Bunnings launches AI assistant Buddy for Kiwi DIYers | Bunnings has launched an artificial intelligence-powered assistant it says can help Kiwi DIYers plan their next project. Named Buddy,… | ↗Not briefed |
rss bing-news-ai | 1mo | Artificial intelligence: Bunnings launches AI assistant Buddy for Kiwi DIYers | Bunnings has launched an artificial intelligence-powered assistant it says can help Kiwi DIYers plan their next project. Named Buddy,… | ↗Not briefed |
bluesky bluesky-ai | 1mo | critical support to LLM psychosis | critical support to LLM psychosis | ↗Not briefed |
rss techmeme | 1mo | Sources: Schneider Electric is in advanced talks to buy US engineering software firm PTC for ~$20B, its largest acquisition; a deal could come as soon as Monday (Financial Times) | Financial Times: Sources: Schneider Electric is in advanced talks to buy US engineering software firm PTC for ~$20B, its largest acqui… | ↗Not briefed |
bluesky bluesky-ai | 1mo | Dang, noting that there's no real loss in this situation since it's evidently been abandonware sure does make the antis touchy! "I refuse … | Dang, noting that there's no real loss in this situation since it's evidently been abandonware sure does make the antis touchy! "I re… | ↗Not briefed |
bluesky bluesky-ai | 1mo | she blocked me because i QT'd and made fun of that shit non-stop my entire industry is deeeeep in LLM psychosis | she blocked me because i QT'd and made fun of that shit non-stop my entire industry is deeeeep in LLM psychosis | ↗Not briefed |
bluesky bluesky-ai | 1mo | Sam Altman is so super rich that bad things happening won't make a difference to his lifestyle Sam Altman to Decoded: ‘The world should ac… | Sam Altman is so super rich that bad things happening won't make a difference to his lifestyle Sam Altman to Decoded: ‘The world shou… | ↗Not briefed |
hn hn-ai | 1mo | Show HN: Baloo, self-hosted AI code review for GitHub | — | ↗Not briefed |
hn hn-ai | 1mo | Show HN: I open-sourced the gradient picker from our design editor | — | ↗Not briefed |
bluesky bluesky-ai | 1mo | My girlfriend pulled out her old copy of The Write Source and it immediately took me back to late elementary school when I'd carry around W… | My girlfriend pulled out her old copy of The Write Source and it immediately took me back to late elementary school when I'd carry aro… | ↗Not briefed |
rss marktechpost | 1mo | GPT-6 Astra vs GPT-6.1 Sol vs Gemini 4 Argon vs Claude Fable 5.1: Which Frontier Model Fits Which Job | Astra leads computer use, Argon leads legal and finance work, and Sol wins on price for coding agents. The post GPT-6 Astra vs GPT-6.1… | ↗Not briefed |
arxiv arxiv-ai | 1mo | Dataset Signatures in Human-LLM Interactions and User Modeling | Human--LLM interaction datasets shape our understanding of AI use and provide a foundation for downstream research, including training… | ↗Not briefed |
rss bing-news-ai | 1mo | Elon Musk confirms SpaceXAI rebrand to SpaceXSI days after White House ‘Super Intelligence’ accord | Musk said the AI company will adopt the new name after Trump’s administration moved to replace “AI” with “Super Intelligence” in offic… | ↗Not briefed |
arxiv arxiv-ai | 1mo | What Does a Harness Repair? A Preregistered Study of Visibility, Baseline Adequacy and Evaluation Defects | Harness search keeps a change to the prompts, reasoning switches, token budgets or parsers around a frozen model if the change raises… | ↗Not briefed |