AI
Artificial intelligence: LLMs, agents, diffusion, vision, NLP and the latest from top labs.
593 published articles
AI liability
OpenAI's review system flagged the shooter. Its leaders allegedly stayed silent
Thirty new lawsuits from the students, teachers, and principal at the Tumbler Ridge shooting accuse OpenAI and Sam Altman of helping the suspect after its review system flagged his ChatGPT conversations. The suits allege the safety team wanted to contact Canadian authorities; leadership stayed quiet instead.
2026-09-14
China's enterprise AI agent land grab
Alibaba's QwenWork claims 30 million users in a month, with no outside audit
Alibaba says QwenWork passed 30 million users a month after entering beta, with enterprise accounts over half the base. The figures come from Alibaba's own newsroom, and the customer case studies behind them ship without methodology.
2026-09-14
AI Safety
OpenAI's Astra may hide its 'thinking.' Safety researchers are alarmed
The Information reports Astra uses an opaque, looped architecture that hides its reasoning. A top Redwood Research scientist calls it the worst development for AI safety to date, and OpenAI has not denied a single detail.
2026-09-14
Design Tools
Open Design 0.18.0 ends the 'is this the latest version?' meeting
Open Design 0.18.0 ships a shared Team workspace with live read-only mirrors and lets Codex call the design engine headlessly. 115 pull requests from 22 contributors landed in two days. Here's what changed and the agent-tooling trend it fits.
2026-08-21
EU AI Act: marking AI-generated content
Claude's invisible watermark is coming. A full rewrite erases it
Anthropic is adding an invisible watermark to future Claude models to meet EU AI Act rules on marking AI-generated content. It is free and untraceable, but it passes over exact code, short passages, and fully rewritten text.
2026-08-21
AI Music Generation: Qwen's answer to prompt drift
Music 3.0 swaps the one-tag prompt for a timeline that keeps AI songs on track
AI tracks tend to drift from the prompt as they unfold: instruments drop out, emotion flattens, the vocal style comes and goes. Music 3.0 swaps the one-tag description for a time-sequential Structured Caption, backed by an 8B/0.6B Hybrid-LM that splits structure from detail.
2026-08-21
Open-weight AI
Xiaohongshu's dots3-note: a 280B open MoE that only activates 16B
Xiaohongshu's dots studio has released dots3-note preview, its first open-weight model: a multimodal MoE with 280B parameters, 16B active, and a 512K context window. The sparse design targets low serving cost on one 8-GPU node, but the card has not published benchmark numbers yet.
2026-08-20
AI-Generated C++ Code Quality in Production
AI-written C++ runs 5-8% hotter and reviews longer. The fix is in the prompts
A 12-month study of 3.52 million code changes to an enterprise C++ codebase finds AI code is measurably worse on coupling, allocations, and loop style, costing 5-8% more compute. Prompt-level feedback recovered 11.1% of static analysis warnings.
2026-08-19
Agent Memory
TEPA: for AI agents, a stale memory is worse than none
Agent memory has a falsifiability problem, a new arXiv preprint argues: stale facts stay retrievable and pollute the prompt. Its TEPA mechanism revokes superseded memories, and in drift tests naive memory scored below no memory (0.210 vs 0.309) while TEPA reached 0.950.
2026-08-19
Test-Time Scaling
CoBa routing matches best-of-16 voting with 58.9% fewer tokens
CoBa, a compute-balanced routing policy from a new arXiv paper, matches best-of-16 majority voting within 0.01 points while cutting parameter-weighted tokens by 58.9%. On 3,129 evaluations across MATH-500, AIME, and AMC, it also beats single-sample decoding outright, though a small best-of-16 edge remains when budget is no object.
2026-08-19
AI Research
Muon groks modular addition faster, then its solutions collapse
Muon-trained transformers grok modular addition faster than AdamW, then lose the solution in every configuration tested. The paper pins the collapse to the representation-readout interface, where freezing either parameter group prevents it. Fourier analysis shows the task circuit survives and simply gets outvoted.
2026-08-19
Fisher-R1 | Hypothesis testing
AI agents run flawless statistics and still draw the wrong conclusions
Agents that automate hypothesis testing can execute analyses flawlessly yet reach wrong conclusions, because most benchmarks never check whether the p-value is valid. A 425-task benchmark quantifies the gap, and an RL-trained open-weight model closes a chunk of it.
2026-08-19