Daily AI news, sourcedUpdates / All
Updates 1959 items · newest issue first · page 19 of 98
AreaAllModelsAgents & codingApps & productsChips & infraResearchPricing & limitsBusiness & fundingPolicy & safetyOther
- Builder.io open-sources Agent-Native, a TypeScript framework where agents and UIs share the same actions
- OpenCode 2.0 moves to a shared background service with Code Mode, background subagents and an SDK
- Microsoft's October security update credits researchers 'using GLM from z.ai' on 23 vulnerabilitiesOfficialMicrosoft Security Response CenterOct 15Policy & safety
- Apple prepares to debut its first Apple Intelligence features in China in December with iOS 27.3, per Bloomberg
- YC-backed Instinct opens early access to a team of agents that monitors markets 24/7 for traders
- ByteDance Seed sets up an Alignment & Preparedness department focused on AI safety and is hiring researchers
- Stanford's Hazy Research releases NCM, a non-autoregressive language model it says decodes up to 31x faster at matching quality
- OpenAI alleges in court that a former Apple employee who joined it took files about an unreleased device, per ReutersReportedCNA (Reuters)Oct 15Policy & safety
- Claude Code desktop adds saved environments, nested sessions and a /btw side chatOfficialX @ClaudeDevsOct 15Agents & coding
- Nvidia's share of China's AI chip market fell to 34% in the first half as Huawei rose to 28%, IDC data shows, per ReutersReportedReuters (via U.S. News)Oct 15Chips & infra
- Paper Instruments releases Ultramarine 1 Flash, an RL post-train of GLM-5.3 Flash for knowledge work, via API with weights promisedOfficialPaper InstrumentsOct 15Models
- Researcher traces where LLM "slop" writing style lives in Qwen3-4B-Instruct and which training stage installs itUnconfirmedr/MachineLearningOct 15Research
- Anthropic tells prospective investors third-quarter revenue reached $16.8 billion and was profitable on an adjusted basis, per Reuters
- Developer's PulsarLM inference engine drops old, low-attention tokens to keep long chat and agent contexts boundedUnconfirmedr/LocalLLaMAOct 15Chips & infra
- LocalLLaMA user builds three four-GPU Radeon AI Pro R9700 machines on $127 used EPYC CPUs for about $10,700 each and 384GB of VRAMUnconfirmedr/LocalLLaMAOct 15Chips & infra
- LocalLLaMA user benchmarks AMD's MI350P against Nvidia's RTX Pro 6000 Blackwell after tuning AMD kernels to double its out-of-box speedUnconfirmedr/LocalLLaMAOct 15Chips & infra
- OpenAI says Block's security team uses Daybreak to triage every alert and propose patches, turning weeks of work into hours
- Fish Audio releases Drama 3, a 35B open-weight mixture-of-experts text-to-speech model directed in plain language across 83 languagesOfficialFish AudioOct 15Models
- Datalab releases OmniParseBench, an open document-parsing benchmark of 16,288 pass/fail tests across 92 languagesStatementDatalabOct 15Models
- McKinsey will have some entry-level candidates spend 10 minutes with an AI-simulated client during interviews, per The Wall Street JournalReportedThe VergeOct 15Business & funding