llama.cpp PR adds chunked GDN kernel for CUDA and HIP

Leash Bio's Minos turns years of DNA-encoded library data into tens of thousands of lab-in-the-loop simulations for agent scientific RL. The release includes a full preprint, blog, and open dataset.

Reddit user komarco created "Liquid Geometry," a 3-minute immersive show for a "Museum Night" exhibition, generating keyframes first and then using Minimax H3 for video generation.

Chainalysis reports cumulative x402 payments on Base jumped from ~1,500 in mid-2025 to over 138 million by Q3 2026. More than 87% of payments were between a hundredth of a cent and 10 cents; US-linked wallets were 88.1% of identifiable AI inference spending.

Anthropic analyzed 250,000 Claude conversations over one week: politics, policy and social issues make up 5.75% of US Claude.ai chats, elections and voting 0.51%, and requests for voting recommendations just 0.01%. Policy-violating election requests appeared in 0.002% of conversations.

Barclays downgraded Dell, HPE, and NetApp over concerns that enterprise AI adoption is lagging, triggering a data center sell-off. Dell shares fell 11%.

Databricks says its enterprise customers receive millions of documents, and Document Intelligence is aimed at powering real-time applications and agents on that data.

Hugging Face compares Apple's shipping M5 Ultra Mac Studio against Nvidia's RTX 5090 for local model inference, using Qwen3.8-27B prompt processing and token generation benchmarks. The video also covers multi-token prediction speedups.