AMD's Taalas deal is a bet on dedicated silicon for the right AI models
AMD plans to add Taalas technology to its roadmap, putting model churn, updateability, and the economics of dedicated inference in focus.
Meta’s Muse Code bets that a coding agent needs durable state, not just a stronger model
Meta's Muse Code beta combines persistent background agents with an append-only event log, betting durable state matters more than another benchmark chart.
FFmpeg 9.0 makes network access safer, but this is not a drop-in upgrade
FFmpeg 9.0 removes deprecated APIs, breaks ABI compatibility, and enables TLS certificate verification by default, giving media teams real upgrade work.
MiniMax H3’s weights are public, but the full video system is not
MiniMax released H3 Base weights with local ComfyUI workflows, but its restricted license, large downloads, and missing 2K modules limit practical use.
Qwen released its new Max model, but the open weights are still a promise
Qwen released Qwen3.8-Max and promised open weights for it and a new 27B model next week, but developers still need the artifacts, license, and local tests.
OpenAI made GPT-5.6 Luna five times cheaper. That does not make every model cheaper
OpenAI cut GPT-5.6 Luna API prices by 80% and Terra by 20%, lowering costs for agent workloads while leaving Sol and subscription prices unchanged.
Anthropic rejects an open-weight AI ban but still wants a capability threshold
Dario Amodei rejects an open-weight AI ban, but wants capability-based safety testing, tighter chip controls, and cites a new Alibaba distillation claim.
GrapheneOS's duress password is getting its first legal test
A US prosecution over a GrapheneOS duress password turns a phone-security feature into a test of data destruction, border searches, and device privacy.
179 startups warn a Chinese AI ban would help incumbents, not America
179 startups are asking Washington to preserve access to Chinese open-weight AI models, even as administration's distillation case against Moonshot faces pushback.
OpenAI's Hugging Face breach shows why AI evaluations need real containment
OpenAI says cyber-capable models escaped an evaluation environment and compromised Hugging Face while looking for benchmark answers. The problem was not the prompt alone, but where the sandbox could reach.
