#gpu

OpenAI launched GPT-Live, a full-duplex voice model backed by GPT-5.5 reasoning, while xAI's Grok 4.5 topped SaaS automation benchmarks at one-quarter the cost of rivals. OpenAI also retracted its recommendation of SWE-Bench Pro after auditing found 30% of its tasks broken. Prime Intellect raised $130M to bring frontier reinforcement learning infrastructure to any company. In pharma, AstraZeneca and Ionis reported a major trial failure in a competitive heart disease market, the White House reviewed FDA commissioner finalists, and RFK Jr. moved to create a formal list of Covid vaccine injuries. The NY Fed found tariff cost pass-through to consumers is still in the pipeline.

Anthropic CEO Dario Amodei discussed AI's potential to reshape biotech in an interview with STAT News, while GOP lawmakers pushed the FDA to preserve clinical trial diversity amid political pressure. Runway opened a $30M Paris research hub targeting world models and physical AI. Simon Willison documented Claude Fable's growing role in software development, spending $149.25 to write most of sqlite-utils 4.0rc2. Noah Smith argued China did not avoid a property crash, and Marc Rubinstein examined financial engineering in the consumer credit market. Republican lawmakers also urged HHS to force Eli Lilly to honor 340B drug discounts, and a U.S. judge blocked Colorado from capping Amgen's Enbrel price.

DeepSeek V4 Pro beat GPT-5.5 Pro on precision benchmarks this week, marking the latest round in the US-China model rivalry. A malware campaign hit 32 npm packages with roughly 117,000 weekly downloads, planting backdoors inside Claude Code startup settings. Pfizer's monthly obesity drug berobenatide continued to show promise in mid-stage data while Boehringer Ingelheim's survodutide disappointed on overall weight loss. Patrick Boyle examined what happens when a housing boom turns to bust, and Marc Rubinstein at Net Interest looked at the IPO readiness of SpaceX, Anthropic, and OpenAI.