#performance

Thinking Machines launched Inkling, a 975B open-weights multimodal model that immediately drew comparisons to closed frontier labs. xAI's Grok CLI sparked a data-privacy backlash after users reported it uploading entire home directories. Eli Lilly acquired psychedelics biotech AtaiBeckley for $2.8B, and CMS signaled a structural overhaul of how it pays for clinical AI.

OpenAI's GPT-5.6 family launched across three tiers (Luna, Terra, Sol) and immediately reshuffled the model landscape: Anthropic extended Claude Fable 5 access on paid plans in response, Perplexity swapped Claude for Grok 4.5 at half the cost, and one production team reported a 2.2x speed gain and 27% cost reduction after migrating an agent to GPT-5.6. Nathan Lambert argued open-weight models face a six-month window to prove viability against a closing frontier. In pharma, HHS pressed ahead with guidance to curb antidepressant prescribing, Medicaid funding cuts threatened family caregiver wages across multiple states, and daraxonrasib produced what oncologists called the biggest pancreatic cancer breakthrough in decades. Patrick Boyle examined Canada's economic warning signs and Russia's war finances.

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.

The US government issued an export control directive suspending all access to Claude Fable 5 and Mythos 5 for foreign nationals, a move that rattled the AI industry and prompted Tyler Cowen to note the equilibrium was hard to solve for. Separately, Arvind Narayanan and Sayash Kappoor published a detailed argument that software engineers won't be replaced by AI, citing task fragmentation and the compounding cost of errors. The FDA approved Sanofi's teplizumab for children with stage 3 type 1 diabetes after it cleared a dispute between career staff and political appointees. Noah Smith asked whether Chinese corporate debt could zombify the economy, while Patrick Boyle examined how SpaceX humiliated Wall Street by retaining founder control across 25x more fundraising than typical startups.

Anthropic launched Claude Fable 5 this week and immediately faced a backlash over two policies buried in its 319-page system card: a provision allowing the model to silently reduce its helpfulness for frontier AI research without telling users, and a 30-day data retention requirement for Fable and Mythos class models. Anthropic walked back the silent-nerfing policy within days, issuing an apology to Wired. Separately, a federal proposal would let OMB and 40 grantmaking agencies rewrite rules governing all federal research grants, raising concerns about the nationalization of American science. Medicare Advantage insurers were found to be blocking rehabilitation care and reversing denials only on appeal, per a federal OIG report. OpenAI filed to go public.

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.

Microsoft unveiled two proprietary MAI models at Build, including a 1-trillion-parameter reasoning model and a 137B coding model for GitHub Copilot, as Uber hit its AI budget ceiling after burning through its full 2026 allocation in four months. Alnylam struck a $2 billion deal with Inceptive Nucleics to build AI foundation models for RNA interference therapeutics. Trump signed a narrower AI oversight executive order requiring a 30-day review before powerful open-weight models can be released. The Trump administration also moved to strip job protections from roughly 8,000 NIH staff, including officials who oversee research grants, while Medicaid work requirements drew criticism for going beyond what advocates had anticipated.

The FDA's leadership vacuum deepened this week as the drug center head followed Commissioner Makary out the door, leaving the agency with no confirmed leadership at its two most critical posts. Open-weight models kept multiplying with Gemma 4, DeepSeek V4, and Kimi K2.6 all dropping in the same month, while Sebastian Raschka's analysis of new KV-sharing and compressed attention techniques explains why long-context costs are falling fast. A rare Bundibugyo Ebola outbreak in Congo's Ituri province prompted a WHO public health emergency declaration, and the WHO's move landed as hantavirus conspiracy theories were already cycling through social media. On the economy side, Trump's China visit produced mixed signals on decoupling, while Noah Smith argued that Europe's comfortable stagnation is no longer viable.

Bun rewrites itself in Rust, Codex spreads across surfaces, biotech readouts in tau and Duchenne, and FDA leadership churn.