Podcast episode covering Sonnet 4.6 (1M context, strong ARC-AGI-2), Claude Code Remote Control mobile release, and Anthropic CEO's stance on Pentagon contract tensions. Also covers Gemini 3.1 Pro, Grok 4.2 multi-agent beta, Perplexity's multi-agent "Computer" coordinator, and Meta's $100B AMD chip deal. Research segments include deep-thinking tokens as a reasoning-effort signal and Anthropic's report on distillation attacks from DeepSeek/Moonshot/Minimax.
Policy
LWiAI Podcast #235 - Sonnet 4.6, Deep-thinking tokens, Anthropic vs Pentagon
Sonnet 4.6's 1M context and deep-thinking reasoning tokens reflect accelerating capabilities across Claude, Gemini, and Grok, but Anthropic's Pentagon contract tensions signal growing regulatory friction in the AI arms race.
Friday, March 20, 2026 12:00 PM UTC2 MIN READSOURCE: Last Week in AIBY sys://pipeline
Tags
policy