#244 - GPT-5.5 Instant, Grok 4.3, OpenAI vs Musk
May 11, 2026
More Technology podcasts
Summary
Hosts Andrei Kurenkov and Jeremy Harris discuss Anthropic’s Mythos, defending its impact against skeptics while warning of future cyber and bio risks. They cover OpenAI’s GPT-5.5 Instant, analyzing its improved capabilities and the quirky "goblin" training anomaly. The episode also reviews new releases from xAI’s Grok and Mistral, plus Anthropic’s agent updates featuring self-improving "dreaming" and multi-agent orchestration.
Topics discussed
Intro ads and podcast welcome
Anthropic's Mythos release and AI progress skepticism
ChatGPT 5.5 capabilities and the 'gremlin' training bug
New models: Grok 3, Mistral, and agent orchestration
Elon Musk vs OpenAI drama and Siobhan Zilis interview
Anthropic/SpaceX deal, valuation, and data center financing
DeepSeek funding, Anthropic interpretability, and OpenAI MRC
US Gov AI policy, DoD deployment, and Microsoft security
Safety adapters, recursive transformers, and Connect4 AI
More from Last Week in AI
#261 - (belated) Argon, Muse Spark, Anthropic S-1
Nov 20, 2026
#260 - Muse, Misalignment, Math
Oct 19, 2026
#259 - Dots, Sonnet, Self-Safety, Rogue AI
Oct 9, 2026
#258 - Opus 5.5, Sol and Luna, Muse, DeepSeek-V4.1-Flash, Xi
Oct 2, 2026
#257 - GPT 6 Astra, AI Extinction, Security Incidents
Sep 24, 2026
Listen ad-free on Castria