#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 AI’s rapid cyber capabilities, defending Mythos against skeptics while analyzing OpenAI’s GPT-5.5 Instant and its quirky "goblin" training glitch. They review new releases like XAI’s cost-effective Grok 4.3, Mistral’s unified Medium 3.5, and Anthropic’s enhanced Claude agents featuring self-improving "dreaming" modes and multi-agent orchestration.
Topics discussed
Intro, sponsors, and Box AI agent platform
Mythos biosecurity risks and AI safety skepticism
OpenAI o1 model capabilities and training quirks
Grok 3, Mistral, and Cohere model updates
AI agents, dreaming, and AI music generation
OpenAI lawsuit, Elon Musk, and board drama
XAI compute deals, Palantir, and enterprise AI
Data center sales, bank debt, and financial risks
Anthropic funding, NVIDIA, and US AI policy
NSA Mythos hack and government AI security
LoRA adapters and neural network interpretability
Recursive models, stability, and episode wrap-up
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