The Next Wave of Enterprise AI
Summary
This episode analyzes the confusing rollout of a new AI executive order requiring voluntary model sharing for cyber safety, alongside Anthropic’s expanded Mythos access and SK Hynix’s chip capacity plans. It then explores the shift to enterprise AI efficiency, highlighting OpenAI’s Codex updates for knowledge workers and Microsoft’s cost-optimized model releases. The discussion emphasizes how organizations are adapting to token scarcity and rising costs.
Topics discussed
Intro, sponsors, and show overview
Trump's AI executive order: context and details
Reactions to the EO from media, labs, and politicians
Anthropic expands Mythos access via Project Glasswing
SK Hynix doubles memory chip capacity amid shortage
Sponsors: KPMG, Robots and Pencils, AssemblyAI, OutSystems
OpenAI Codex shifts to non-technical knowledge workers
The 'strange abundance' of modern knowledge work
New Codex features: Annotations and Role Plugins
Codex Sites: Disposable web apps for collaboration
Microsoft Build: New MAI models and benchmarks
Microsoft's cost optimization strategy and conclusion
Listen ad-free on Castria