Grok 4.6 Shows How Fast Your AI Options Are Expanding
Summary
This episode analyzes the shifting AI frontier, highlighting xAI’s Grok 4.6 as a competitive, cost-effective contender against OpenAI and Anthropic. It covers major funding rounds for Cognition and Lovable, strong earnings from neoclouds CoreWeave and Nebius, and Tencent’s increased infrastructure spending. The discussion also addresses Google’s internal restructuring, mixed reviews for DeepSeek V4 Pro, and evolving US government policies on model safety testing.
Topics discussed
Intro: Shifting frontier model landscape and show updates
Cognition seeks $40B valuation amid coding agent boom
Lovable raises $400M, pivots to business platform
Neoclouds CoreWeave and Nebius report massive demand
Tencent triples AI capex; China's infrastructure boom
Samsung uses AI to drastically speed up chip design
Trump admin expands AI safety testing to open models
Sponsors: KPMG, Blitzy, HyperAgent, and Harbor
Grok 4.6 release: Benchmarks and performance analysis
Grok 4.6 pricing and cost efficiency vs competitors
Community reactions and mixed user experiences with Grok
Elon Musk teases Grok 4.7 and XAI's comeback
Google's AI race: Sergey Brin's return and Gemini 4
DeepSeek V4 Pro leaks: Benchmarks vs real-world use
Ramp data: Business adoption limits of Fable 5
Conclusion: Increased choice in the AI model landscape
Listen ad-free on Castria