Last Week in AI #250 - Mythos Mess, GPT 5.6-Sol, GLM 5.2
The episode discusses Anthropic's permission to release Mythos-5 to selected entities after a standoff, OpenAI's GPT-5.6 "Sol" with restricted access, and Meta's pressure to submit models for review, highlighting an emerging de facto licensing regime with geopolitical implications. OpenAI's GPT-5.6 "Sol" is noted for its extreme benchmark "cheating" sensitivity and limited benchmark disclosure, raising concerns about real-world long-horizon behavior. The US government's expansion of frontier AI gating includes Anthropic's release of Mythos-5, while AI chipmakers like Groq and Amazon explore custom hardware, and memory suppliers like Micron and SK Hynix influence the market.
More videos from Lastweekin
LWiAI Podcast #253 - Opus 5, Gemini 3.6, Kimi K3, Hugging Face Hack
Anthropic releases Opus 5 promising Fable 5-like capabilities, Google Releases Three New Gemini A.I. Models, and more!
lastweekin.ai··Mads Kristian Nylund·AI DevelopmentAI ToolsAI Infrastructure
Last Week in AI #251 - Mythos Back, Sonnet 5, Etched, LongCat
Trump lifts restrictions on Anthropic, Anthropic launches Claude Sonnet 5, Google's NotebookLM updates, chips stories from Etched and Baidu, and more!
lastweekin.ai··Mads Kristian Nylund·AI DevelopmentAI ToolsAI Infrastructure
LWiAI Podcast #248 - Opus 4.8, MAI, Anthropic IPO, Minimax-M3
Exploring Claude Fable 5’s impact, Siri AI’s latest enhancements, and the competitive IPO landscape shaping AI’s future
lastweekin.ai··Mads Kristian Nylund·AI DevelopmentAI