Anthropic’s Mythos is Back, OpenAI Releases GPT 5.6, Apple’s Price Increases
Summary
Host Alex and guest Ranjan Roy discuss the US government lifting restrictions on Anthropic’s Mythos model and OpenAI’s limited preview of GPT 5.6, highlighting concerns over regulatory control and open-source competition. They analyze how rising costs are driving enterprises to cheaper models, threatening the IPO narratives for OpenAI and Anthropic. The episode also covers SpaceX’s post-IPO valuation drop and Apple’s significant price hikes on Macs and iPads due to memory costs.
Topics discussed
Intro: GPT-5.6, Apple price hikes, and memory crunch
US lifts restrictions on OpenAI's Mythos model access
GPT-5.6 model names: Sol, Luna, and Terra controversy
Geopolitics of AI: Restricted access vs. open source
Model distillation risks and Anthropic's safeguards
GPT-5.6 cybersecurity performance and pricing
OpenAI IPO timing and market sentiment shifts
Enterprise pushback: Switching to cheaper AI models
AI billing opacity and overcharge audits
SpaceX valuation drop and market froth
Apple price hikes: Memory costs or greed?
Tribute to Om Malik and closing remarks
Listen ad-free on Castria