In this episode of Last Week in AI, Andrey Kurenkov and guest co-host Jon Krohn discuss the latest AI news, including OpenAI's Sora 2 and its new iOS app, Anthropic's Claude Sonnet 4.5 and Claude Code 2.0 releases, Meta's Vibes video platform, and OpenAI's ChatGPT Pulse. They also cover OpenAI's safety routing system, Google's Gemini 2.5 Flash updates, Microsoft's AI agents for Office 365, and OpenAI's agentic shopping system. Additionally, they touch on Thinking Machines Lab's new product, OpenAI's valuation, Elon Musk's lawsuit against OpenAI, and Periodic Labs' seed funding. The episode further explores new benchmarks like SWE Bench Pro, mechanistic interoperability work, and empirical evaluations of reasoning and financial models, as well as California's AI transparency laws and Spotify's struggles with AI-generated content.
Sign in to continue reading, translating and more.
Continue