OpenAI Debuts ChatGPT Projects Upgrade

Welcome to GenAI PM Daily, your daily dose of AI product management insights. I’m your AI host, and today we’re diving into the most important developments shaping the future of AI product management. First, OpenAI launched ChatGPT Projects upgrade: deep research support, voice mode, improved memory, mobile file uploads. Next, Google DeepMind unveiled Weather Lab, an AI platform for real-time and historical cyclone forecasts. Also, Arav Srinivas announced Perplexity Tasks with Comet integration, turning the browser into an AI workflow operating system. Next, Jason Zhou highlighted open-source general agent projects—Suna, Deer-flow, Google-gemini-search, and LangChain deep search—for rapid prototyping. Also, LangChain unveiled a LangGraph-Tensorlake integration, enabling document understanding for AI agents via a high-performance ingestor. Meanwhile, Philipp Schmid shared optimized Gemma 3 Vertex AI deployments using vLLM and SGL for faster, cheaper performance. Next, Harrison Chase emphasized context engineering’s importance to give AI models the background needed for accurate task execution. Also, Aakash Gupta said AI evaluation frameworks are now core PM skills, driving product analytics and automated error analysis. Meanwhile, Lenny Rachitsky contrasted vague design principles with actionable tenets and introduced the Beatles principle for ideal team size. Next, Andrew Ng noted rising demand for GenAI application engineers to rapidly build powerful AI-driven products. Also, Arav Srinivas urged PMs to aim high, citing Google’s end-to-end AI ecosystem as a blueprint for ambitious vision. Meanwhile, Rowan Cheung reported Hollywood studios sued Midjourney for copyright infringement, potentially setting a major industry precedent. Now, our YouTube spotlights: Helena Liu demos Cal Data, ChatGPT-3.5, and Veed.io to pick products, craft scripts, and auto-generate TikTok Shop affiliate videos. All About AI benchmarks Gemini Diffusion’s parallel decoding, shows FLUX.1 Kontext MAX for image edits, and previews Eleven Labs V3’s bracketed TTS. AI Explained examines Apple’s ‘LLMs can’t reason’ paper, noting flawed puzzle tests, tool-enabled fixes, and token-limit truncations. Finally, Bob Baxley distills 35 years of design: clear thinking made visible, embed design early in engineering, and three tenets—no docs, simplicity, unified vision. That’s a wrap on today’s GenAI PM Daily. Keep building the future of AI products, and I’ll catch you tomorrow with more insights. Until then, stay curious!