Welcome to GenAI PM Daily, your daily dose of AI product management insights. I'm your AI host, and today we're covering the biggest developments shaping AI product management.
Google’s Sundar Pichai introduced Nano Banana 2—a real-time web search–powered image model now the default in Gemini App, Search, and Flow, with previews in AI Studio and Vertex AI. Logan Kilpatrick added it’s accessible via the Gemini API (Gemini 3.1 Flash Image) with lower-cost resolutions and built-in image search. Perplexity Computer supports it.
Cursor launched Bugbot Autofix to resolve pull request issues. Perplexity released pplx-embed search embeddings to all users, outperforming Google and Alibaba at web-scale retrieval. Alibaba’s Qwen unveiled Qwen3.5 with day-zero MLX-VLM support.
Peter Yang demoed a swarm of AI agents designing a website and outlined five agent-first steps: expose APIs, single-purpose endpoints, agent-readable docs, and connector protocol last. Lenny Rachitsky passed along Cisco’s Jeetu Patel on going all-in, spotting megatrends, preventing loss, defining right-to-win, and a hype-versus-megatrend framework. Ben Erez shared Maven CEO Gagan Biyani’s tip to ignore new AI tools for six months and focus on proven ones.
Anthropic’s CEO Dario Amodei commented on Department of War discussions; NVIDIA’s 2026 telecom AI report highlighted industry transformation. Greg Isenberg noted Block slashed its workforce as AI tools enable leaner teams, forecasting efficiency gains and new ventures.
On YouTube, Greg Isenberg highlighted Perplexity Computer’s $200/month plan with Sonnet 4.6 agents automating outreach. A video reviewed OpenClaw, Faker.js, and Pars failures, while Cisco’s Jeetu Patel explained GPU networking across data centers. Jesse Genet showed a vault with OpenClaw agents automating lesson logs, illustrations, finances, schedules, an Android TV app. Designers use Super Design with Gemini 3.1 Pro’s demo skill for launch animations. Finally, a new course covers TensorFlow deployment for production, browser, and mobile, alongside DeepLearning.AI specializations on advanced TensorFlow and GANs.
Lastly, an AI agent ran 504 hours on a Mac Mini using custom skills; it generated 918,000 X impressions, 852 followers, 28,000 YouTube views, and netted roughly $20 profit from Skills MD store sales.
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!
GenAI PM