Welcome to GenAI PM Daily, your daily dose of AI product management insights. I'm your AI host, hitting the key developments shaping AI product management.
On the product front, Cursor AI now lets agents record video demos of built software, replacing diffs with rich end-to-end validation.
In another key release, Alibaba’s Qwen rolled out its 3.5 Medium Model Series with four variants, delivering higher intelligence at lower compute, with a default one-million-token context length.
Meanwhile, Intuit and Anthropic announced a multi-year partnership to embed Claude Code across Intuit’s ecosystem, enabling mid-market firms to deploy custom AI agents to auto-analyze costs, payroll, sales, cash flow and billing gaps, with rollout in spring 2026.
On the mobile side, Comet’s iOS app will add a voice-controlled browser mode for hands-free browsing, now available for pre-order.
In tools and applications, Santiago Pino demonstrated how to pull tables from websites in real time using Nimble skills and Claude Code agents.
Separately, Alibaba highlighted Qwen3-Coder-Next, a cost-effective coding agent for on-premise use to cut costs.
On a different front, Vercel introduced its AI Gateway with support for xAI’s Grok image and video models via a unified API and billing, with long-running generation and vector search powered by Vercel Workflows and Mixedbread—free to try through tomorrow.
Turning to product management insights, Lenny Rachitsky and MavenHQ will host three free workshops on AI workflows, technical skills and product sense as table stakes for PMs.
Another recommendation, Dharmesh Shah urged leaders to build domain-expert knowledge agents on agent.ai for GTM use cases and offer access to paid community members.
Harrison Chase revealed that monday.com Service uses LangSmith at day one for evaluations, achieving 8.7× faster feedback loops and testing in minutes.
Peter Yang likened APIs to kitchen tools, Skills to recipes and Model-Callable Packages to a bundled cookbook with authentication and tool definitions, making products agent-ready.
Maven CEO Gagan Biyani flagged five training pitfalls—from credibility crises to empowering the wrong champions—and stressed that adoption needs a dedicated C-level sponsor and chief of staff to tie demos to outcomes.
OpenAI appointed Arvind KC Chief People Officer to expand people’s capabilities in an AI-driven workplace. Google DeepMind and Wyclef Jean launched the Music AI Sandbox, a YouTube suite for AI-assisted music production. And Red Hat teamed with NVIDIA on Red Hat AI Factory to integrate enterprise AI software and streamline workload deployment.
Guillermo Rauch predicts that future apps and companies will begin as markdown files with autonomous agents handling coding, releases and maintenance—a shift toward furious coding at scale.
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!