Google is moving beyond AI as a mere analysis tool. The concept of an 'AI co-scientist' envisions AI as an active partner that helps sift through information, generate novel hypotheses, and outline ways to test them. This reframes the human-AI collaboration to fundamentally accelerate the scientific method itself.

Google's research head distinguishes between innovation—the continuous, iterative process of improvement applied across product and research—and true breakthroughs. Breakthroughs are fundamental shifts that solve problems not previously solvable in principle, such as the Transformer architecture that underpins modern AI.

The partnership where OpenAI becomes an equity holder in Thrive Holdings suggests a new go-to-market model. Instead of tech firms pushing general AI 'outside-in,' this 'inside-out' approach embeds AI development within established industry operators to build, test, and improve domain-specific models with real-world feedback loops.

Conventional innovation starts with a well-defined problem. Afeyan argues this is limiting. A more powerful approach is to search for new value pools by exploring problems and potential solutions in parallel, allowing for unexpected discoveries that problem-first thinking would miss.

In AI, low prototyping costs and customer uncertainty make the traditional research-first PM model obsolete. The new approach is to build a prototype quickly, show it to customers to discover possibilities, and then iterate based on their reactions, effectively building the solution before the problem is fully defined.

To avoid choosing between deep research and product development, ElevenLabs organizes teams into problem-focused "labs." Each lab, a mix of researchers, engineers, and operators, tackles a specific problem (e.g., voice or agents), sequencing deep research first before building a product layer on top. This structure allows for both foundational breakthroughs and market-facing execution.

A key strategy for labs like Anthropic is automating AI research itself. By building models that can perform the tasks of AI researchers, they aim to create a feedback loop that dramatically accelerates the pace of innovation.

Moving from a science-focused research phase to building physical technology demonstrators is critical. The sooner a deep tech company does this, the faster it uncovers new real-world challenges, creates tangible proof for investors and customers, and fosters a culture of building, not just researching.

While a tight product-research link is beneficial, it creates a management challenge where teams get so excited about implementation they neglect the next big research question. The research leader's role includes making the difficult judgment call to shift focus back toward long-term discovery, even amid product success.