A new product development principle for AI is to observe the model's "latent demand"—what it attempts to do on its own. Instead of just reacting to user hacks, Anthropic builds tools to facilitate the model's innate tendencies, inverting the traditional user-centric approach.
Related Insights
Intentionally create open-ended, flexible products. Observe how power users "abuse" them for unintended purposes. This "latent demand" reveals valuable, pre-validated opportunities for new features or products, as seen with Facebook's Marketplace and Dating features.
Traditional SaaS development starts with a user problem. AI development inverts this by starting with what the technology makes possible. Teams must prototype to test reliability first, because execution is uncertain. The UI and user problem validation come later in the process.
Users often develop multi-product workarounds for issues they don't even recognize as solvable problems. Identifying these subconscious behaviors reveals significant innovation opportunities that users themselves cannot articulate.
The essential skill for AI PMs is deep intuition, which can only be built through hands-on experimentation. This means actively using every new LLM, image, and video model upon release to objectively understand its capabilities, limitations, and trajectory, rather than relying on second-hand analysis.
Beyond standard benchmarks, Anthropic fine-tunes its models based on their "eagerness." An AI can be "too eager," over-delivering and making unwanted changes, or "too lazy," requiring constant prodding. Finding the right balance is a critical, non-obvious aspect of creating a useful and steerable AI assistant.
A truly "AI-native" product isn't one with AI features tacked on. Its core user experience originates from an AI interaction, like a natural language prompt that generates a structured output. The product is fundamentally built around the capabilities of the underlying models, making AI the primary value driver.
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.
The traditional SaaS method of asking customers what they want doesn't work for AI because customers can't imagine what's possible with the technology's "jagged" capabilities. Instead, teams must start with a deep, technology-first understanding of the models and then map that back to customer problems.
AI is evolving from a coding tool to a proactive product contributor. Claude analyzes user feedback, bug reports, and telemetry to autonomously suggest bug fixes and new features, acting more like a product-aware coworker than a simple code generator.
Building a true AI product starts by defining its core capabilities in an AI playground to understand what's possible. This exploration informs the AI architecture and user interface, a reverse process from traditional software where UI design often comes first.