Answer extracted from the Invest Like The Best podcast — listen to the full episode below.
What decision-making framework should investors use when evaluating AI companies in a period of rapid technological change?
Start by mapping what the technology actually makes possible, then determine what is genuinely valuable within that possibility space, and finally identify founders aligned with that vision. Rather than evaluating market problems in isolation, you must understand the underlying technology deeply from first principles—recognizing structural fits between what AI can do and where it solves real problems.
In a landscape moving at extraordinary speed, most investors evaluate AI opportunities by asking: what customer problem does this solve? But that frame, Sarah explains, misses the essential step that must come first. You cannot know what problem an AI company should solve until you understand what the technology fundamentally enables.
The real work is recognizing where technology and value align naturally. Consider Harvey, the AI law company—a canonical example of this framework in action. Law is structured language. Lawyers spend their days reading documents and generating text from precedent. Language models excel at exactly these tasks. The fit is not accidental; it emerges from understanding what language models do at a technical level and recognizing how that capability maps onto a real professional workflow. This alignment didn't require inventing a new market; it required seeing what was always latent in the work itself.
This approach stands in sharp contrast to chasing problems first. When you lead with customer pain, you risk optimizing for last-generation solutions to last-generation problems. In AI's current moment, where capabilities are still rapidly evolving, that backward-looking frame can blind you to genuine innovations. As Sarah notes, the landscape feels "frenzied" precisely because the rules are changing—and your evaluation method must account for that instability.
The three-step framework is deliberate: First, achieve deep understanding of what the technology can do. Second, map where that capability creates real value (not imaginary value, but genuine structural advantage). Third, assess whether the team building the solution has truly internalized both steps and is aligned with the opportunity as you now understand it. The founder who understands their domain at first principles and grasps the technology's true constraints is far more likely to navigate the inevitable pivots ahead.
"If you think about the opportunity for the ecosystem in the future, is there going to be a competitive Western open source model?"
Sarah — Investor specializing in early-stage AI, biology, defense, and robotics, with deep involvement in venture capital. Sarah co-founded or leads a fund with partners including Mike, Pranav, and Bella, and has spent her career drawn to high-agency people and the structural forces shaping technology outcomes at scale.
This question reflects the breadth of Sarah's thinking: she doesn't just evaluate individual companies in isolation. She traces how foundational questions—like whether the West can sustain competitive open-source AI development—shape the entire investment landscape. The full conversation explores how geopolitical and industrial policy questions cascade down into founder selection and capital allocation decisions, grounding her abstract principles in real strategic choices.
For investors grappling with backtest-resistant markets, this framework offers something concrete: a method that doesn't rely on historical patterns. You cannot predict where AI will go, but you can understand what it does today. You cannot guarantee a market will emerge, but you can recognize when technology and human work are genuinely misaligned. You cannot bet on narrative; you must bet on founders who see clearly.
Key takeaways
- Understand technology from first principles before evaluating market problems—this is the sequence that prevents false negatives on transformative opportunities.
- Look for structural alignment between what the technology does and what the work itself demands, not just acute customer pain points.
- Assess founders on their depth of understanding in both the domain and the technology, as clarity about constraints is more reliable than confidence in growth.
- Recognize that evaluation in a rapidly changing landscape requires a framework independent of backtesting or historical precedent.