(2025-09-08) Barber Career & startup advice for the AI era

Chris Barber: career & startup advice for the GenAI era... AI makes some things cheap and easy. But not everything. The things that it doesn’t make cheap and easy but that are still demanded by customers then become the bottlenecks, and are generally more valuable.

With more competition and more leverage, the returns to doing the right thing (for you) go up - as Andy Rachleff would say, do the thing that some people desperately want that you uniquely provide.. But finding that isn’t easy with top down planning. Top down planning works for stable environments with known territory, not high uncertainty.

With uncertainty, the move is to run lots of small experiments. Contact reality and try things. (Tiny Experiments)

*Intuition led decision making tends to naturally be suited for this kind of bottoms-up navigation of uncertain territory. Good small experiments often make intuitive sense but may be hard to logically justify to others, they might seem silly or have no good reason. *

*Cycle between doing experiments, noticing where you get disproportionate (exponential) returns on effort, and doubling down on those. Start with small experiments, then intuitively do bigger ones and explore adjacencies from things that showed promise, and repeat. *

(Note that this advice requires AI fluency as a pre-requisite.)

See also: Amdahl’s law, Baumol’s cost disease, Theory of Constraints, bottlenecks, Jevons paradox, Moravec’s paradox, Lottery Ticket Hypothesis, Hirsch’s positional goods, Red Queen hypothesis.