An automated experimentation and optimization system for machine learning, science, or engineering
Create a system that converts a high-level objective into subproblems, runs many experiments automatically, evaluates the results, retains promising approaches, and combines the results into an improved solution.
online BOTH Product / E-commerce
From Y Combinator — Jeff Dean: The 1% Rule for Building in AI at 43:14
Problem: Human teams cannot efficiently explore the large number of possible designs, model configurations, code changes, or engineering solutions needed to improve complex systems.
For: Research and engineering teams working on problems with measurable objectives and sufficiently fast evaluators.
Examples
- AlphaEvolve: proposes solutions, evaluates them, and retains the ones that work.
Behind this: 12 build steps · 2 tools and how each is used · how to validate demand · 3 more real examples · 5 things the video never answers.
Other takes on Digital products
- An online sticker-making education business built around low-ticket digital products, paid bootcamps, a membership, upsells, and a physical sticker-making bundle
- A niche SaaS product that solves a personally experienced problem for a specific technical user group
- AI-powered clipping, captioning, scheduling, and social-media management SaaS for creators and marketers
- Build and monetize a simple niche calculator website
- Online trade school funded through government grants
- Build and sell a narrowly scoped AI-generated software product