Automating AI research
Thesis is building Darwin, a machine that runs its own machine learning experiments and uses the outcome of each one to choose the next, pointed at the grand challenges of biology and materials science. Luigi is a second-time founder. He built Sphere, a fintech that processes billions of dollars a year in cross-border payments and was valued at more than $250M in its Series A, scaling stablecoin payments infrastructure from zero. He studied computer science and math at Columbia. Thesis (YC F25) is state of the art on OpenAI's MLE-Bench, reached in one month on $10k of compute, ahead of research teams at Microsoft, Google, Meta and Baidu.
Luigi Charles is a co-founder of Thesis, an AI research lab automating AI R&D. He previously built Sphere, a cross-border payments company that processes billions of dollars a year and was valued at more than $250M in its Series A.
A Y Combinator F25 company building Darwin, a hill-climbing machine that searches vast combinatorial spaces of machine learning experiments, using the outcome of each experiment to optimize the next.
It is state of the art on OpenAI's MLE-Bench, the benchmark for how well AI systems train machine learning models autonomously. Thesis reached that in one month with $10k of compute, ahead of research teams at Microsoft, Google, Meta and Baidu.
Sphere, where he scaled stablecoin payments infrastructure from zero into a business processing billions of dollars a year in global cross-border payments.
At biology and materials science. The stated ambition is a machine that discovers the next Transformer or the next AlphaFold rather than a tool that assists with one experiment at a time.
His brother Sergio Charles, Thesis co-founder and CEO, who worked on AI R&D at Google X, Nvidia and Stanford's AI Lab.
