10/02/2026
Long before artificial intelligence could learn from experience, Blackwell was studying the mathematics of how to make the best decisions when the outcome is uncertain.
Born in 1919, David Harold Blackwell became one of the most influential mathematicians and statisticians of the 20th century. His research included decision theory, probability, game theory, dynamic programming and Markov decision processes.
Here’s where it gets interesting.
Modern reinforcement learning, the branch of AI where a system learns by taking actions, receiving rewards or penalties, and improving its decisions, relies on this broader mathematical framework of sequential decision-making.
Blackwell also developed the Approachability Theorem, which showed mathematically how repeated decisions can move outcomes toward a desired goal, even when you cannot predict what the other side will do.
His work on Rao-Blackwellization also helped establish ways of improving statistical estimates by making better use of available information, an idea that remains relevant across statistics and machine learning.
Blackwell wasn't trying to invent artificial intelligence. Much of his groundbreaking work came decades before today's AI revolution.
But there is a remarkable ending to the story.
In 2024, NVIDIA named its Blackwell computing architecture after David Blackwell, bringing the mathematician's name into the era of generative AI.
He developed mathematics for making better decisions under uncertainty.
Decades later, machines are learning to do exactly that.
And now some of the technology powering them carries his name.
