Skip to content

Questions about Ray Solomonoff

Short answers, pulled from the story.

What did Ray Solomonoff invent?

Ray Solomonoff invented Algorithmic Probability and the General Theory of Inductive Inference, also called Universal Inductive Inference. He is also a founder of algorithmic information theory and is credited as an originator of the branch of artificial intelligence based on machine learning, prediction, and probability.

When did Ray Solomonoff first describe Algorithmic Probability?

Solomonoff first described Algorithmic Probability in February 1960, in a report titled "A Preliminary Report on a General Theory of Inductive Inference" and at a conference at Caltech that same year. He published a more complete treatment in 1964 in two papers titled "A Formal Theory of Inductive Inference."

How is Ray Solomonoff related to Kolmogorov complexity?

Solomonoff's 1960 work established the basic theorem that later became known as Kolmogorov complexity. Andrei Kolmogorov independently published related ideas in 1965 and acknowledged Solomonoff's priority. The scientific community nonetheless attached Kolmogorov's name to the complexity measure while associating Solomonoff's name with algorithmic probability and universal induction.

Was Ray Solomonoff at the original 1956 Dartmouth AI conference?

Yes. Solomonoff was one of the ten original invitees to the 1956 Dartmouth Summer Research Conference on Artificial Intelligence. He, John McCarthy, and Marvin Minsky were the only attendees to stay the entire summer. He circulated a report at that conference titled "An Inductive Inference Machine," now regarded as one of the first papers on probabilistic machine learning.

What is Solomonoff Induction?

Solomonoff Induction is a method of predicting the next event in a sequence by adding up the predictions of all models that describe the sequence, weighting each model by the length of its description, with shorter descriptions receiving higher weight. It uses Algorithmic Probability within a Bayesian framework and is the only probability system known to be complete, meaning it will find any describable regularity given enough data.

What award did Ray Solomonoff receive in 2003?

In 2003, Solomonoff became the first recipient of the Kolmogorov Award, given by the Computer Learning Research Center at the Royal Holloway, University of London. He delivered the inaugural Kolmogorov Lecture on that occasion.