Gerald Tesauro

American computer scientist whose backgammon program TD-GammonTD-Gammon, prop. n.: A backgammon program built at IBM in the early 1990s that taught itself to play at near world-champion level by playing more than a million games against itself. (1992), developed at IBM's Thomas J. Watson Research Center, gave reinforcement learningreinforcement learning, n.: A way of teaching a computer program by trial and error: it tries actions, earns points for good results and penalties for bad ones, and gradually learns which choices pay off over time. with neural networks its first major success. Trained as a physicist, he first built Neurogammon, a network trained on expert moves that won the 1989 Computer Olympiad, then trained TD-Gammon by self-playself-play, n.: Training an AI by having it play against copies of itself, so it always has an opponent at its own level and can keep improving without human examples. alone, updating a multilayer perceptronmultilayer perceptron, n.: A neural network made of several layers of artificial neurons, each connected to every neuron in the next layer. It is the simplest kind of deep network and a component of almost every modern model.'s evaluations of positions with temporal difference learningtemporal difference learning, n.: A way of learning predictions by adjusting each guess toward the next, slightly better-informed guess instead of waiting for the final outcome — like revising a weather forecast every day as the date approaches.. Starting from random play, later versions trained on some 1.5 million games played close to the level of the best human players and changed expert opinion about how to play certain opening positions. Its success, in a game with dice and an enormous number of positions, inspired later deep reinforcement learningdeep reinforcement learning, n.: Reinforcement learning in which a large neural network does the learning, letting an AI learn tasks such as video games or robot control directly from raw inputs like screen pixels or camera images. and self-play systems such as AlphaGoAlphaGo, prop. n.: A DeepMind program that in 2016 beat Lee Sedol, one of the world's best players of the board game Go, a decade or more before many experts expected a computer to manage it.. He later devised the wagering strategy of IBM WatsonIBM Watson, prop. n.: An IBM question-answering computer that beat two champions of the US quiz show Jeopardy! in 2011. IBM then sold business AI products under the Watson name, many of which disappointed. for its 2011 Jeopardy! matches.