AI model out-strategizes humans

“An AI model has surpassed the best human players at the game of Stratego. The strategy game, which includes a ‘fog of war’ mechanic that conceals crucial information about the opponent’s pieces, has long posed a difficult challenge for artificial intelligence because players must make decisions with incomplete information. In 2022, a previous model called DeepNash was able to reach the level of top human players but did not demonstrate clear superiority over the very best. This new model, Ataraxos, won 38 out of 40 games against top players during demonstrations at the Stratego World Championship and won 15 out of 20 games against Pim Niemeijer, the strongest human player,” reports The Doomslayer.

MIT News explains:

“Using advances in machine-learning, researchers from MIT, Carnegie Mellon University, New York University, and Stanford University developed an AI that defeated top-ranked human players of the board wargame Stratego by a large margin — something no AI system had been able to achieve.

Stratego, a two-player game of imperfect information, in which the opponent’s piece identities remain hidden, is often used as a benchmark to test the strategic thinking abilities of powerful AI models.

To build their model, the researchers combined efficient training algorithms with new techniques tailored for calculated decision-making in hidden information settings.

The AI system achieved greater performance at Stratego than the next best models, while being far cheaper and less computationally demanding to train. The system also outperformed top human players in other strategic games with different rules and designs, demonstrating how it can be generalized for a variety of use-cases…

Past efforts, such as Google’s DeepMind, relied on sophisticated operations that were computationally demanding and costly. But even with millions of dollars in training costs, these models were still not strong enough to beat top human Stratego players.”

Nature adds:

“Real-world decision-making generally involves hidden information, that is, information that is unknown to one agent but possessed by another. Unfortunately, the presence of large amounts of hidden information renders established reinforcement learning and search approaches ineffective. Even with multimillion-dollar industrial research efforts¹, top-human-level play at Stratego—a board wargame with hidden information on a massive scale—has remained beyond the reach of artificial intelligence (AI). Here we introduce Ataraxos, an AI for Stratego based on general techniques that we developed for both self-play reinforcement learning and test-time search under hidden information. Ataraxos defeated the most decorated human Stratego player of all time by a large margin—achieving, to our knowledge, the first superhuman result in the game’s history—while consuming orders of magnitude less compute and data than previous efforts. Using the same techniques, we built a superhuman AI for Barrage Stratego and state-of-the-art AIs for Hanabi and dou dizhu, all with low cost and high sample efficiency. The success of this approach across adversarial, cooperative and team games establishes a design pattern for reinforcement learning and search that is effective under large amounts of hidden information, a longstanding desideratum of the field of strategic decision-making.”

Artificial intelligence appears to be have created more jobs than it destroyed.

Artificial intelligence is also getting cheaper.

Artificial intelligence helped a company detect a big source of geothermal energy.

The nation of Zambia used artificial intelligence to find new mineral wealth.

Kenyan farmers are using artificial intelligence to produce much more food.

Artificial intelligence is greatly improving the detection of many different diseases.

The FDA has approved a robot to draw people’s blood.

Robots with artificial intelligence are spreading on Japanese farms. On some American farms, there are drones with artificial intelligence, and robots that use artificial intelligence to kill 100,000 weeds per hour.