Real-time strategy, role-playing, and game-building have become fruitful training grounds for cutting-edge artificial intelligence systems in recent years with digital competitors easily giving their human opponents everything from StarCraft II and Dota 2 to Minecraft and go. Now, Facebook is asking for the help of the AI community to bring down NetHack – one of the most notoriously difficult titles in game history – and perhaps help computers learn to simulate instances faster using fewer resources.
NetHack is a tactical curb that passes to a sling-like dungeon crawler. Originally developed in the 1980s, but still actively updated today, the game does not expect you to win ̵
As part of the NeurIPS 2021 NetHack Challenge, Facebook invites researchers to design, train and release AI systems capable of “developing agents that can reliably beat the game or (in the more likely scenario) achieve such a high score” as possible, “according to a Wednesday FB blog post. In this way, Facebook hopes that this will not only show the NetHack learning environment as a viable reinforcement learning system, but also enable a range of potential AI / ML solutions based on both neural and symbolic methods.
“The candidate agents will play a number of games, each with a randomly drawn character role and fantasy run,” explained Wednesday’s post. For a given set of evaluation episodes for an agent, the average number of episodes in which the agent completes the game is calculated, along with the median score at the end of the episode in the game. Entries are ranked by average number of wins and, if they are a draw, by median score. ”
The competition runs from this month to October 15 with winners announced at NeurIPS in December.
All products recommended by Engadget are selected by our editorial staff, regardless of our parent company. Some of our stories include affiliate links. If you purchase something through one of these links, we may earn an affiliate commission.