Tensor: A Game of Equilibrium

In this game players are managers of teams that are designing artificial intelligences. You are all working for the same research university, which will only give precious computer processing time to the teams whose AIs are doing the best. The game takes its inspiration from real-world “deep learning” techniques used when designing and building real AIs.

Unlike many games, Tensor is game of equilibrium, not efficiency, and you do not win by VPs. You will balance a board of 5 “weights” across many rounds through careful tile selection and placement, as well as special cards that let you change your tiles in unique ways.

On their turn the player takes one action, often placing a hexagonal tile on their board. The game ends when the draw pile of hexagonal tiles is exhausted.

At the end of the game, the player whose lowest weight is the highest wins! Be careful, as other players can hack your board!