“Say What You See” —Artificial Intelligence at Play

As artificial intelligence (AI) moves into every corner of modern life, it has enhanced how we have fun and seek connection. AI’s errors and odd responses can be super fun, and that understated fact has created a giant game environment.

Some of this “gaming” is literal, and these games lay out the basic framework of the broader, global game. For instance, Google has created a literal AI game called Say What You See. In this game, a little cartoon brain mascot prompts the player with AI-generated images, which the player must then describe in 120 characters or less. That description is then fed back into the image generator, and if the new generation is a plausible replica of the original, the player gets a passing score. It’s like a visual, AI-mediated game of Telephone

The rules of Say What You See are simple, and yet the game increases in difficulty as it goes along, forcing players to dig deep into their brains for words typically used to describe objects of beauty and transcendence, and cram them into a cold, emotionless machine. 

For example: The player may be shown an object they might call a “little brown horse” if they saw one on a shelf. But the prompt “little brown horse” will produce nothing close to the image the player needs. So the player’s thought process may start to resemble the following: Is that a wooden horse figurine on a wooden stand, depicted walking? Is it actually loping? I think it might be loping, but does the robot need the player to be that pedantic? Would the AI respond better to “statuette” than to “figurine”? It does!

Invite students to experiment with this AI game. Can they “say what they see”? How did AI respond to their descriptions? In what ways did their process of thinking evolve?

Share this post: