In “The Toynbee Convector,” a Ray Bradbury short story, a time traveller tells everyone that in 100 years the world will be a utopia:

“We did it! The future is ours. We rebuilt the cities, freshened the small towns, cleaned the lakes and rivers, washed the air, saved the dolphins, increased the whales, stopped the wars, tossed solar stations across space to light the world, colonized the moon, moved on to Mars, then 1 Alpha Centauri. We cured cancer and stopped death. We did it—Oh Lord, much thanks—we did it. Oh, future’s bright and beauteous spires, arise!”

It turns out the man was lying, but society, inspired by his vision, successfully builds the world he proposed. Today’s most popular sci-fi narratives about AI—Her, The Terminator, 2001: The Space Odyssey—are limiting our imagination of what’s possible. What might happen when master storytellers and technical experts collaborate to envision expansive AI futures?

This spring, I hosted a convening at Stanford for AI researchers and sci-fi filmmakers. Each researcher-writer pair translated a piece of research into a possible future, character, and scene. We’re currently working on publishing the stories and sharing learnings.

More broadly, Stories for the Future is an initiative to create nuanced, forward-looking narratives about AI. The work is funded by Stanford HAI.