AI 2027 is an essay that outlines one scenario for how artificial intelligence advancement might go. It was written by Daniel Kokotajlo, Thomas Larsen, Eli Lifland, Romeo Dean, and perhaps most notably Scott Alexander, and makes a number of concrete predictions. I’m sure that over the next couple of years many people will point out where these predictions fall short- but if you’re going to criticize people for sticking their neck out and being wrong, it behooves you to stick your neck out with them.

Today we’re going to meet up and share our predictions. You’ll want to bring a phone or laptop, and have a Fatebook account. You’ll also want to have read AI 2027 before the meetup- we’ll be going over it again section by section, but we don't want to go at the speed of the slowest reader and you’ll want to be able to discuss it with people.

Plan on the essay discussion taking about an hour, and then we'll open things up to general socialization.

We'll be in the Roof Garden- look for the stairs to the west of the Google entrance, then look for the guy in the tall striped hat.