I’ve been having a lot of fun making AI videos—slop, if you like—with Sora this past week.
I’m pretty sure you’ll have fun with it too when it becomes public.
(At time of writing the app was only available in the US App Store and is invite only and I have a spare invite if you’re keen.)
It’s been fascinating seeing all of the negative takes roll in from friends and strangers. All are welcome here, whether in the social media comments or my inbox—this is a free-to-comment space.
I’ve found Sora genuinely interesting for the same reasons Mike Gioia wrote about this week.
The Sora app is particularly addicting and fun. Sora offers the familiar addiction of short-form video feeds along with a social network and notifications about friends making videos with your likeness. There are new unclear social norms to test as well. Is it uncouth to cameo with a married woman? Can you offend someone with their own AI avatar? How should you feel that someone made your avatar do that? Almost immediately I made videos of acquaintances doing bizarre things. Then I made videos of me slapping a friend in increasingly surreal ways. In both cases they seemed to think it was funny.
Most of the criticism I’ve seen falls into two camps.
The first is about quality—the look of the video, the writing, the prompting. Matters of taste. And that’s fair. Good writing, direction, and prompting will always stand taller than whatever chaos I’m cooking up. Taste matters. That’s the point.
The goal of Sora is not to create entertainment for disinterested viewers. The purpose is to entertain yourself and your friends. It’s peer-to-peer media.
The second kind of criticism is the one that sticks with me. I can’t quite pin it down, but it carries a certain protective snobbery. A whiff of creativity used to belong to us—because we studied, bought the gear, put in the years, built the following—and now anyone with an app can join the party. This is where I find Hank Green’s popular take. It’s really cool to be anti-AI at the moment. I personally just think AI is the current catchphrase for the current state of computing.
Peer-to-peer media is a huge shift away from media that’s made by specialists as a product to be consumed by a wide audience.
That attitude will age like milk.
Mike’s most interesting question is the last one in his post:
Why did OpenAI make this? Why would OpenAI build a social app designed to entertain you? Only a few years ago Sam Altman was trying to scan everyone’s eyeball for World Coin. People called him weird. Now everyone is voluntarily scanning themselves into his social video app.