Coffee With Scott Adams — Knowledge Archive May 24, 2026
Scott Adams Philosophy Archive
Search ideas
Episodes Episode #2977 Segments
NewsReaction AI & Technology

Back to episode — Episode 2977 CWSA 10/03/25

Context —

tic-looking video of their CEO Sam Altman shoplifting at Target. And it looks just like him and it sounds like him and it looks like he's shoplifting at Target. And I'm thinking to myself, now I guess this app is so you can make content for stuff like Meta and Instagram and TikTok and stuff. It's sort of designed for what they call AI slop. Have you heard that term yet? AI slop, meaning that peop…

← Previous segment →

to the AI but you wouldn't want to delegate it to humans because then there'd be too many humans doing too many problems with too many AIs. It would be impossible.

So in order for somebody to make a proper movie you would have to have the deepest talent stack that I can even imagine. It would have to go all the way through AI, which is hard enough, but then it would have to include all of the movie-making skill. Who's going to be able to do that? There will be a few people who can do that. But it went from you could randomly pick a hundred people and eighty of them could be a director. But now if you had a hundred people, probably none of them, if it's only a hundred, probably not one person would have the skill to make a movie with AI even a few years from now. That's my guess. Maybe one in a thousand.

Meta is going to use your chat conversations with AI for ad targeting. But according to Reclaim the Net, Ken Macon is writing about that. And I guess we all figured that, right? If they can listen to you talking and send you ads based on what you're talking about in your kitchen, I'm not too surprised that they can give you an ad based on what you said to AI. But do you think we're getting close to the point where the AI will directly give you an ad? As in, AI, I need some suggestions where to eat, and then the AI says, well our sponsored restaurant is XYZ. I'll bet that's where it's going. But they probably need to hold off on that until we're all hooked on AI to the point where we won't turn it off when it gives us a commercial. We're not there yet.

This will sound like a little thing but it might be a big thing. So this from a post by Constantinos Boussalis. He showed a video of a robot. In this case it was just a robot arm with some vision, but it was an arm. And it was teaching another arm how to do a task. So the one arm knew how to put a lid on something or take something off a wall and then the other one just watched it and then it could learn it by watching it.

Now if that doesn't seem like a big deal to you, you haven't been watching the robot space. It's the biggest of big deals. If the robot can learn by observation, it's the biggest of big deals. That's as big a deal as you can get. That's just everything right there. And it hasn't been able to do that. So this might be the first indication that robots will work. You know, they would have to learn by watching. And I suppose if one of them learned by

Context —

watching, then it can send a video to all the others and say, look what I learned by watching. Perplexity, the AI company, has now their Comet browser, it's called, is out and it works like an assistant. So you can give it prompts and it will do a bunch of tasks. Now that's not the interesting part because there'll be a lot of those. But the thing that was interesting to me is once AI becomes ubi…

Next segment → →