Coffee With Scott Adams — Knowledge Archive May 24, 2026
Scott Adams Philosophy Archive
Search ideas
Episodes Episode #2824 Segments
MainContent Moist Robot Framework

Back to episode — Episode 2824 CWSA 04/29/25

Context —

uclear secrets." Of course not. Here's some more fake news. Trump is claiming he's already made 200 trade deals, but he won't specify. He won't give any examples so you can't check on him. Does anybody think that Trump already made 200 trade deals? Does anybody think that's even a little bit true? So some of the claims that Trump is making are just hilarious. I saw Scott Bessent was in an intervi…

← Previous segment →

greeing with them, that's all you need to know to know that they were fake. 48% of voters believe that the problem with the polls is that some pollsters were deliberately adjusting their numbers to support Harris while 35% think it was because they don't know how to accurately poll. Do you think the problem is they don't know how to accurately poll? No, that's not the problem. I'm pretty sure they do know how to accurately poll because the internal pollster knew how. Do you think only the internal polls know how to poll and all the external public ones don't? No, they all know how to poll. So whatever you see that looks crooked, it's probably intentional.

So here's a scary story. Apparently the University of Zurich had been running a secret project with AI bots on Reddit and they were trying to see if they could secretly manipulate Redditors' opinions. They've been running it since November 2024. I saw this on the Reddit Lies account. Apparently it worked. So the bots, which would be just AI characters that are not real, that pretend to be real, the bots were six times more likely to change the minds of Redditors than the baseline, often by leveraging misinformation. So all the bots did is lie. And they were really successful at changing people's minds. They just lied. Then people thought, "Oh, okay. Well, they changed their minds." But apparently they had personas but they used information about their targets for persuasion. So they would know kind of what would work for each person. They would figure that out and then they would change their minds.

Now if you can change people's minds with an AI bot, did those people have free will? Because the thing that changed thei

Context —

r mind was not their minds. The thing that changed their mind was the AI. So what was it that caused them to have the opinions they had? Was it their free will or was it the AI? It's obviously the AI. What causes anybody to have the views that they have? Pretty much the news that they've seen and the information they've absorbed. So for the most part, we're very close to proving that free will is…

Next segment → →