Back to episode — Episode 2213 Scott Adams - How I Could Easily Program AI To Spot Fake News. It's Pattern Recognition
Context —
ous report? Okay. Do you think you could teach AI to spot a Rupar? Do you know what a Rupar is? So Rupar is named after journalist named Rupar. It's when you take out some context from a story to reverse its meaning. If you asked AI was the fine people hoax a hoax, how hard would it be for AI to figure out what it was? Because it would know to look for context that's missing. It would just go loo…
← Previous segment →ody. So I don't know that he's, I don't think he's got TDS at all actually. In my interactions he has a point of view and sometimes he disagrees with mine. But I'm not talking about changing somebody's mind. So the reason Mark would not be a good choice for deprogramming is I don't think he's hypnotized in the first place. Bill Maher clearly has something going on. There's some kind of distortion there. Mark Cuban is just somebody who has a set of preferences that he argues very well. He doesn't seem to have any kind of illusions going on. I've never seen any.
Now and then, you know, the problem with a Rob Reiner, obviously wouldn't do it. But if somebody like a Rob Reiner wanted to experiment, how could you trust an actor? Don't you suspect that Rob Reiner is acting and that he doesn't necessarily believe everything he tweets but it's good for his side, right? It's just he's supporting his team. So I can't read his mind so I don't know what Reiner is thinking. I know that the way he acts doesn't look rational but I don't know that he doesn't know that. I don't know that he's unaware that he's playing a role. To me it looks like a role. So you can't really trust an actor because you never really know because they'd be too good at fooling you because they're actors.
Pick someone young 'cause someone old would be too difficult to deprogram. But I think it's a good show whether they get deprogrammed or whether you see cognitive dissonance. It would be interesting either way 'cause I would tell you what to look for. I would say before we start the deprogramming I'll tell you what to look for if cognitive dissonance happens. So I'd say well one thing to look for is the filibuster where the person keeps talking and won't let me orient the conversation. That would be a sign. Another sign would be the word salad where they're talking but the sentences are like, I'm not even sure those are making any sense. And there are a few others. But you know it'd be fun to tell you what to look for and then create the cognitive dissonance and then call it out. And the person experiencing it will just get mad because they won't see it. You know, they'll think you're mocking them or something. All right, well if I find the right person for that I'll do it.
So Vivek is being called some kind of a Russian asset. That's the latest one. Let me see. Let's put this through the AI that doesn't exist yet. Typing into the AI. Let's see if you can find a pattern. There's a Republican candidate who's a populist. He's rising in the polls, exceeding expectations. And he's being accused of being a Russian plant, that really he's working for Russia. So see if I can come up with any pattern. Yep, oh it came up with a pattern recognition. Yeah, yeah. It turns out that when populist Republicans who don't want war, when they're involved there seems to pop up a Russia collusion hoax. How often? Let me check how often. 100%. 100% of the time.
Yeah. Now seriously, you don't think an AI
Context —
could spot this? Anybody, is there anybody who thinks AI can't spot this? Come on. The pattern could not be more obvious, right? If I can spot it I'm pretty sure AI can spot it. But this is based on, I guess there's some Russian news channel where the Russians said their first choice would be Trump because he wants to end the war and their second choice would be Vivek because he wants to end the w…
Next segment → →