Coffee With Scott Adams — Knowledge Archive May 24, 2026
Scott Adams Philosophy Archive
Search ideas
Episodes Episode #2941 Segments
NewsReaction Politics as Persuasion

Back to episode — Episode 2941 CWSA 08/28/25

Context —

it's pretty bad. Black plastic is made from recycled electronics such as old televisions, computers, and other electronic waste. So, allegedly, the black plastic contains flame retardant chemicals. There are just all kinds of chemicals, and you're eating it. Now, how many of you have had that experience where the news serves up exactly what you were thinking and it wasn't even like a normal thing…

← Previous segment →

thinking about those damn plastic containers changed reality until something was presented to me on that topic. Maybe. So, that's all part of why affirmations might work. Maybe.

Well, here's another one. According to Canadian Affairs, I guess that's the publication, there's a senator in Canada who wants alcohol to have warning labels on it because, as the headline says, alcohol is poison. So, Canada might label alcohol as poison. He may not be using that word, but that's the sense of the story about it is that alcohol is poison.

Now, if you're new to me, you don't know that I've been saying for quite a number of years, alcohol is poison. It's a refrain that helped a lot of people quit drinking. They just have to hear those words, alcohol is poison. And it's based on the idea that human brains are really like AI and we're just programmed by the words that are most frequently repeated in our heads.

So if you say, "Hm, alcohol is a beverage. I sure would like to have a beverage." You're going to do a lot more beverage drinking than you're going to be doing poison drinking. So, if it seems like, well, that couldn't possibly work 'cause all you did is call it a name, say everybody knows what alcohol is. The fact that you're calling it a poison, how's that going to help me stop drinking? And the answer is because that's all it takes. The word that you most associate with it will reprogram you.

So, if every time you think of it or someone offers it to you, you say, "No thanks. Alcohol is poison." Most of you, not all of you, but most of you that would be enough to never have another drink again. It works because I hear all the time from people who used it successfully. You'll probably see a few in the comments.

And apparently Gen Z, as you know, is not drinking nearly as much as in past generations, but that's also worldwide. So, Germany is having a problem because, you know, they got a big beer industry there and the young people are turning away from beer and alcohol in general. Sure enough, now only 38 percent of men in Germany under 25 drink at least once a week. It used to be 55 percent a generation earlier and it was 85 percent in the mid-70s. In the mid-70s in Germany, 85 percent of the population had a drink at least once a week. 85 percent. And now that's down to 38 percent with men under 25. That's a big change. Wow.

If you're worried about AI affecting your privacy, well, I got a story for you. OpenAI says it scans user conversations with its AI and can report some of them to the police according to an article in Futurism. Now, the things they would report would be the obvious things like if somebody was asking how to end their own life, they might report that so the person could get help. Or if they were saying something like, you know, how to hurt people or, I don't know, make a nuclear bomb or create a poison or something like that that obviously is subjective. But if AI spots that sort of thing, it surfaces it to some humans and the humans decide whether or not that should be turned over to law enforcement.

To which I say that would really mean the AI is listening to everything you say and is using a filter to judge whether you should be getting a contact from law enforcement. That would really change the things I'm willing to use AI for because I always thought one of the great advantages of AI is that it wouldn't be censored in any way and that I could ask all of those banned questions. That doesn't mean I'm gonna do something. But sometimes you're just curious, you know, you're just curious about a domain that would be very bad if you were to take that action, but sometimes you just wonder about it. And I guess you'll get turned into the police if you wonder about the wrong things while you're in the presence of the AI.

And it might not even be something that you asked the AI. It might be just something you heard. You know, if you had it in voice mode accidentally and you said something on another topic to another person, it could just overhear it and then next thing you know, knock knock. So that's pretty creepy. I'm not sure they should do that. I don't know how to judge that one.

Well, as you know, Trump fired Lisa Cook, one of the F

Context —

ed governors, because she's accused quite credibly, and I don't believe she's denied it, that she did some mortgage fraud when she was a little bit younger. And she claimed two homes as her primary residence to get better rates, I guess, and that's illegal. So, she's fired, but I think she's going to fight it in court. And I saw a post by Eric Daugherty talking about how the experts were imaginin…

Next segment → →