Coffee With Scott Adams — Knowledge Archive October 9, 2026
Scott Adams Philosophy Archive
Search ideas
Episodes › Episode #2402 › Segments
NewsReaction Media & Fake News

Back to episode — Episode 2402 CWSA 03/03/24

Context —

Hispanics it's voters. Is that possible? Let me make a prediction. I'm going to make the most radical prediction but I'm not going to predict it as a certainty for the first time ever and only because of the special situation with Biden falling apart and the border being open under that weird special situation. Trump could win the black vote outright. I'm not going to predict that as in like a cer…

← Previous segment →

ld take it from let's say a pure manager perspective. You know if you knew at the moment it was better than the alternatives I'd probably do it and try to figure out something later. But here's my question. Since you cannot use math and economics to support limiting natural gas because it's a questionable proposition economically and even CO2 wise would you say that the Biden administration is following the science? Their entire thing is follow the science. So don't you think that they owe us an analysis perhaps a model for climate change? They like to use models. They should produce a model. They should. Here we are cutting the natural gas and here you can see how this follows through to reducing this line on CO2 if that's their argument. And their whole deal is you got to be compatible with science. Why don't we hold them to their own standard and make them compatible with science? Well it's because they've been gaslighting and they've never been compatible with science because you know what is not a science? Models. Everybody who thinks that a projection model is science has really been brainwashed. It's closer to the opposite.

Do you think that I could create a model or here's a challenge for you. Would you bet a million dollars against my claim that I can create a model that will predict the GDP of America in 10 years? So that's my claim. I can create a prediction model that'll hit the GDP of America 10 years from now on the same dot. Who would bet against me? It's easy. Do you want me to tell you how I could do it easily? And I know you're thinking hey nobody can predict the future like that. I could create a model that I would bet a million dollars will be right. Do you know how I do it? I would create a thousand models. Every one of them would be off by a tenth of a percentage point and then I would give you all thousand models and I'd say in 10 years I'm going to tell you which one of those was right and then I'm going to claim that I created a model that predicted it. And you know what it would be true because that would be the model. You had it for 10 years. I'm pointing to it. It's the exact number that is the reality. There you go.

Now you're saying but Scott that's not like climate change models. No it is. That's how they do it. You didn't know how they do it. They have hundreds of models and they do something called hindcasting which is making sure that it would have worked if it had existed in the past up to today. Do you know how they do that? They make so many models that some of them by coincidence will be in the range of what is reality and then they tell you and here's the big scam that therefore it can predict the future. Nope. Think about my thousand models about the GDP. One of them was right. Can it predict the future? No it wasn't right because it was right. It was right because there were a thousand of them and one was going to be pretty close. I threw away everything that wasn't within that range of something it coul

Context —

d have been. How many of you did not know that you can make the models do anything including match the past? It has no predictive value. None. Yeah some of you do but if you don't do this for a living or you don't listen to somebody like me who did it for a living which is projecting the future as funny as that is you would never know. If you're an ordinary person you don't know that a meta-analy…

Next segment → →