I have a more positive opinion of scientists than you seem to, and I'd like to explain why. A big part of the problem here is selective media reporting of scie

I have a more positive opinion of scientists than you seem to, and I'd like to explain why. A big part of the problem here is selective media reporting of science. All fields have good practitioners and bad practitioners. ("bad" here can mean overconfident or bad at stats or outright fraudulent) As the parent article points out, in fields where objective feedback is less easily accessible, it becomes easier and easier for bad practitioners to survive. Many media sources will report only the opinions of scientists that support the positions they agree with. And the fields with the least objective feedback are often the very same fields that the media has strong opinions about, meaning that they can easily find somebody supporting their position. Even well-intentioned media sources can have difficulty with mistranslating the results or misinterpreting the confidence level of some piece of research. I think the reason I have a more positive opinion is because I more often tend to look at primary sources: papers, or at least blog posts by researchers. That kind of stuff is often linked on HN, which is a big part of the value of this place. Looking at primary sources makes it harder to tell what's important, and what's just minor details, but the advantage is it allows you to see more of what's actually going on, instead of just getting the media's view of "the scientific consensus". Of course you still have to watch out for the replication crisis and all that, but it's easier to tell if the scientists did their stats right when reading their paper than to try and guess based on a journalist's write up of the results. I guess the problem I have with your "permanent cancellation" proposal is that it doesn't allow for uncertainty. Have a groups of people try to predict the outcome of 7 coin flips in a row, and less than 1 percent will remain, along with maybe Persi Diaconis. You need to have some way for people to say "there's a 50% chance of heads", without getting fired. I agree that we should still keep track of people's predictions, and stop listening to them when they've been confident and wrong many times. That kind of "prediction checkup" is important, and I wish that it was more mainstream to publish checkups of that kind. Expecting people to keep track in their heads means that most of them won't bother, and will end up listening to the same doofuses over and over.