[danShumway]

There are two possible dangers that you can fall into with science. One is to assume a conclusion with such force that it causes you to avoid certain questions or to suppress certain research. The other danger is to be so fascinated with doubt or so averse to a particular outcome that you ignore consensus, and place less weight than you should on the obvious conclusion that research is pointing to (and yes, I think the author is guilty of this).

I have been hearing for a long time about how scientists are scared to question climate change, or gender differences between men and women. From what I've seen, the reality is that there are pretty clear conclusions that we can draw in those areas: that climate change is human-caused and dangerous, and that purely biological, mental differences between men and women are usually overstated.

Frankly, I don't think that there's a particularly strong culture in science that is scared to ask those questions, and while cancel culture is a real thing, I don't think it's a real problem here. I do think there's a culture at the edge of science that doesn't like the answers researchers have found, and that is aggressively underestating the degree of confidence in those answers in the hopes that certain debates can be prolonged forever rather than used to influence policy changes right now.

In other words, pervasive doubt can be just as dangerous and just as politically motivated as pervasive certainty. There are certain topics (such as climate change) where we effectively know the right answer. Of course we never reach 100% certainty, of course there are areas where we want to learn more, but we're a heck of a lot closer to 100% certainty than we are to 50% certainty.

Because of that, some of the taboos you notice on the far-edges of the scientific community are actually justifiable defensive measures -- because running the clock down while introducing impossible standards of certainty is an effective strategy to circumvent scientific and social progress. We saw this happen with the sugar industry, we saw it happen with the tobacco industry, we saw it happen with race science, and we're seeing it happen today with gender studies, climate science, and anti-vaxers.

If you were to look at the subjects I just listed to pick out a common theme or lesson, I would say that it is, 'beware isolated demands for scientific rigor, particularly when those demands are selectively applied in ways that benefit a political or socioeconomic status quo.'

[mikedilger]

Scientific and social progress should continue with the near certainties we already have -- I think we are in agreement here. So long as we don't cancel contrarian scientists. We need to accept their long-shot research and be grateful for it, because we cannot predict ahead of time which one of them will turn out to be correct, as long as they couch their statements when they present them publicly. As a defender of free speech, I think that if some people want to believe in the long shots, they are making that ill-advised decision for themselves. When it comes to anti-vaxxers, they shouldn't have that choice as the effect is upon all of us. If a voting majority choose to believe the ill-advised long shot research then something is definitely broken, but it's not the existence of long-shot research. I won't pretend to know what that 'something' is.

[thu2111]

From what I've seen, the reality is that there are pretty clear conclusions that we can draw in those areas

The problem is this is circular: you reached your conclusions based on the output of the academic system (presumably).

I used to think like this, that if most scientists agreed on something it was very likely to be true, that mistakes by whole fields were exceptionally rare and remarkable events, and that groupthink wasn't very powerful.

Over the years I've been faced with evidence that I was wrong about those things, over and over again. Now I think if most scientists agree on a topic where you can't run extremely rigorous experiments then it's quite likely to be wrong, that mistakes by whole fields are very common and groupthink is extremely powerful.

I do think there's a culture at the edge of science ...

Science is almost entirely funded by governments. In some fields there's more diversity than others, but it's notable that the fields that seem to have the biggest problems and most controversies attached are the ones where there's little private sector involvement. If groupthink sets in, and due to the lack of any feedback loops from outside the academic system that is clearly a problem in academia, then almost by definition anything that's taboo will be at the "edge of science" even if it's right.

[danShumway]

Maybe background plays a role here.

My perspective is that I used to think like you growing up as a Creationist, and I can't see a major difference between the arguments OP is making and the arguments that were made to me by members of my church. I used to be highly dismissive of group consensus, I used to say that if you couldn't explain your point in a way that convinced me in specific, that I didn't care how many people agreed with you.

I now think that consensus is an important metric that should be at least considered, particularly in areas where I am not an expert.

Of course, we've seen areas where the broader scientific community was wrong. Before race science became a fringe argument used by fringe segregationists, it was a generally accepted conclusion. But again, underconfidence is just as dangerous as overconfidence. You can't pick one, you have to reject both.

> you reached your conclusions based on the output of the academic system (presumably).

At the end of the day I have to base my conclusions on something, and I think basing them on scientific output is a better place to start than basing them on ideology. Even from the much more subjective metric of "do I trust the communities who argue for this", climate-science doubters and gender-science doubters don't come out looking well.

During the early rise of neural networks, where consensus was that this was extremely promising for AGI, I could find neutral, intelligent people that I respected who disagreed with that conclusion. That made me feel much more confident rejecting the general consensus myself. On the subject of gender-science, I can't find the same number of quality defectors. To me, the people saying that race/gender are heavily tied to IQ look a lot like Creationists, and I ain't getting pulled into that trap again.

[thu2111]

Yes, I suspect it's different backgrounds. I was never exposed to creationism as a child.

I'm curious what led you away from creationism. Was it purely a matter of observing most people aren't creationists and deciding to go with the crowd? Or did people argue with you about it (perhaps implicitly via things you read or watched) and you came to realise the arguments didn't hold water?

Given no competing evidence against it, I'm happy to go with the apparent consensus on a topic, especially if it doesn't matter to me. Where there's disagreement combined with a question that matters though, argument-by-apparent-consensus doesn't do it for me anymore.

Partly this is because of what I saw in the Bitcoin community. That whole community lost its mind after a small minority of people hijacked the communication channels, started deleting any posts they disagreed with and loudly insisted that their (crazy, unsupported) views were actually the scientific consensus, that anyone who disagreed was either uninformed or - when they couldn't quite get away with that argument - simply in a tiny minority and should be ignored. None of that was true but because concepts like "consensus", "intellectual minority", "edge of science" etc are totally subjective they were able to effectively create in people's minds that belief despite that it was false.

I see a lot of that sort of behaviour in academia now. There are people who say, wait a minute, does that claim stack up? And in response what they get is "Shut up, you aren't qualified to have an opinion, 97% of scientists all agree so they can't be wrong". Such claims of consensus usually fall apart when examined, but you can't get the word out because those same people are doing everything they can to silence disagreement.

That's why I don't think we can really trust much academic output. The signs of groupthink are all there. Note: I distinguish between science and academia. Lots of great science is done by corporations, e.g. in the field of AI. It's the institution of academia that has the problems, not science as a concept.

To me, the people saying that race/gender are heavily tied to IQ look a lot like Creationists, and I ain't getting pulled into that trap again.

Look in what way? The people I've seen say that are all scientists or people quoting them. I don't like these conclusions either, because I'd like to believe my own intelligence or IQ is related to hard work and not DNA. As would everyone! But I can't just blow the people off who have research showing these things because of how they look. Surely that'd make me the whatever-ist?

[Hercuros]

I think the answer to this is not to start having general feelings of doubt about generally accepted conclusions in science. Rather, we need to be much more specific about where we think that groupthink is playing a role, looking at things on a case-by-case basis, rather than casting doubt on everything without justification. We also need to be extra careful with accusations about groupthink in fields where there is a political or economic motivation for making such accusations.