I changed my mind about some things. These examples are illustrative of potential weaknesses of focusing on error correction and critical discussion like Karl Popper advised. I don't think the weaknesses are inherent or unavoidable. They're practical issues that don't require different epistemology principles to address. They're just ways you can go wrong if you don't know enough.

The errors involved expectations around other people as well as underestimating the amount of knowledge needed for things.

Sharing Evidence

I used to think if I said something to people, and they had evidence that I was wrong, they would tell me. I took a lack of any negative responses as indicating they didn't have key evidence that I was missing. That has implications. E.g. if there are a dozen reasonable people present (who are active forum posters who say they value rational, critical discussion) who have thousands of hours each of workplace experience, and I say "sexism is rare", and no one contradicts me, I thought that meant that all of their workplaces lacked blatant sexism. I now think I was wrong. Even people who are aware of sexism, in their opinion, often don't share their evidence.

In general, we all have a pretty limited amount of experience with the world. Many people only have much experience with a few workplaces which isn't a good random sample, so information about other people's experiences is important. But I now believe this type of information, based on people failing to share their experiences that contradict your claim, is highly unreliable.

To some extent, I was expecting people to share information that I would share or that Popperian epistemology says is valuable to share. I particularly had expectations if they said they were Popperians or said they had values like mine. I wasn't expecting people who weren't interested in criticism or rationality to share contradictory evidence; I was specifically talking on forums for rational discussion and debate. I also had low expectations for people who rarely or never posted anything or who were on a break; my expectations were mostly for people who were currently actively participating in discussions. It's common for a forum to have over 200 members but under 20 people actively posting this week.

Finding Good Arguments

A related issue is that I thought that you could look at the arguments made by a school of thought and then judge it accordingly. Now I believe that the arguments made by thought leaders are often bad even though better arguments are available. People, including authors, pundits and academics, will use bad examples and bad evidence despite much better stuff being available. They'll make low quality arguments while ignoring better arguments made by people on their own side which end up never being popular or well known. So the best arguments that exist can be hard to find, even if they're published and publicly available, especially when just reviewing a school of thought you disagree with and don't study in depth.

Overconfidence

The issues about sharing evidence and finding good arguments are two factors that helped lead me to overconfidence. It's crucial to recognize when the readily available information and discussion isn't good enough, and then to either be aware that you don't know much or do deeper research. It's important to have a sense of how much knowledge is needed to deal with an issue well, and to know you might not reach that bar even if you consider things said by public figures, professors, authors, experts, and other successful people with good reputations who work in the field and ought to have a lot of knowledge. Information available in internet discussions is also frequently inadequate.

My attitude was roughly that you take the current state of the debate – books, papers, essays, online discussion or debate from the best people actually willing to participate – and you evaluate that, and that's how you reach the best conclusions available given currently existing human knowledge. I overestimated how good current human knowledge is in some areas, but that's only the secondary problem. The main problem is a lot of good knowledge is hard to find and there's not enough productive debate taking place. Merit often doesn't rise to prominence.

You can read the books that people on a side of an issue recommend and still miss better information; you're trying to be fair by listening to what they say their best arguments are, but there are actually much better arguments that they don't know about. In a lot of cases, better knowledge exists but few people have it (including many "experts" don't have it). Often some good knowledge was created by someone but it never spread to most people. If you get information by talking to people and looking at popular recommendations (including reading books that summarize topics or review many ideas), you'll often miss good but unpopular ideas.

Another common issue is that people can't tell which literature and arguments on their side are good or bad. So they might recommend twenty books, one of which is good, and nineteen of which are bad, but they don't understand which is good or why. So even if you check several of the books they recommend, you could easily miss the good one. Often, they'll just tell you one or several books, not twenty, so there's a good chance they won't even tell you the name of the good book, even though they're familiar with it and like it. Either way, whether they give you a long list or narrow it down for you, it's easy to miss a good one that they don't recognize as superior to the rest.

It helps to read citation chains: look up multiple sources that a book or paper cites, then for each of those look up multiple things it cites, and keep following citations back to early work repeatedly. This is one example of the kind of much higher effort research that can be more effective. But it's still not enough: lots of good work isn't receiving citations.

These problems affect all popular positions on all sides of all controversial issues.

Pesticide

I missed some important knowledge that is well known and popular. I thought Silent Spring was a bad book. I believed secondary source summaries criticizing it. I put too much trust in people who were either dishonest and irrational or else were themselves misled by secondary sources. This is a hard issue because I don't have time to personally review every book that I hear is bad. In this case, I managed to fix my mistake myself. On my own initiative, I read Silent Spring and, despite my skepticism, after a few chapters I discovered that I liked it. I try to sometimes read things I expect to be bad to check whether they actually are bad. I try not to only read things that I agree with or that people similar to me recommend. But I was more than five years too late to actually get a response from Alex Epstein about Silent Spring. Now it turns out that I can't really get anyone to debate me about it.

When I thought Silent Spring was a bad book, I couldn't get any useful criticism of my position or productive debate from fans of the book. And now that I take the opposite position, I also can't get useful criticism or debate. The underlying theme here is it's hard to get good feedback about errors or disagreements, and productive debate is scarce. Judging the pro-Silent Spring people for not debating would be a mistake because the anti-Silent Spring people also don't debate. I do debate, but I'm the weird exception and my willingness to debate shouldn't lead me to conclude that the side of an issue I agree with is open to debate; I should check whether other people besides me will debate (preferably people who aren't my fans and aren't on my forum).

So I was wrong about pesticides and DDT (even if it turns out I'm wrong now, and they're somehow good, I was still also wrong before because I didn't know enough to refute Silent Spring). I was also wrong about nutrition. I believed a bunch of things my former mentor, David Deutsch, told me about food (and about environmentalism; he's on the anti-Silent Spring side). I was open to debate and error correction about nutrition but that didn't lead to me being corrected. I only got better ideas about nutrition after finding and reading some books and papers that aren't the popular, mainstream recommendations. It took a lot of work to find less well known knowledge and learn about it.

Misogyny

I was also wrong about feminism, sexism and misogyny (and also racism). My mentor David Deutsch, like many right wingers, is the type of person to deny that sexual assault is a common problem today. I thought that, if it was common, I would find out from debate, from discussion, from being open to error correction, and from hearing and engaging with some arguments from the other side. I did read some books, have some discussions, consider some criticisms, and so on. But I was missing a lot. It turns out that many women have compelling stories but that evidence wasn't reaching me. I've now seen a lot of evidence primarily on Reddit and TikTok. I'm now convinced that sexual assault is rampant, misogyny is present at most workplaces (from coworkers, bosses, and systemic company policies), there is a gender pay gap (and a racial one), and that pick up artists (PUAs) like Mystery encouraged sexual assault and misogyny.

As with many people, Mystery's misogyny mistakes were partly unintentional but not done with full innocence. There are tons of other people who are bad too, and many are worse, so I don't think Mystery should be singled out; I just brought him up because I liked some of his work. I still think his attempt to study social rules with a semi-scientific attitude, and talk about them explicitly, had value. Mystery-era PUA wasn't as bad as the manosphere and "red pill" content that's much more popular today than old school PUA ever was (the manosphere has been flooded with late adopters). A lot of the manosphere stuff also is a lot more blatantly and intentionally misogynist than the misogyny of Mystery or of most internet posters in 2000.

One of the types of PUA misogyny I didn't recognize well enough was how much PUAs were applying social dynamics ideas only to women (and dating), not to men (and other contexts like job interviews and office politics). I always applied the social dynamics concepts to men and to other contexts, and to me the broader applicability was a major part of the appeal. In retrospect, I think applying some of the PUA social dynamics ideas to men on my forum offended them: they intuitively felt like I was calling them irrational women.

Why didn't I get social dynamics ideas from somewhere else that wasn't focused on sex and dating? Because I still don't know where to find enough knowledge elsewhere. Useful, explicit discussion of social rules is unusual, especially with significant counter-culture, anti-mainstream themes. You can learn about social dynamics from psychologists but that has various advantages and disadvantages rather than just being better.

Some of my old posts use the term "red pill", so I also want to say that I think "red pill" (or "manosphere") in 2026 is awful. It was always flawed, but the meaning of the term changed over time and got worse. I guess there used to be multiple different meanings of "red pill" in different internet subcultures, and now red pill and manosphere are a larger movement with more of a clear, bad meaning.

I'm interested in social dynamics ideas from other sources when I manage to find something that I think is good, like alimcforever's idea and analysis of low talk. I've also liked some of Eliezer Yudkowsky's discussion of social dynamics, but there isn't a lot; he focuses more on other topics like Bayes, AI and rationality. I do think there is some good information about social dynamics mixed into his books Inadequate Equilibria, Rationality: From AI to Zombies, and Harry Potter and the Methods of Rationality (disclaimer: his books contain some terrible ideas, too). Yudkowsky approaches social dynamics more from a theoretical point of view with logical analysis, whereas PUA and alimcforever approach it more from experience. I think both approaches have value.

Another place I've found useful analysis of social dynamics is neurodivergent content on TikTok. People who identify as autistic or similar will sometimes explicitly discuss social rules because they've been trying to learn the rules in a fairly explicit way as an adult rather than learning them intuitively as a child. Sometimes they're trying to figure out the social rules so they can get along with society better and sometimes they're critical of social rules.

Right Wing

All large political groups are super flawed, and I've thought that since shortly after meeting David Deutsch. You can't just stay away from the flawed stuff because then you'd never engage with anything. Deutsch basically told me that it goes without saying that everything has lots of flaws and that you have to focus on the positives to get value where you can. He even applied this to Ann Coulter: he encouraged me to focus on her good points. I think that was bad advice that actually affected my life. I read a bunch of her books, which I thought had some good parts, but now I think she intentionally lies sometimes and I don't know which facts to trust from her books. I did make multiple efforts to check if her information was true, search for criticism of it, and fact check her citations (which on average are noticeably better than most authors) more than once. But I now think that checking was inadequate and she's deceptive and manipulative sometimes and (like many well known smart people) uses her cleverness and scholarship skills to help support her deceit.

I think some of Coulter's bad points (like misogyny, homophobia and attacking evolution) are red flags that can serve as useful warnings. In retrospect, I think Deutsch was wrong to advise me to ignore them and just focus on the parts where she appeared to be using facts and logic.

I now think Deutsch was always a biased right-wing tribalist and he fooled me in the past with his Popperian talk about rationality. I withdraw my past recommendations for right-wing material that Deutsch got me to read like Coulter and Frontpage Magazine. Sorry.

This retraction doesn't include Objectivism, Austrian economics, and other material which is closer to being classical liberal than right wing. I remain a fan of classical liberalism, and I think it should be emphasized more that it's not actually right wing. I think Ayn Rand had various flaws including misogyny and bad ideas about sex, but I still like her overall.

As one more related comment, I think Deutsch's non-aggressive Christianity and Judaism have significant value type of atheism had some good points, and I think one shouldn't be hateful towards religion, but I also now think Deutsch partly held that view because of his right wing biases. He was dismissive of all other religions, including major religions like Islam, Hinduism, Buddhism. He was only actually friendly towards two specific religions (the same two that the Republican party likes).

Cars

I recently heard the idea that the U.S.A. is car-centric, without enough use of trains, because of systemic racism. I thought that was plausible but would take too much research to actually reach a conclusion about (and it'd be basically impossible to get useful online debates about it, or to put much trust in most things I could find to read about it). I think I was always open to this kind of idea (given a few paragraphs of explanation which I'm not including here) but in some sense I was waiting for it to come to me, or be popular, instead of searching it out in effective ways (which is hard). I've always done a lot more than most people to seek out unpopular ideas, but there was (and no doubt still is) room for improvement.

Israel

Another thing I may have been wrong about is Israel and its violent conflicts. My former mentor David Deutsch taught me that Israel is in the right. He was actually quite aggressive and pushy about that early on in our discussions. I think that was poor mentorship: even if he was right about it, it wasn't a topic I needed to learn about at that time (or maybe ever). It would have been better for me if he'd focused on epistemology more in our discussions.

I have some skepticism of the morality of Israel's violence now, but I haven't gone back and researched it enough to say much. I'm trying to be more humble and recognize that it's complicated and I don't know enough about it. I used to think my participation in political debates, engaging with people on both sides, and reading materials on both sides (pro-Israel and anti-Israel articles and books) was adequate to reach a good conclusion. Now I think that was inadequate. Regarding Israel's war that started on October 7, 2023, I have some major concerns about it, and I also have major concerns about Hamas and Hezbollah, and I know that I don't know enough about it. Finding out enough would be a ton of work, in part because I couldn't rely on a lot of accessible resources and discussion that I no longer trust. And this isn't a priority for me to research.

I also dislike how a lot people on the left pressure social media content creators, who don't know much about it, to speak out against Israel. In general, I think people should be respected for knowing their limits, recognizing their ignorance, and being neutral on issues, especially issues that aren't directly part of their life. I wish people would have more respect for studying issues and being thoughtful, and also for having the humility to know that they don't know about a topic. Instead, a lot of respect is given for being on the same side as someone, being in the same tribe, supporting the same conclusions as them, like it's about cheering for the same sports team rather than an intellectual issue where you ought to understand things.

Note: I think what I said here is capable of seriously offending many people on both the left and the right. There are substantial incentives not to say it. I think that's a big problem with debate and free speech: too many people are way too intolerant. Nuanced, neutral or modest (admitting ignorance) views can get you hated by both sides, so a lot of the smarter people with nuanced views don't want to talk in public, which is one of the reasons it's hard to get high quality debates. In general, I think most people should have fewer strong views, but they're pressured into taking sides and joining a big tribe.

TCS and ARR

My former mentor David Deutsch was a founder of Taking Children Seriously (TCS) and of Autonomy Respecting Relationships (ARR).

I changed my mind about TCS and wrote criticism of it. I also changed my mind about the wisdom of polyamorous ARR relationships. I did criticize polyamorous relationships early on after Deutsch advocated them to me, but my views have also evolved more since then.

Corporations and Capitalism

I changed my mind about what big companies are like and about how capitalist, rights-respecting and law-abiding our society is. I wrote Capitalism Means Policing Big Companies. I lowered my opinion of billionaires in general. And I lowered my opinion of anarcho-capitalism. I see errors in the anarcho-capitalist literature that I don't want to associate with.

These changes go against ideas I learned from my former mentor, David Deutsch, who is a libertarian anarcho-capitalist.

Note that I did not change my core values or principles regarding freedom, rights, non-aggression, limited government, peace, social harmony or abstract capitalist economic theory.

Willing to Change my Mind

I think, for each of these issues, I was mostly rational in a key sense but many of the people around me (friends, debate partners, authors I read) were less rational than me. The key way I was being rational is that if I found out about new evidence or arguments, I was willing to change my mind without a lot of resistance. After changing my mind, I've witnessed other people heavily resist change when encountering a lot of the same information that changed my mind. So they were less open to error correction than I was. I didn't know in advance that they'd be like that, and it's relevant to why past discussion with them was less effective at learning new ideas than I'd expected. Their resistance means some of them probably saw or found evidence and arguments in the past that would have changed my mind, but instead of telling me or changing their own minds they were dismissive. That's relevant to my estimates about what evidence and arguments exist and how available they are.

I'm not going to go into details on all these topics because I don't want this to turn into a lecture where I try to tell you the right answers on these hard topics that are outside of my expertise. I don't have the time to research every topic adequately! I've been writing primarily about epistemology and rationality for the last few years, and I purposefully brought up philosophical themes in this essay too. But I'm still open to some discussion about these other topics on my forum. If you think you have adequate knowledge and you want to share it, you can go ahead; I'll ask questions and share criticisms and if your knowledge meets my standards I'll appreciate the help.

Rational Learning

In some sense, I thought you could use the methods of rationality to organize existing knowledge and debate and reach good conclusions. But now I think a lot of the inputs to that process are too biased and problematic, and a lot more work is needed to have effective knowledge. Just getting a list of the arguments and evidence, and then evaluating the right conclusion given that list, isn't a good enough approach because a lot of important knowledge won't make it onto the list. It's still a good, useful skill to be able to do that, but better search strategies are also needed, as well as better research.

Some areas have never had anyone competent do much work, so there's still room for a lot of new knowledge to be created quickly. Focusing on learning existing knowledge makes more sense for areas where a lot of people have already tried really hard and found almost all the ideas that are easy or medium difficulty to create, but I now think few areas are actually like that since there aren't enough competent people doing productive work.

Misquotes

Also, overall, I've become less trusting of what other people say. People lie about factual matters more than I realized. People give more inaccurate summaries about ideas they dislike than I realized. And they misquote. I didn't used to know that David Deutsch put misquotes in his books and academic papers. Now I know that lots of intellectuals do that, so it's hard to find anything to read where you can trust much, even quotations or cited facts.

Deutsch gave me the false impression that putting misquotes in books was socially unacceptable, that he would never do it, and that criticizing misquotes would be received well. And he communicated similar things about other issues like factual and logical errors. That was all false.

Astrology

Also, the better I get at thinking, the less I can expect other people's work to be the same quality I would do. So, for example, I've never researched astrology. People like David Deutsch say astrology is stupid: basically just an unscientific mindset combined with Barnum statements. But I don't think I should trust the haters without doing any of my own research. I've never read a pro-astrology book to check whether astrology is being misrepresented by its opponents. I've certainly never done deeper research to see if perhaps there are some more reasonable astrology ideas mixed in with some more popular dumb ideas. I don't consider astrology a promising lead and have no plans to research it. I've always been open to debate about it but I don't recall ever having a debate about it. I think I shouldn't be aggressively anti-astrology since I don't really know anything about it. I should just not bring it up in general and I generally shouldn't use it as an example to attack.

This is just one example where I'm trying to recognize my inadequate knowledge for making confident statements. I purposefully used an example (astrology) where a lot of people are extremely dismissive, and I would have been more dismissive in the past. I'm saying that even for astrology you shouldn't just trust the haters and secondary sources without doing any of your own research. You should also be aware of your own lack of substantial knowledge for many other topics too. Any topic where you're less confident about jumping to conclusions than for astrology is a topic where more research or acknowledgment of your ignorance may be warranted.

Approximations

Thinking of issues in terms of sides or tribes is not ideal but is a useful approximation for writing in a short, understandable way. Similarly, my language like "good arguments" is short, loose, understandable speech, not advocacy of indecisive, weighted factor epistemology. Using concepts familiar to our culture is important when writing so that you can focus on a few issues without arguing about fairly-irrelevant off-topic issues that most people would find confusing and disagree with (they're the kind of topics that require their own essay or book to explain well, not something suitable for explaining in a one paragraph note). For more information about my epistemology ideas, see Critical Fallibilism.

Conclusion

I've raised my standards and decided that a lot of knowledge isn't good enough, such as the criticisms of Silent Spring I knew about. Good ideas can be hard to find even when they exist. Thought leaders often don't talk about the best ideas on their own side. In discussions, people often don't share important, relevant evidence with you, even though they have the evidence and aren't blocked from sharing by privacy concerns. Methods like debate and looking at the other side's arguments are still good methods, but they're less effective than I thought. There aren't easily available better methods to switch to. So be more humble.