1) Learning to WRITE well is a primary way people learn to THINK well. There are other ways to learn to think well (e.g., a strong culture of oral debate and rigorous discussion), but that’s largely not how things are set up, so without writing, there’s a vacuum. Until schools change, students are sacrificing learning to think.

2) Normalizing cheating in one domain normalizes it in other domains too.

There are lots of ways to use AI to improve your thinking (e.g., ask an AI to critique a belief you hold or to help you explore points on all sides of a debated issue). But when a teacher says, “Write this without AI,” and you have an AI write the essay, it’s preventing you from engaging in significant thinking.

Thinking well involves a number of components, such as:

– gathering evidence
– considering arguments
– formulating a viewpoint
– honing your viewpoint
– presenting your viewpoint clearly

Replacing thinking with AI is not analogous to replacing doing multiplication with a calculator. That’s a memorized algorithm. Thinking well, on the other hand, is core to understanding the world, figuring out what goals to set, not being duped by others, and many other essential aspects of life.

This piece was first written on May 23, 2025, and first appeared on my website on June 5, 2025.

Here’s what I’ve found to be useful:

1) Shrink the group

The best group conversations typically happen (I claim) in groups of 3 to 5. If your group is bigger than that, you can split it into smaller groups (e.g., by engaging with the people closest to you). With 7, it’s hard to get a word in or stay on any thread.

2) Introduce a topic

If the group seems to be bored, an easy way to steer the conversation is to wait for a lull and then say, “I’d be curious to hear what the people in this group think about…” and then raise a topic that’s likely to be very interesting and relevant to folks there.

3) Invite an exceptional conversationalist

Some people are very good at keeping everyone engaged. If you can, bring someone like that along who can make the discussions more interesting for everyone. Socially skilled extroverts who really care about others having fun are a good bet.

4) Leave out or inform the obliviously over-talkative

Some people talk a lot but don’t pay attention to whether others are interested (e.g., extroverts who are narcissistic or who lack strong social perception). Just one can ruin a large group conversation (even if they’re fine in other settings). When they are well-meaning and are likely to take the feedback well, you can point out this tendency and hope they improve. If they are unlikely to want to hear the feedback or unlikely to take it well, you could simply not invite them to events focused on group conversation.

5) Encourage interesting introverts

Some people have very interesting things to share but are quiet in group settings due to shyness or not wanting to draw attention to themselves. Involve them by asking for their perspective or directing questions to them (if you think they wouldn’t mind).

6) Keep things on track

Sometimes, a group conversation will be interesting, but then, suddenly, it gets derailed by someone’s stray comment that leads the group in a random direction. Consider jumping in to make a comment or ask a question related to the prior thread to return to the more interesting topic.

I hope some of these ideas help you improve your next group conversation or dinner party!

This piece was first written on November 3, 2024, and first appeared on my website on November 7, 2024.

Use my lists of “Idea-Inducing Questions” to generate nearly endless ideas to write about, think about, or discuss!

Questions about learning and truth-seeking

• Recently learned: what’s a powerful idea, concept, or mental model that you’ve been learning about recently that you think is worth knowing?

• Changed opinions: what is a strongly held belief you used to have that you changed your mind about? What caused you to change your mind? Why do you think you were wrong before?

• Influential book: what is the name of one book that substantially influenced the way you think about things? What did you learn from it that you can pass on to your audience?

• Debates: what’s something you disagree with a certain group of people on? What do you think is the core of the disagreement?

• Thorny problems: Is there a complex problem, situation, idea, concept, or a set of competing ideas that you’re still trying to understand or figure out your opinion on? What are the factors that are driving your opinion in different directions or that make the issue tricky to figure out? What are the open questions or confusions you have about it still?

• Third perspectives: For any pair of opposing ideas that most people in the public sphere take either one side or the other on, can you think of a third perspective or synthesis of both ideas that could actually be better than taking either side?

• Misconceptions: What’s a commonly believed idea that you think is actually wrong or a misconception?

• Underrated or overrated ideas: What’s a powerful or useful idea that you think is significantly underrated? Or conversely, what’s an idea that is talked about a lot in a positive light that you think is overrated or that isn’t actually a good idea?

• Epistemics: how do you think about what to believe versus what ideas to reject? How do you approach understanding hotly-contested, thorny, or highly complex topics? What mental models or approaches do you use to help you think more clearly or analyze questions or evaluate evidence?

Questions about ideas that have useful applications

• Beneficial ideas: what’s an idea that, if it became widely known and adopted/used, would greatly improve the world?

• Versatile ideas: what powerful idea or concept do you think has many different useful applications across many life domains?

• Psychology tools: what’s a powerful idea, concept, mental model, or tool from psychology that you think is useful to people’s lives?

• Tools for making sense: what’s a powerful idea, concept, mental model, or tool that you think can help people better understand or make sense of the world?

• Scientific principles: what principle from a mathematical or scientific field (e.g., economics, statistics, evolutionary biology, etc.) do you think is important or valuable to know about (because it helps you understand the world or because there are applications of it to daily life)? How can you apply this idea in life?

Questions about your own ideas and experiences

• Your ideas: What’s an idea you’ve come up with that you think would be valuable for your audience to hear about?

• Ideas you’ve applied: What’s an idea that you’ve found to be very useful or powerful in your own life? How have you applied it successfully?

• Unique experiences: what’s something you have experienced that very few people have experienced (whether it’s a good thing, a bad thing, or just something strange or surprising)? What did you learn from that experience?

• On your mind: What’s an idea you’ve been thinking about a lot lately? What are your current thoughts about it?

• Area of expertise: what topic are you very knowledgeable about? What is the most valuable idea from that field that you think many people would benefit from knowing about?

Questions about unusual or neglected ideas

• Powerful obscure ideas: In your opinion, what is one of the most valuable or important ideas or concepts that most people don’t know about?

• Contrarian ideas: What’s something you disagree with most smart, educated people about (according to your own definition of smart and educated)? Or what’s your answer to the Thiel question: “What important truth do very few people agree with you on?”

• Overlooked areas: What’s a topic area that very few people have an opinion on or knowledge of, that you think it’s important to have an opinion on, or that is well worth learning about?

This piece was first written on August 2, 2020, and first appeared on this site on May 27, 2022.

It’s tough to make real-life disagreements productive, but here are my favorite techniques for making it easier to do so. These help more if you are significantly motivated to use the disagreement to deepen mutual understanding of the issue.

I’m assuming here that you have control over your own behavior, but not over the other persons, because that’s the reality of almost all such situations. I’m not assuming that the other person is motivated to figure out the truth (e.g., they may be mostly motivated by “winning” the debate).

TIPS FOR HAVING MORE PRODUCTIVE REAL LIFE DISAGREEMENTS

(1) Avoid debates in front of others – if others are watching, it tends to create unproductive social dynamics. You or the person who disagrees with you may want to look good in the debate in front of those other people, which tends to push in the opposite direction of being open-minded and conceding when the other person has made a good point. What’s more, having other people involved increases the chance that the conversation gets derailed.

UNPRODUCTIVE: “Bob, what do you think about what we’re discussing?”

PRODUCTIVE: “Want to sit over there with me and discuss this in more detail?”

(2) Stay on topic – very often, the topic of disagreement will start to drift. Resist the urge to go follow tangents or switch what the debate is about, or you’ll lose focus. If someone says something new that you disagree with, resist the urge to address it if it isn’t related to the main point of contention.

UNPRODUCTIVE: “You’re wrong about that too.”

PRODUCTIVE: “We can discuss that point of disagreement later, but I’d be interested in returning to the subject we were discussing if you don’t mind.”

(3) Investigate first, don’t attack – when someone says something we strongly disagree with, we can be tempted to immediately attack their position. However, doing so sets up the discussion as a war. If the goal is to have a productive debate (where both you and the other person have a reasonable chance of making your views more accurate), then it’s much better to set the conversation up as a discussion where you can learn from each other instead. Avoid immediately jumping into attacking the other person’s position so that they aren’t put on the defensive and so that the conversation isn’t framed in terms of winning. This relates to Julia Galef’s concept of Scout Mindset (see: http://bit.ly/2fPYRJI), which I think is the right mindset to have when entering into a disagreement. You should be trying to understand the lay of the land, not firing cannons. If you fire cannons, you can expect cannon fire right back at you.

UNPRODUCTIVE: “There are at least three reasons you’re wrong about that.”

PRODUCTIVE: “Could you tell me more about that?”

(4) Clarify the other person’s claims – it’s surprisingly easy to have a long disagreement without fully understanding what the other person really meant. Typically you’ll need to ask a lot of clarifying questions early on so that you really understand what’s being claimed. If you refute the other person’s points, without having clarified, there is a good chance that you are arguing against something that isn’t quite what they believe. One of the best ways of making sure you’ve understood the other side is to repeat back the other person’s claims in your own words and ask if you’ve accurately reflected what they are saying.

UNPRODUCTIVE: “But X is not true because…”

PRODUCTIVE: “I think that what you’re saying is X, am I understanding that right?”

(5) Clarify definitions – language tends to be quite ambiguous. If the points you are making hinge on someone understanding your definition of a word, then take a moment to clarify what you mean by it. Or, if you notice that the way the other person is using a word may not be quite the way you use it, stop for a moment to explain your usage and to ask them to clarify their own. If you and the other party mean different things by word, it’s very hard to constantly keep that in mind without getting confused. It’s typically much better to either decide to switch over to the other person’s definition (once you’ve asked them to explain it) or to ask to switch to another word entirely that you both agree on the definition for. It’s a lot harder to get someone to successfully switch over to your definition than to simply resolve to switch over to theirs.

UNPRODUCTIVE: “We obviously don’t have free will because we don’t choose the things we fundamentally want.”

PRODUCTIVE: “Could you elaborate on what you mean by the phrase ‘free will’? I’m not sure I understand what you are using that phrase to mean.”

(6) Identify reason(s) for belief – don’t just stop at understanding what the other person is claiming. If you actually want to make progress in the debate, you’ll need to dig into the reasons behind their beliefs. Knowing why they believe what they do may cause you to agree with the other person more (because it helps you understand their reasoning) but, even if it doesn’t, it can also help you understand where the debate needs to focus in order to make progress. Without knowing why they believe what they do, you don’t know what points are most critical to discuss.

UNPRODUCTIVE: “I don’t think X is true because….”

PRODUCTIVE: “What are the main reasons that you think X is true?”

(7) Gently probe the reasons you’re given – if someone tells you that the reason they believe X is Y, don’t assume that if you were to change their belief in Y, then belief X would change too. First of all, we are not always aware of the reasons we believe things, so while the person might think Y is the reason they believe X, they might be wrong (it could just be their best on the spot guess or the first argument their brain could put together – they may have even forgotten what caused them to believe X). Second, our beliefs may be propped up by multiple other beliefs, so even if the other person really does believe X because of Y, there may be other beliefs P and Q that would need to change at the same time for the other person to stop believing X. What’s more, a person may have an emotional attachment to X that isn’t really supported by conscious “reasons” at all (e.g., the idea of not believing X causes them anxiety, so if they start to sense that X is being challenged they try to defend it with whatever argument comes to mind first). So if the main point of contention in your debate is whether X is true, and the other party says they believe X because of Y, you’ll want to gently probe Y to understand better what’s really going on. This can be done with questions such as:

-“Is Y the main reason you believe X, or are there other important reasons you believe X as well?”

-“If you were no longer convinced of Y, do you think that would be enough to cause you to stop believing X?”

-“If it turned out that X was true, would you see that as a bad thing? Why?”

UNPRODUCTIVE: “But Y is not a convincing reason to believe X.”

PRODUCTIVE: “If you stopped believing that Y was true, would that change your mind about X?”

(8) Look for the critical points of disagreement – there may be a lot of reasons that you and the other person disagree, but they are unlikely to be equally important. Chances are, there are a small number of important points of contention that the disagreement hinges on. (see CFAR’s “double crux” technique: http://bit.ly/2xYjqOI for more about finding these “crux’s”). It can be helpful to frame finding the core points of disagreement as an explicit goal in the conversation and to enlist the other person’s help in figuring it out.

UNPRODUCTIVE: “Let’s go one by one through all the reasons we disagree.”

PRODUCTIVE: “I think our main point of disagreement is actually just Y. does that seem right to you? If not, I’d be really interested in finding out what our core points of disagreement really are.”

(9) Find common ground – agreeing tends to make people feel closer to each other and to make a debate feel less heated and awkward, so it’s good to point out the things that you do already agree on. Pointing out what you think you already agree on can also be helpful for clarification because you may discover hidden points of disagreement that you didn’t even realize were there.

UNPRODUCTIVE: “It looks like we really don’t agree on this topic.”

PRODUCTIVE: “It looks like we already agree on a number of things, including…”

(10) Look for ways you can learn – one of the best things about a disagreement is that it presents an opportunity to make your own views more accurate. Even if you are right overall on the topic, you can still correct minor errors in your beliefs and arguments and hone them for the future. Chances are the other person has at least SOME good points. Try to learn as much as you can from them. At the very least, you can learn about how someone gets to the point of believing something that you disagree with on this topic, which is useful to understand in its own right. On complex, controversial issues, it is almost always the case that both sides have some reasonable points. It’s valuable to know what the reasonable points on the other side are. Keep in mind also that, like everyone, you are totally wrong about some of the things that you strongly believe in. This might just be one of those topics. If that’s true, it’s better to find out now and then start being right about it, rather than continuing being wrong about the topic indefinitely.

UNPRODUCTIVE: “I’m still not convinced.”

PRODUCTIVE: “I’d be interested in hearing more of your thoughts on Z because I’ve never thought about Z the way you’re presenting it.”

(11) Give credit when a good point is made – if the other person makes a good point, or changes your mind about something, tell them that. This might feel like conceding ground in the “battle,” but if you’re viewing the disagreement as a battle, then you are probably already not in a productive mode of discussion. Telling the other person that they made a good point or that they changed your mind about something shows good faith. It demonstrates that you are actually interested in listening to what they say and that you are not simply trying to beat them. That tends to make the other person less defensive, more open-minded, and more likely to be willing to admit that you changed their mind as well. It can help to frame the entire discussion as a collaborative enterprise to figure out the truth, rather than as a conflict between two people. Remember that finding out you are wrong about something is a gift that makes you more powerful because, the next time around, you’ll have truer beliefs and better arguments, as well as a deeper understanding of the world.

UNPRODUCTIVE: “What you said was not 100% true.”

PRODUCTIVE: “You made a great point about Z, which helps me understand this topic better.”

(12) Keep your ego out of it – if your ego is invested in coming out on top in the disagreement, or at least in not having your points refuted, then you are less likely to learn and more likely to turn the disagreement into a battle where the goal is winning. One nice trick for keeping your ego off to the side is to put forward claims without saying that you necessarily believe them. For instance, to keep an argument that you have uncertainty about at arm’s length, you can say, “Some people argue that Z, what do you think of that claim?” or if you aren’t sure you believe something, say “I’m not sure yet what I think about this yet, but one thing that comes to mind is Z.” You can also bring up the views of other people, as in “I was reading an article that said Z. What’s your reaction to that?” Using softening language can help too, such as “I’m thinking that” rather than “I believe that” or even worse, “I’ve always believed that.” The more your language makes the belief seem like a core part of your identity, the harder you may later feel it is to back down, even if you turn out to be totally wrong.

UNPRODUCTIVE: “I believe in Y.”

PRODUCTIVE: “I’m still figuring out what I think about Y. Some people say Z about it; what do you think of that argument?”

(13) Keep the other person’s ego out of it – if the other person’s ego is invested in winning (or at least not losing), it is likely going to tend to reduce the quality of the discussion. Make it as easy as possible for the other person to concede a point, for instance, by acknowledging that you understand why they think the thing and explaining how you see someone might come to that conclusion, or by emphasizing that another point they made was really good while you’re pointing out that this other point they made is weaker. Mentioning ways that you realize your own views were wrong can also be helpful, as it neutralizes the embarrassment of being wrong. If you used to believe the same thing that they do (but changed your mind about it), it can be helpful mentioning that as well to show that you understand why someone would take their side.

UNPRODUCTIVE: “You’re wrong about X.”

PRODUCTIVE: “I can understand why you said X, and you made a lot of good points about it. I think, though, that there is an important consideration that the argument you gave isn’t taking into account.”

Examples of common “ambiguous questions” include debates like:

• Is healthcare a right?

Well, what is a ‘right,’ and is that truly the question here that we mean to be asking?

• Is milk healthy for kids?

Well, what does ‘healthy’ mean, and when we say ‘healthy,’ what alternative are we comparing it to implicitly?

• Is immigration good? 

Well, what is ‘good’? Do we mean what advantages and disadvantages are there to immigration?

• Is obesity a disease?

Well, what is a ‘disease,’ and what are we hoping to determine by asking this question?]

• Is abortion murder?

Well, what are we truly trying to get at by asking if it’s ‘murder’?

• Is marriage defined as being between a man and a woman?

What do we mean by ‘defined’ here, and how will determining the ‘definition’ answer any questions for us?

• Is a trans person who identifies as a woman actually a woman?

What does ‘actually’ mean here, and in what ways does that bear on the reasons we’re having this discussion?

• Should performance-enhancing drugs be banned from sports?

What is a ‘performance-enhancing drug,’ and for what reasons do we want to ban them?

• To focus more on one specific example, consider the question:

“Is it healthier to have a diet that has no meat or that includes meat?”

Different disambiguations of these questions could begin to be tackled empirically, but the empirical methods would differ, and the difficulty level could range from relatively easy to answer to nearly impossible, depending on what we mean by the question. For instance, the questions could be disambiguated as:

• “If we take meat eaters in a society and make them go vegetarian (i.e., meatless), would that make them healthier?” Of course, this is still ambiguous because it will depend on what vegetarian food they are made to eat, and what society we started with, and how we define health (e.g., longevity? absence of chronic disease?). Even if we settle these subquestions though, the causal question is quite challenging to answer and very expensive to study, but could, in principle, be answered with randomized controlled trials (i.e., randomize some people to change to one vegetarian diet, some to change to another vegetarian diet, etc.).

• “On average, when a person decides to switch from a meat-eating diet to a vegetarian (i.e., meatless) one, does it make them healthier?” This could, in principle, be answered (once we decide on a metric for health) with a longitudinal study tracking people across time, but we’d still have to rule out effects of other changes besides the dietary ones that tend to coincide with the dietary changes, which could be tricky.

• “If we were constructing the ‘optimally healthy on average’ human diet, would it include any meat or be all vegetarian?” Well, even if we decide on how to define “healthy,” nutrition scientists probably just don’t know enough about humans to answer this question. Plus, there are SO many possible diets; there is no way we could test them all. But in principle, we could start testing many combinations of foods in carefully controlled experiments where we make study participants switch to a prescribed diet to find combinations that seem to work really well for health. This would at least make a small amount of progress. It could, of course, even turn out that the “healthiest” diet is strongly specific to the individual’s body or that the “healthiest” diet will one day involve foods specifically manufactured for each person’s body.

• “If we look at vegetarians within non-vegetarian cultures, are they healthier than the non-vegetarians around them?” Well, this could depend on which cultures we’re talking about but could be answered with individual-level survey data once we pick a culture and a metric for health.

• “If we look at traditionally vegetarian cultures, are they healthier on average than non-vegetarian cultures?” Well, what do we mean by health? If we settle on a definition of health (e.g., longevity or rate of heart attacks), we could then answer this empirically with group-level statistics about different regions (e.g., what’s the average longevity in the mostly vegetarian vs. mostly non-vegetarian regions). This is probably not what most people mean by the question, but it’s relatively easier to answer than most of the other disambiguations.

Our society cannot reach conclusions effectively with out getting a clear view of our issues.