I believe for some people it’s very important to have a moment of realization that one can get to the frontier of knowledge in a given field of interest. It feels intimidating if others are making contributions that seem decisively out of your league. Because people might intuitively underestimate how far you can get with focused reading and learning, it could be good to give tailored advice to people newer to (e.g.) AI risk for how/where they can make contributions that will feel encouraging. For illustration, a few years ago I was playing a computer game for fun for quite a while until I was by chance matched up with the one of the better competitive players and I almost won against them, getting lucky. That experience showed me that I’d have a shot if I actually tried, and it encouraged me to immediately start practicing with the aim of becoming competitive at that game. It changed my mindset over night. Similarly, I think there’s a difference in mindset between “reading and talking about research topics for fun” and “reading and talking about research topics with the intent of seriously contributing”.
I agree with others that a rewarding social environment and people in a similar range of competence you can bounce ideas back-and-forth with are extremely important. If you collaborate with people who are similarly driven to figure things out and discuss ideas with you, that automatically forces you think about your ideas for much longer and in more detail. By yourself you might stop thinking about a topic once you reach a roadblock, but if every morning you wake up to new messages by a collaborator adding criticism or new bits to your thinking, you’re going to keep working on the topic.
I also suspect that people are sometimes too modest (or in the wrong mindset) to develop the habit of “taking stances”. Some people know about a lot of different considerations and can tell you in detail what others have written, but they don’t invest effort coming up with their own opinion – presumably because they don’t consider themselves to be experts. Some of the community norms about not being overconfident might contribute to this failure mode, but the two things are distinct because people can try practicing taking stances with personal “pre-Aumann opinions”, which they are free to largely ignore when deferring to the experts for an all-things-considered judgment.
Speculation about personality traits conducive to generating ideas: OCD was mentioned in the comments. There’s also OCPD and hyperfocus. Carl Shulman’s advice for researchers among other things mentions something about having a strong emotional reaction to people being wrong on the internet (in communities you care about) – I think this might be a symptom of being very invested in the ideas, and it can help further clarify one’s thinking while trying to articulate fervently why something is wrong. Need for closure also seems relevant to me. It has its dangers because it can lead to one-sided thinking. But in me at least I’m often driven by feeling deeply unsatisfied with not having answers to questions that seem strategically important. And, anecdotally, I know some people with low need for closure who I consider to be phenomenal researchers in most important respects, but these people are less creative than I would be with their skills and backgrounds, and their obsessive focus maybe goes into greater width of research rather than zooming in on making progress on the “construction sites”. Finally, I strongly agree with John Maxwell’s point that a “temporary delusion” for thinking that one’s ideas are really good is a great reinforcement mechanism (even though it often leads to embarrassment later on).
Re temporary delusion that one’s ideas are important (or new): a good example is individuals who file patents. They think they have a world-beating invention. In fact 90% of patents are never used, and getting a patent costs a fortune. There’s a well-known phenomenon I’ve heard called ‘search shock’, when a naive inventor goes to a patent attorney, who conducts a patent search and reveals to the incredulous inventor that every supposedly original, brilliant aspect of the invention has been both thought of and patented before.
I believe for some people it’s very important to have a moment of realization that one can get to the frontier of knowledge in a given field of interest. It feels intimidating if others are making contributions that seem decisively out of your league. Because people might intuitively underestimate how far you can get with focused reading and learning, it could be good to give tailored advice to people newer to (e.g.) AI risk for how/where they can make contributions that will feel encouraging. For illustration, a few years ago I was playing a computer game for fun for quite a while until I was by chance matched up with the one of the better competitive players and I almost won against them, getting lucky. That experience showed me that I’d have a shot if I actually tried, and it encouraged me to immediately start practicing with the aim of becoming competitive at that game. It changed my mindset over night. Similarly, I think there’s a difference in mindset between “reading and talking about research topics for fun” and “reading and talking about research topics with the intent of seriously contributing”.
I agree with others that a rewarding social environment and people in a similar range of competence you can bounce ideas back-and-forth with are extremely important. If you collaborate with people who are similarly driven to figure things out and discuss ideas with you, that automatically forces you think about your ideas for much longer and in more detail. By yourself you might stop thinking about a topic once you reach a roadblock, but if every morning you wake up to new messages by a collaborator adding criticism or new bits to your thinking, you’re going to keep working on the topic.
I also suspect that people are sometimes too modest (or in the wrong mindset) to develop the habit of “taking stances”. Some people know about a lot of different considerations and can tell you in detail what others have written, but they don’t invest effort coming up with their own opinion – presumably because they don’t consider themselves to be experts. Some of the community norms about not being overconfident might contribute to this failure mode, but the two things are distinct because people can try practicing taking stances with personal “pre-Aumann opinions”, which they are free to largely ignore when deferring to the experts for an all-things-considered judgment.
Speculation about personality traits conducive to generating ideas: OCD was mentioned in the comments. There’s also OCPD and hyperfocus. Carl Shulman’s advice for researchers among other things mentions something about having a strong emotional reaction to people being wrong on the internet (in communities you care about) – I think this might be a symptom of being very invested in the ideas, and it can help further clarify one’s thinking while trying to articulate fervently why something is wrong. Need for closure also seems relevant to me. It has its dangers because it can lead to one-sided thinking. But in me at least I’m often driven by feeling deeply unsatisfied with not having answers to questions that seem strategically important. And, anecdotally, I know some people with low need for closure who I consider to be phenomenal researchers in most important respects, but these people are less creative than I would be with their skills and backgrounds, and their obsessive focus maybe goes into greater width of research rather than zooming in on making progress on the “construction sites”. Finally, I strongly agree with John Maxwell’s point that a “temporary delusion” for thinking that one’s ideas are really good is a great reinforcement mechanism (even though it often leads to embarrassment later on).
Re temporary delusion that one’s ideas are important (or new): a good example is individuals who file patents. They think they have a world-beating invention. In fact 90% of patents are never used, and getting a patent costs a fortune. There’s a well-known phenomenon I’ve heard called ‘search shock’, when a naive inventor goes to a patent attorney, who conducts a patent search and reveals to the incredulous inventor that every supposedly original, brilliant aspect of the invention has been both thought of and patented before.