Dwarkesh Patel: evaluation

Transcript context

…It's worth considering that I'm maybe not very good at this. First off, I'm not that successful as a researcher. I object to the LeBron James characterization. It's true that I'm maybe the most successful crowdfunded researcher in tech stuff, and that's kind of weird. In the last couple of years, I've figured some stuff out, but I wouldn't say I've had any spectacular hit publications. One thing that is true of this is that when I have big publications, I get a lot of new subscribers. There is some kind of market force that could be higher if I were having more spectacular success with my research. It's also true that I systematically avoid marketing it. That's a self-protection thing. I am really worried about the corrosive influence of audience and marketing on honest inquiry. It is very easy to distort my work. It's almost a default to try to make it be something that people would be more likely to like rather than the thing that I actually want to investigate, or to do the boring simple version of it rather than the interesting deep version so I can publish more stuff more often. One thing that I've chosen not to do, and it's a choice that's definitely cost me financially, is to publish what academics would call minimum viable units of paper. They have a pithier phrase than that. Minimum viable papers. It's very common to take any new marginal insight that is above a particular bar and publish that. I just haven't done that. I've written informal letters to my patrons, “Hey, I figured this thing out this month.” If I were an academic, I would have published that as a paper. If I were a marketing-oriented crowdfunded researcher, I would have done some glossy thing and promoted it “Look at this thing I figured out.” But actually, I just don't think it's that big a deal and I'd rather get on to the next thing. I have that choice of waiting to publish. That's not really what I'm worried about. Really what I'm worried about is marketing, man. Marketing. It makes it so hard to be honest with oneself, at least in my experience. Not only to be honest, with what I think is interesting and important, but even to be honest about the results. Every paper is, in some sense, a little marketing piece trying to make the case that it's significant, that its results are really exciting, really important. That is really corrosive to discovery. It's true that you need a really strong emotional connection to the work in order to do good work. Part of that emotional connection comes from a sense of excitement of maybe being hot on the tail of something really good. There's a temptation to portray what you found in the best possible light, to downplay its limitations, to take up space and to totalize. All of this is just death for discovery. It is interesting to hear that from somebody who inadvertently and without intentionally trying to do so has done a good job of spreading your material. I've known about you for a long time. I do wonder if there's an element of, if you get to a certain level of quality, trying to market your stuff, not only doesn't help, but probably hurts you. If you can try to think of somebody like Gwern trying to post YouTube shorts of his blog post, it would just be like, “What are you doing, man?” It's just so good that he doesn’t need to promote it. Gwern is an interesting example because there's a simpler failure mode. I still routinely run into people who will tell me, “Oh, I've really liked your work for a while. I didn't know you had a Patreon.” That's a simple failure of a certain kind of marketing on my part. Gwern actually has this even worse. I adore Gwern. I have learned so much from him. You can go to his Patreon page and he actually makes public his revenue. He makes a tiny fraction of what I do on Patreon. I think this is inappropriate. Gwern is a much more impactful researcher than I am and he has a much bigger audience than I do. The fact that they aren't converting into patrons is mostly a matter of the way that he talks about it and the way that he presents it. It's not that he needs to market more people to his webpage. I expect he has plenty of traffic and a large audience. It's much larger than mine. There are a bunch of variables about the way you talk about this membership offering. None of us really want to think about them. I've ended up at a slightly more effective part of the space, but I'm pretty sure that there's much more effective ways to do whatever it is I'm doing.…

Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.