The Exchanges

Q it. Uh, I think one of the most important books I've read in the last few years. Let's just start off a little bit of the history and things you've talked about in the past. I think it's fair to say you're one of the most forward thinking kind of bloggers and thinkers online today. You went to Harvard and you, like you studied government. How'd you get into blogging?

A You know, I, um, I, I was first, I was running a small business, like an ed tech company with my friend, Andrew Finn. And we, uh, and I, and I was, I was blogging on the side as kind of just a, uh, a hobby, um, like five hours a week. But I, I, I had liked that more than I liked running the business. So at one point we just Decide we don't need both of us here. And I'm going to go and see like, what, what if I put 60 hours a week instead of five hours a week into blogging? So I started this new blog, wait, but why? Um, and, uh, what year was this? This was 2013. I mean, I guess the thesis was the, you know, people said blogging's dead. That was like a common thing to say back then. And like the internet was just flooded with content. You had Buzzfeed now producing a trillion articles a week. Um, and, but the, the thesis we had was kind of that There's actually not very much high quality, like really high quality stuff. There's a lot of like B minus content, C plus content out there. And if you can just work really hard and, and really put out, you know, a level content that people will notice. That was the idea.

AI assessment note: “I was blogging on the side as kind of just a hobby”

Q let's talk about it a little bit. You're, you're obviously prophetic because AI hadn't really been as much in the consciousness in 2015. A bunch of us were trying to invest in it, trying to figure it out, but obviously it didn't really explode like as something that was working until the last few years. So what did you see that led you to this conclusion? What, what'd you know?

A I think that, uh, actually it's, I think that, uh, Nick Bostrom and Ray Kurzweil and some others, I think they're, they're the prophets, um, because, um, all I did was read Some of their book. Basically my judge for whether it's a good thing to, something's a good thing to write about, um, is I just use my own self. Like it, do I find it interesting? I assume that my audience is a big group of me, um, because they're, they're close enough that usually that's a good gauge. So when I hear about, when I started hearing about AI in 2015, it was really, you know, that's pretty early where I think people forget how recently it still seemed just like a sci-fi concept. AI seemed like something that is in sci-fi movies only. Um, And in 2015, that started to change. It was actually 2014. I first started hearing about it. I wrote about it at the very beginning of 2015. And I started to hear about real serious people talking about it. I said, what's what's going on? So I read Nick Bostrom's book. I read, you know, The Singularity is Near by Kurzweil. Nick Bostrom's book is super intelligent. I mean, these are these are these books really, I think, were way ahead of their time and my eyes were wide. And so I said, holy shit, you know, and if I feel this way, I'm sure my readers are also kind of hearing about it, wondering what's going on and they're going to have their minds blown. I decide…

AI assessment note: “So I read Nick Bostrom's book. I read, you know, The Singularity is Near”

Q Do you really think that's the biggest threat? Because for me, I still think it's people using AI for the next decade. It's going to probably cause the problems you, but you're, you're worried about the AI itself at some point.

A I'm worried about both. I mean, I think that in the near future, of course, you're going to have a higher risk of Bad guys with using AI, you know, to do things like manufacturing scary bioweapons, um, things like that, that you really, you know, things that could really kill a billion people. I mean, people with, you know, with AI can, you know, there's obviously the smaller concerns where, you know, they, you can start to creating deep fakes and all of that. I mean, that, that's a big concern for civilization as well, but I'm thinking like, you know, bioweapon or, you know, autonomous drones, you know, you know, swarms of drones that can go into a city and specifically, you know, so. Um, but I think as time goes on, I just think it's totally unprecedented. We're literally like monkeys building humans. I mean, like we should be scared. We should be scared of that. Like it could turn out well, but you kind of have to get lucky. And there's a lot of people that think we're just not, we don't know how to do this. We don't know how to align this thing before it becomes too powerful to change.

AI assessment note: “I'm worried about both. I mean, I think that in the near future”

Q So this is a decade later since you wrote this and obviously a lot's happened now. I think a few years ago you said in a Reddit AMA that you were in favor of pausing any research because it's kind of scary and we don't know what we're growing here. Like, do you still feel that way? Like, how do you, how do you think about this?

A I mean, we're in a quandary here because We have to ask, what is the scarier prospect? An arms race that, that throws caution to the wind and tries to accelerate the process of building these, you know, superhuman brains, or Risking or slowing down and risking that China gets ahead of us and now has a massive technological advantage that might be, you know, permanent. Those are both terrifying, I would say to an American or they should be. And the question is what's scarier, you know, what's the bigger risk. And so you can try to find that middle ground and say, listen, we should, we should try to stay ahead, but go cautious while staying ahead. But now, you know, China has shown that maybe it's going to be harder to stay ahead of them than we thought. Maybe the, this limit on chips is not going to.

AI assessment note: “we're in a quandary here because We have to ask, what is the scarier prospect?”