How The AI Boom Is a Pascal’s Wager

So you're watching a video in a series about the fundamentals of artificial intelligence or at a high level, what ideas and concepts and trends you should understand to know what's going on in the world of AI and where this technology might go, where it is going, what it means for you and how to get the most use out of it for you personally and in your career. And so on that note, one of the interesting questions to ask about AI that is very good It's good in making you think about These dynamics is that of Pascal's wager as applied to artificial intelligence. Now, in this video, I'm going to explain what Pascal's wager is. If you haven't heard of it, how it's similar to A.I. and what kind of second order questions and critical thinking it prompts and then where it's different from the insights that are also unlocked in how it differs from the traditional kind of religious archetype. So that's going to be the video. It's going to be a little more of a philosophical one, but still a very interesting Now, quickly, before we keep going, as you know, I run this YouTube channel and have a full-time job at a startup. And so oftentimes, the things that eat up my time during the week is not the work, it's just communication in my inbox. Right, so for example, every brand deal lives in the inbox. So briefs, contracts, invoices, follow-ups, all happens over email. And so if I needed something like a tracker to organize things, I'd have to build it by hand on a Sunday and then just rebuild it the next week because nothing ever really stuck. And that is why I started using the Superhuman Connectors. And this part of the video is sponsored by Superhuman. So Superhuman Mail and Superhuman Docs, also called Super Docs, are both part of the Superhuman Productivity Suite. And the connectors plug both of them into Claude and ChatGPT. So that means the AI tools I already do most of my work in can write to my inbox and also my documents without having to copy and paste. Now here is how I use it. So I'm on my computer here for Superhuman Mail. I ask it to find every sponsor thread, for example, from the past two weeks, where I'm still waiting on I'll check the red status on what I sent and then draft a follow-up for the ones that were opened but never answered. So the drafts land in my Superhuman Mail inbox, I tweak one or two and then I just hit send. Now second is Superdocs, so I ask it to pull, for example, every contract and invoice from this quarter out of my inbox and build a table in Superhuman Docs with the brand, the amount And the due date, that used to take a ton of time. Now it's one prompt and the editor will see the same table that I do and everyone else on the team. And that used to be genuinely a lot of work. Now it is one prompt inside of Docs. So if your inbox runs your business and your work week like it does for me, then you need to try these Superhuman connectors. You can sign up for Superhuman using my link on the screen here or in the description. So Pascal's wager is a rationalist argument. about religion and what it says is hey if you are not religious and you didn't believe in god and you go to hell for eternity that's really bad and if you are religious and you get to go to heaven because of your belief in a higher being that's really good for you and what's critically important is if you do believe and none of it was real right you have nothing to lose and so a rationalist may Let me look at that and say, well, hey, I have a lot. More to gain, a lot less to lose by believing in a higher power, irregardless of whether that higher power is actually there. Now that induces a whole very gnarly set of questions about faith and religion, and we're not going to get into that, but the concept and the point there stands. And so how does that question and that quadrant have parallels to AI? Well, it does in a lot of ways, and I've made videos about this before, because there are synonyms, there are parallels between AI and God, right? And AI, in many ways, is a religion, a sort of belief about the linear returns to more intelligence, about superintelligence and consciousness. Now, this is getting very abstract. If you've never seen a video on this channel before, I nearly always make videos about AI tools and... platforms and how to get the most views out of them. That's like 90% of this channel. But just bear with me here and you'll see how this all ties together and connects. And so swap out some of the religious aspects with the capital aspects, right? Let's frame it as, okay, we might get to super intelligence to this incredible AI. And that AI is a very effective capital player. It's very good at making money because it's very good at solving problems. problems that might cure cancer or advance future forms of energy. And so traditionally, Pascal's wager, what you say is, should I believe or not believe in God? In this AI paradox, what you say is, should I or do I believe or not believe that we will get to a point of this incredible artificial intelligence that will solve all of our problems and make a ton of money? And some people believe that we're going to get there, and some people don't believe. And that question is the question of... Are we in a bubble? Are we not in a bubble? Are we over-investing into data centers, into AI, or are we radically under-investing? And the strongest parallel here, then the next of thinking that through is, okay, well, if you are Dario, if you are Sam Altman, if you are one of the people who are driving this industry forward, one of the hinges of your belief system about AI in many cases in their own words is that we simply have more to gain than we have to lose, is that the promise of that super intelligent thing that solves our problems is so great, so amazing, that it's worth the risk of it not working out, and it's worth investing an incredible amount of resources now. And what is that, if not similar to that rationalist theory of, well, I should believe in God because the EV is higher? And it's not completely the same, and I'll get to that, but you can start to see what I mean there. The bet on superintelligence is it's worth to invest an enormous amount of money and resources. time and risk bad outcomes because that good outcome is just incredibly good. And so sure, maybe you overbuilt data centers and you wasted a few trillion dollars there, but that's worth it if there's a 1%, 5%, 50%, 80% chance, whatever number you believe it is, the chance that you'll produce this super intelligence that solves extremely valuable endemic problems that humans face that provide an ample return on that money. So those are the parallels that I want to get across. And again, there's a lot of questions there. And you get to this weird kind of abstract, like metaphysical take on AI, and it becomes a question of consciousness and religion. And we're not going to get into all of that. But that dynamic does really affect how people price these companies now and how the actors inside of them Go ahead and move ahead in the industry and with the technology. But what's different about it? And there is a really important difference, which is the arguments in the traditional Pascal's wager is that if you believe in God, you will go to heaven. E.g., if you believe in the positive outcome, then you are guaranteed to get that positive outcome. Again, we're going to put a lot of religious questions there aside. Yet in how most people think about it, including the decision makers in the industry, you can only get to the really, really good outcome of AI if you risk getting to the really bad outcome of AI and super intelligence. And that difference arguably is a large core driver of that debate around AI and whether you should shut it down, whether you should slow it down or do the opposite. And when you hear people in the industry saying, hey, there's a 10% chance that it goes really south with superintelligence and the 90% chance it goes really incredibly well, what they're making is their own version of Pascal's wager. They're saying it's worth it to risk the 10% to go for. And to not believe in that future, to not strive for it, also disregards the really, really beautiful outcome and incurs all of this sunk cost suffering, right? Cancer, death, all of all of the the calamities of humanity and the human experience. Now, again, you get into a lot of Big philosophical questions that we're not going to touch in this video, but ideology is driving a lot of the real world financial debates and impacts, the valuations of certain companies, what's being built or not built, and where the evolution of AI is the philosophy of those who work in it and around it and deal with it applied to the physical and the capital. That is the concept I will end this video on. I hope that is some food for thought for you. These are not questions that anyone is the answer to, much less myself, but I think they are good questions to ask and to think about and to understand the decisions of others. Appreciate y'all being here, and I will see you in the next one.