The Profit, The Sabres, and The Best Christmas Movie

Monday's Chariot Ride: 12/30/24

The Profit

OpenAI has made news with its reported new definition of AGI (Artificial General Intelligence). In an agreement with Microsoft, they defined AGI as OpenAI making $100B in profits.

The definition of AGI is one of those concepts that has spawned intense philosophical and technical debate—and for good reason. One can think of creating an AGI as the ultimate goal of artificial intelligence research. Knowing what it is that your entire field is focused on creating wades into some philosophical territory that is (a) extremely important: if you don’t see what you’re aiming at, are you really going to hit it? And (b) very hard to untangle: is intelligence just one thing or several independent capacities? Does it require actually grasping reality, or is it fine if it converges to reality through an intermediate mechanism (like chatGPT does right now more or less well)? Does AGI require agency, or can it be fundamentally statistical? To even know what we’re talking about—and to even know what AI research is/ought to be aiming for, takes some legitimate philosophical acumen [1].

As a result, OpenAI claiming that AGI=profit dollars might seem a little confusing. Are they taking a philosophical stance here? Are they saying that intelligence=the ability to make money? That would be absurd since oil makes money—and whatever “intelligence” is it certainly should differentiate between things like oil and things like human beings.

No, what OpenAI is doing is creating a business plan. As I understand it, part of OpenAI’s deal with Microsoft says that, partly in exchange for Microsoft’s funding, Microsoft gets a license to all of OpenAI’s products until “AGI” is achieved [2]. So what’s really going on is that OpenAI is telling Microsoft that it’s only until OpenAI becomes ludicrously profitable that Microsoft has to worry about OpenAI’s tech becoming proprietary.

But this whole discussion has gotten heated in large part because of the way that “profit” has become involved in the whole discussion. The idea that profit could be some useful scientific metric is abhorrent to some. It’s a reminder of the seemingly ugly truth that OpenAI now has committed the crime of caring about money.

Well, I think they should. I think they should try to be as profitable as possible and as quick as possible. So even if it’s not really AGI in the philosophic sense, I give my best wishes to OpenAI on this artificial goal of AGI.

The Sabres

China is now proceeding with the largest military buildup since Germany in the 1930s. That’s an ominous headline which, as I pointed out in my stream-of-consciousness style piece on it here, is probably an apt comparison.

They’ve been building up their nuclear weapons arsenal, their evasive missile technologies, and their naval production capacity—as well as clearly working on drone attacks in these terrifying videos.

I think we need to be ready, and we need to be exceptionally smart. The state of warfare is going to look as different from our last major conflict as WWI looked from the Franco-Prussian wars. In the first months of WWI, there were cavalry charges against machine gun encampments; we can’t afford to be on the side of the cavalry charges when nuclear weapons are what we will be met with.

It should worry us that China feels confident enough to start amassing troops like this again. I’ve said before and I continue to believe that if there is a World War in the next 20 years, it will start with China crossing the Taiwan Strait. Luckily, I think the US is worried. We’ve also been modernizing our nuclear weapons, and we’re working on building AI into our weapons systems. We can only hope that our generals—historically not forward-thinking gentlemen—will defy their love for tradition in the name of their passion for victory.

The Best Christmas Movie

After some post-holiday travel absolutely torpedoed my immune system, I spent most of yesterday in a semi-conscious daze of flu-like symptoms. I didn’t do much of value (other than the valuable reparative work that I’m sure was happening on the inside [3]), but I did manage to watch my new favorite Christmas movie—after Christmas!

It’s called Red One—and it's fantastic. Basically, the premise is that Santa Claus, and all of the other Germanic myths associated with Christmas, are real. They’re this Nordic-style “invisible population” of elves and trolls and spirits. Santa gets kidnapped and his jaded chief of security and a hardened crook with family issues have to save him—and rediscover the Christmas spirit.

I liked this movie so much in part because it does all of the things that a good action movie should do. It’s got Chris Evans and Dwayne Johnson in starring roles, and the cinematography reminds me of an Avengers movie in some respects—in that there are action scenes with supernatural forces, and those are all done extremely well.

But what really makes me like this movie is that it does exactly what a Christmas movie should do. It’s benevolent, it’s funny, and it’s somewhat serious. It’s not a great work of romantic fiction, but it is an honest-to-god attempt to really capture what’s so great about the Christmas spirit: celebrating people making real good choices to become better men.

It’s not too late! If you can bear to keep your Santa hat on for one more day before the new year, I’d recommend checking out Red One. If your artistic tastes and spirit of benevolence are anything like mine, I think you’ll enjoy the heck out of this one.

Notes:

• [1] These are all interesting questions that are worth answering. I don’t have the answers to them right now. Maybe I should try to figure them out?

• [2] Do your own research here if it’s important to you. This is my understanding of the things I read, which may or may not be right. A good place to start might be here.

• [3] And of course write this newsletter, if you’re in the camp that finds it valuable :)