Artificially Intelligent by John Mauldin
Paraphrasing Patrick Watson & Palantir's Joe Lonsdale
By John Mauldin
When new inventions turn into market frenzies, the contrarian part of me wants to be skeptical. But the optimistic part of me wants it to be true, especially when the idea promises to change life for the better. Reality is usually somewhere in between. And that, I suspect, is where the current artificial intelligence frenzy will go.
My observation through many cycles is that these fevers go too high, but they aren’t imaginary. The ideas are often real and so are the potential benefits; they just take longer to develop than most investors can tolerate.
AI may prove to be an exception because it is developing so fast—and is already showing tangible, cost-effective applications. This is not the 1990s pets.com sock puppet. AI systems like ChatGPT (released less than two years ago, remember) are doing things most never thought possible. Some of it is questionable quality, some may be socially undesirable or even harmful. But it’s real and improving quickly.
AI came up in many SIC sessions because it’s touching almost every corner of the market in some way. I focused on it more specifically with Joe Lonsdale and Ben Hunt. They look at it from different perspectives but had some remarkably clear thinking on this sometimes murky subject. Today I’ll share some of their wisdom with you.
But to put this in context, let’s review a piece my associate Patrick Watson recently shared with subscribers of our Over My Shoulder service, in which we send succinct summaries of interesting research. Here’s how Patrick described a pithy essay on AI by technology analyst Benedict Evans:
If you aren’t getting Over My Shoulder, you are missing out on tons of material never mentioned in my weekend letters. I highly suggest you subscribe today.
Now, to the AI conversation at the SIC, reduced to a few pages from 40+:
Statistical Engines
Joe Lonsdale has had the advantage of watching AI develop while also knowing the main players. He co-founded Palantir and now runs 8VC, a large venture capital firm with investments in AI and many related disruptive technologies. No one knows the AI big picture better than Joe. Really. As good as he does? Yes. But not many.
Joe started our SIC conversation by noting how this thing we call “AI” is really a form of statistical analysis. You may remember, way back before COVID, stories about Google’s “DeepMind” system learning to beat top humans in games like chess and Go. It did this by playing billions of games against itself, while noting what worked and what didn’t.
This wasn’t rocket science; it just required giant amounts of processing power that had never been feasible before. But something interesting happened. Here’s Joe from the SIC transcript.
“It turns out there's these simple constructs of statistical feedback where when you scale them up, it actually seems to approximate different types of intelligence really well. These are called large parameter models, or on [just] words they're called large language models, and it's basically like an engine that builds many, many abstract layers to predict the next word, the next token… So that statistical engine is the thing we talk about.
“Now most of the time we're talking about AI and it's having a huge impact on the economy, and I think it is important to think of it as a statistical thing, but in some cases it feels like it's reasoning in a lot of areas. You could train it to think in different ways to perform different tasks and we're doing a lot with it right now.”
Note those words carefully. AI “feels like it’s reasoning.” Is an AI system reasoning? Or is it just counting words in such vast numbers that it seems to be thinking? That’s not entirely clear, and on some level is an abstract philosophical question. What is “thought,” after all? We don’t have an objective definition. Though, as a human, I like to think there is a difference between a billion neurons firing and 10 billion bits on thousands of Nvidia chips firing.
In any case, Joe’s point here is critical. AI models—at least those we have now—are just statistical engines so powerful they can appear to be something else. (Ben Hunt had more to say on the dangers of that, which we’ll get to in a minute.)
AI is happening fast. I asked what this meant for jobs and the economy. Will we have time to adjust?
“You never want to say this time is different, but it's going really, really fast. And if you see adoption curves, they're happening quickly. So yeah, I think the rule holds that when you create more wealth, there's still many, many things to fix in the world, and so there's going to be many, many more and better jobs overall with where AI is right now, and it's going to create a lot more wealth, a lot higher productivity in our economies.
“Our view is you're going to actually see that show up from the productivity statistics sometime in the second half of this decade. I'm happy to be proven wrong on that, but it seems very likely from the things we're seeing.
“Is there going to be an adjustment period that's a little bit difficult because it's so fast? And there probably is, and this is a really interesting question. I happen to be in DC today as I'm talking to you, John (I think he was trying to explain AI to senators, which may be a bigger problem than creating AI), and I think one of the big things our country's facing is a rise of populism on both sides and does AI exacerbate that? It very well might at some point in the next five years.
“And that said, it could also create and should create so much wealth and it should make the living standards go up so much that hopefully the productivity thing offsets these problems, but it doesn't mean it's not going to be in for a volatile time.”
That last point struck me because the SIC also featured a panel on historical cycles which also point to a “volatile time” just ahead. Whatever disruptions AI brings will likely coincide with the Fourth Turning’s social and political events, creating whole new sets of elites, geopolitical risks, and economic disruption. I suppose AI might help us get through those times more easily. It also has the potential to go badly south.
Meanwhile, Joe thinks AI will help improve the economy by taking on some boring but necessary drudge work.
“What's the stuff that can be really improved? What's the gap in the economy? It's going after the pre-internet stuff that hasn't been upgraded, and I define that as a lot of these service parts of the economy. There is the legal part of the economy. There's a lot of financial services creating estates and trusts and whatever other work people are doing by hand.
“Healthcare billing is a huge area. We just talked about customer support earlier. Healthcare billing, John, just to give you an example. There's about a quarter trillion dollars a year, by most estimates, spent on healthcare billing in the US economy, $250+ billion. And there's people sitting in office parks and suburbs and there's millions of them and there's tens of thousands of rules per insurance company and thousands of insurance companies. It's a mess. They call it ‘revenue cycle management’ because you're cycling back and forth trying to get these things accepted. And it turns out that using AI could make the workflow… so far, we've proven it at least twice as efficient, I guess the margins are going to go up three or four times.” (It’s one of his major VC initiatives.)
Go paid at the $5 a month level, and we will send you both the PDF and e-Pub versions of “Government” - The Biggest Scam in History… Exposed! and a coupon code for 10% off anything in the Government-Scam.com/Store.
Go paid at the $50 a year level, and we will send you a free paperback edition of Etienne’s book “Government” - The Biggest Scam in History… Exposed! OR a 64GB Liberator flash drive if you live in the US. If you are international, we will give you a $10 credit towards shipping if you agree to pay the remainder.
Support us at the $250 Founding Member Level and get a signed high-resolution hardcover of “Government” + Liberator flash drive + Larken Rose’s The Most Dangerous Superstition + Art of Liberty Foundation Stickers delivered anywhere in the world. Our only option for signed copies besides catching Etienne @ an event.