Choice and Competition Will Determine AI’s Reach

by | Feb 10, 2023

Stop me if you’ve heard this one.

A doctor, lawyer, research assistant, programmer, writer, and MBA walks into a bar. Bartender asks “What you having, ma’am?”

Get it? They were all one lady…ROTFL!

Ok, maybe comedy’s not for me. But what I just described was ChatGPT – the new brain in town that never sleeps, takes a lunch break, asks for a vacation, or calls in sick.

So far, ChatGPT has passed Wharton’s MBA Exam, US Medical Licensing Exam, and the Multistate Bar Exam. It’s taken hours of study and toil – along with personal growth and future prospects – off the backs of middle schoolers and graduate students alike.

It’s writing code in any computer language. It’s writing advertisements. It’s writing books.

And it’s personally spared me the drudgery of researching and checking historical and political facts with concise summaries in a matter of seconds.

Someone posted on Twitter the other day asking what code is the best to learn for the future and my response was, “None. AI will write it.” The only skill required of any of us will be how well we interact with it.

It’s all coming on way too fast, at least for anyone paying attention.

And to help you make sense of it all, I suggest you keep in mind this immutable law of economics…

The Absolute Best

I’m talking about comparative advantage.

This key concept is typically used to explain why countries trade with one another. But it’s a far more fundamental concept than that, describing why any of us bother to engage in economic transactions in the first place.

Let’s say you and I are neighbors and we sell goods at our local village. We make bread and beer, but your bread and beer are both absolutely better than mine.

You have an absolute advantage in producing both goods.

Does that mean I’m out of luck and you get all the business?

Absolutely not. 

It’s a near certainty that you make more profit from either bread or beer. Assuming it’s beer, you’ll focus on brewing, leaving me an opportunity to bake and sell bread.

In this scenario, you have an absolute advantage in terms of quality for both goods but a comparative advantage in terms of efficiency with respect to beer. And your brewing edge creates my baking opportunity. 

We both play to our relative strengths with comparative advantage dictating what we decide to produce.

Thus, a market is born.

Comparative advantage creates the conditions for differentiated production and trade and it all evolves through competition.

Toss in the division of labor, and you now understand how the global economy works at the scale of international trade.

And, to my way of thinking, comparative advantage provides a useful framework through which we can consider the implications of AI.

AI as Economic Agent

Think of AI as another player in the economic game. It has its strengths and humans have theirs.

Compared to humans, the type of AI that ChatGPT represents can provide accurate, comprehensive responses across a broad domain of topics faster than any human can possibly hope. 

Granted, it may have a few kinks. But the more it gets used, the faster it learns. So, it’s not a stretch to assume that ChatGPT (or one of the many contenders for the throne) will work those and many other kinks out soon.

The crazy part is it works those kinks out for itself.

Anyhow, compiling information and even recommending solutions is the essence of ChatGPTs comparative advantage. 

Pair that ability with automation and it might seem that us poor humans don’t have much left to do.

But we do.

The Competition Imperative

AI must be prompted and my bet is that it will always need to be. It can’t predict our wants and needs any better than politicians, bureaucrats, technocrats, psychologists, economists or anyone else can. 

It doesn’t matter how much data an AI possesses or how fast it can process all possible outcomes. Competition, success, and failure will do far better at satiating society than any centralized agent or agents.

Even supposing AI starts to ask questions of itself and decide on a course of action, its useful domain will always be limited to responding to human demand.

Now, some of you may be saying “AI can’t do that yet”… “did you hear about this or that error,” or some other accurate but ultimately irrelevant qualification to help you put off sharpening your comparative advantage today. 

But when I see a man walking down a hall, look away for a moment, then look back, I’m not surprised to see that he’s further down the hall.

AI of the general intelligence kind is basically here. And we all, individually, need to pay attention.

Don’t get lost thinking about robots taking over and enslaving humanity. 

Should AI arrive at the point where it doesn’t need us, then it doesn’t matter how we think about, respond to, and interact with it, in any case.

So long as we matter (which is the only rational assumption), AIs reach will be limited by its absolute inability to meet market demand in the absence of competition. 

Choice and competition, as always, will ultimately dictate what gets produced.

We don’t really have to worry about “Future Shock” type scenarios either. The pace at which the economy evolves will be limited by how fast humans can absorb and adapt to that change.

Whatever AI produces, competition is the only way to determine what gets produced.

Our comparative advantage is that we compete.

Humans are, and always will be, the market.

Take What the Market Gives You.

Don Yocham

What to read next