Opinion: No matter how good AI gets, it won’t beat markets

The economy isn't a vast set of equations. It's a complex discovery process done in real time. Even the best computers aren't up to that

By Peter Boettke. 

"Whenever we see big leaps in computation, would-be central planners come out of the woodwork, claiming this finally makes it possible to organize the economy better than markets do — optimizing tax rates, producing enough to meet our needs, and allocating resources in a way that maximizes well-being for all.

Such arguments gained theoretical prominence in the early 20th century, saw a resurgence with the mid-century advent of modern computing and operations research, and have emerged again with the impressive advance of artificial intelligence (AI).

But this line of thinking rests on a false premise: that an economy is nothing more than a computational problem to be solved with accurate equations and enough data and processing power.

As I argue in a recent paper for the Montreal Economic Institute, this error was understood as far back as the 18th century by Adam Smith (1723-90). In his Wealth of Nations, which just had its 250th birthday, Smith observed that producing even simple goods requires the co-operation of so many different hands that the full network of exchanges would “exceed all computation.” Even the making of a woollen coat, for instance, required farmers, spinners, dyers, merchants, shippers, and so on just to get from raw materials to market.

Such complexity doesn’t stop the coat from being produced. But Smith’s point is that there is no single mind directing every step of production, from raising the sheep to selling you a brand-new peacoat. Instead, it is through the spontaneous co-operation of the many hands and minds that make up the “invisible hand” of the market that such production is possible.

In the late 19th century, Italian economist Vilfredo Pareto (1848-1923) expanded on this point, observing that co-ordinating even a modest economy and matching resources to uses and preferences would soon cause an explosion in the number of equations to be solved. But today’s computers can handle quintillions of computations per second, more than Pareto could possibly have imagined. Doesn’t that make a difference?

This is where Nobel laureate economist Friedrich Hayek (1899-1992) comes in. Hayek explained that the problem is not merely that the relevant knowledge is decentralized — spread out across millions of individuals — but that it is often tacit. Local shopkeepers’ understanding of their customers’ buying habits cannot be translated into one data point to feed into an AI or any other kind of model. Nor can we predict the emergence of an entrepreneur dreaming up a product that did not exist before.

Most important of all is the phenomenon of prices — indispensable signals that guide our decision making. Prices are neither set in stone nor arbitrarily fixed. Instead, they emerge from real exchanges. When the price of wheat rises, it is because buyers and sellers are competing for a limited supply. This price increase signals something about relative scarcity. It also provides an incentive to adjust consumption and conserve the resource, to look for a substitute, to increase production and to innovate.

In short, prices are not lying around in the wild, waiting to be harvested and fed into an algorithm. Rather, they are the result of constantly evolving discovery. Without this process of discovery, the knowledge embedded in a price simply doesn’t come into existence.

Hayek called the price system, with its ability to generate knowledge in the market, a “marvel.” He described competition as a “discovery procedure” that does much more than allocate resources. When entrepreneurs bring new products to market, for instance, they are making informed bets. If they’re wrong, they bear the cost. If they’re right, they reap the rewards. Through this process, we all learn a little more about what is possible, what is valued and what works.

As for AI, it can process truly vast quantities of historical data to detect patterns, forecast trends and optimize within given parameters. But it can only look backward to find data, whereas economic life is forward-looking and creative. The growth of the social-media influencer market, to choose but one example, could hardly have been predicted by an algorithm 20 years ago. In the same way, today’s algorithms can’t accurately predict what or how much we’ll consume tomorrow, since much of what will matter tomorrow hasn’t been imagined yet.

As powerful and helpful a tool as AI can be to improve logistics, better manage inventories and analyze markets, it remains just that, a tool. It can help us gain a better understanding of markets but only markets themselves can predict and co-ordinate the results of the billions and billions of voluntary exchanges that take place every day."