Even seemingly benign forms of centralization have myriad economic and political costs, depending on who ultimately controls them.
read more…
For decades, Hayek’s argument has provided a basis for rejecting regulation of all kinds. If the regulation of economic activity (such as measures governing new product launches) or price regulation (such as caps and regulations) impedes the functioning of the pricing system, it will hinder the decentralized process of adapting to the ever-changing world. Become. .
Friedrich von Hayek is best known for the seminal and controversial 1944 Road to Serfdom. However, his most famous work in economics is “Society’s Use of Knowledge”, which describes how societies use and acquire distributed information about economic fundamentals such as preferences, priorities, and productivity. A fairly short article about
The article lays out a powerful critique of central planning, which states that any central agency properly collects and processes “an incomplete, frequently contradictory, scattered piece of knowledge owned by every individual individual.” argues that it cannot. Central planners are doomed to fail if they can’t capture individual preferences among millions of products, much less ideas about where each individual’s talents can be most productively and creatively utilized.
In contrast, a market economy can efficiently and effectively process and aggregate such information. Price signals seamlessly convey data about the preferences and preferences of market participants. As tin becomes scarcer, its price rises. “What tin users should know is that some of the tin they once consumed is now being used more profitably elsewhere, and as a result tin must be conserved. It’s just that.”
Also, this is more than just processing existing data. Hayek argues that market systems are also good at discovering or generating new and relevant signals. “Data, which is the starting point for economic calculations, is never ‘given’ to a single mind that can solve it for the whole society.” Its impact is never given. “
Hayek is known for offering a knowledge-based (or “computational”) critique of central planning, but his argument is best understood as a call for broader decentralization. “If we can agree that the economic problem of society is primarily one of rapid adaptation to change, the final decision must be left to those who are familiar with these situations,” he said. It has said. Ultimately, Hayek concludes, “this problem must be solved by some form of decentralization, i.e. through market economies and pricing systems.”
For decades, Hayek’s argument has provided a basis for rejecting regulation of all kinds. If the regulation of economic activity (such as measures governing new product launches) or price regulation (such as caps and regulations) impedes the functioning of the pricing system, it will hinder the decentralized process of adapting to the ever-changing world. Become. .
But now, artificial intelligence, especially generative AI models that encode, process, and deploy large amounts of pre-existing information (via hundreds of billions of parameters) pose two challenges to Hayek’s argument.
First, given AI’s ability to absorb, organize, and interpret large amounts of data, one might wonder if central planning could be more efficient than today’s market systems. This is the hope behind “AI socialism” (or “fully automated luxury communism”). AI gives central planners the means to determine optimal and (possibly) charitable economic allocations.
But while AI socialism is an interesting thought experiment, it offers only a superficial critique of Hayek. Even if AI could do all the calculations and data collection that market economies already do (a very big assumption), the concentration of power in the hands of central authorities would be a major concern. prize.
The starvation that killed 5 million Ukrainians in the early 1930s was not the result of Stalin’s failure to calculate proper apportionment. On the contrary, he had enough information to use it to extract as much grain as possible from the area (because of greater political motives and possibly his desire to devastate Ukraine). .
Moreover, Hayek’s criticism of central planning goes beyond just calculating existing numbers. As we have seen, it is primarily focused on adapting to change and thus on creating information as well as using it.
“The kind of knowledge I was interested in is, by its very nature, the kind of knowledge that cannot be incorporated into statistics,” Hayek writes. This means that even an all-powerful Large Language Model (LLM) cannot handle the nature of distributed information.
But AI also poses a second, deeper challenge to Hayek’s argument. In the age of generative AI like ChatGPT-4, should we assume that the market will encourage decentralized use of information? and led by Microsoft. Even if other companies can compete with this duopoly, LLMs by their nature may require a high degree of centralization. It’s very easy to imagine a scenario where most of humanity gets their information from the same model.
Of course, Google’s and Microsoft’s control over information is not the same as that of the Chinese Communist Party. But as Simon Johnson and I argue in our new book, Power and Progress: Our millennial struggle for technology and prosperity, Even seemingly benign forms of centralization have myriad economic and political costs, depending on who ultimately controls them. In the U.S., these costs include increasing tech monopoly as control over data creates barriers to entry, developing business models based on ongoing online engagement and personalized digital advertising. and this creates emotional rage, extremism and echo chambers online. , adversely affects democratic participation.
Decentralization is therefore still desirable. But promoting it in the age of AI may require turning Hayek’s argument on its head, or at least changing its face, by embracing regulation rather than focusing solely on its potential costs. yeah. — Project Syndicate
Daron Acemoglu, Professor of Economics at MIT, is co-author (with Simon Johnson) of Power and Progress: Our Thousand-Year Struggle Over Technology and Prosperity (PublicAffairs, May 2023).
