Thursday, May 14, 2020

Would AI make socialism finally work?

The Soviet economy suffered from chronic problems in planning, in production management, and in distribution. Let’s briefly analyze which of these problems lend themselves to Artificial Intelligence solutions, and which to a paradigm shift toward a more flexible economy. Another thorny problem to address: the design of an efficient automated Planning and Scheduling capability for the entire economy—a task facing a nearly infinite array of possibilities at the onset.

The task quickly runs into a problem AI programmers call “Combinatorial Explosion,” where the amount of time and computer resources needed to solve a problem grow exponentially and overwhelm finite cyber capabilities.

For example, should capital goods A and B go to factory X or Y before being processed as feedstock to factories C and D, or can raw materials E and F partially serve as feedstock substitutes to factories C and D, etc., etc.?

To avoid Combinatorial Explosion, Soviet managers would have to exercise human judgment and limit the range of possibilities. And so, a general, all-encompassing AI program could scarcely run the entire economy by itself—one with workers merely cogs in the machine, rather than active planning participants. …

Soviet Communism’s ailments were more ideological in nature than technological, ailments that AI computer technology cannot entirely resolve. But if a Communist government were indeed to address these problems by adopting the suggested market solutions, the economy—or parts of it—would transform into something suspiciously similar to a western economy—which is exactly what Soviet premiers Nikita Khrushchev and Alexei Kosygin had once separately proposed, and what the Chinese have in fact actually done.
Source: “Artificial Intelligence and the Rebirth of the Soviet” by Darian Diachok

No comments:

Disclosure

Links to Amazon.com are affiliate links and earn commissions.

Your support is appreciated.

Blog Archive

Subscribe — Follow by Email