In a post-capitalist society, the integration of artificial intelligence into the economic system will drastically reduce the role of the human factor. New systems for generating artificial thought and speech, such as ChatGPT, which compete with human intelligence, are already posing a number of mysteries across the spectrum of the social sciences. What if these smart machines become smarter than their creators? How will our legal and economic system deal with them? Will they have “rights” or will they be our digital slaves?

“K” sought answers from the heretic intellectual, theorist of anarcho-capitalism, 78-year-old David Friedman. The son of Chicago School founder and Nobel laureate Milton Friedman (1912-2006), who inspired the economic policies of Reagan and Thatcher in the early 1980s, David Friedman argues that societies can live harmoniously without a central government. In our discussion, he tries to predict the problems that society will face with the advent of artificial intelligence.

– In my first book, The Freedom Machine, I described that all the useful functions of the state should be performed privately. In my opinion, these main functions are the enactment and application of the law, that is, the enforcement of individual rights, as well as the settlement of disputes.

In the system I have outlined, this was done in a decentralized way, where there were businesses that sold services to protect the rights of their customers and also to resolve their disputes. Thus, every person would like to become a client of such a business. Each pair of firms, realizing the possibility of a conflict, agreed in advance on an arbitrator.

On the other hand, neither company would like to start a conflict with the other, because the “war” between companies is very expensive, so it makes sense for both companies to commit in advance to accept a private court, an arbitration body, where both sides of the company will monitor it results and decisions.

In fact, we have a model for what I’m describing in today’s world: auto insurance companies. If my car crashes into your car, my insurance company will want to believe that you are at fault, and your insurance company will want to believe that I am at fault. And they do have two alternative ways to resolve the dispute. One of them is to sue each other and go to court. This is a very costly process, another way is to establish mutually agreed rules to decide who is responsible and who is not. In today’s America, these cases never go to trial. This is the main logic.

– As for legal needs, I do not see a problem. I have described how the legal system is treated in the market because the exercise of rights is inherently a personal good. I am someone who benefits from protecting their rights, and I am willing to pay a private company for it. But protection doesn’t work that way, and that’s going to be a problem.

Some time ago someone pointed out to me a piece of news about one of the Baltic countries, probably Estonia, where training for guerrilla warfare is a popular sport. They seem to have a fairly well organized system, which is subsidized by the state, but this education is mostly done privately.

In this way, through charitable donations, it was possible to support a small core of professionals who coordinated and then organized the war and the battlefield. But this is a difficult problem, and I would be much less optimistic about applying my system to Estonia than to the US, because the US has neighbors that are militarily weaker and therefore do not need to build up their power to defend themselves. . In addition, neither the Canadians nor the Mexicans gave any indication that they wanted to invade America. Therefore, I think that my system will work best in a country where there are no powerful aggressive neighbors.

In 40 or 50 years, we may have a much smarter artificial intelligence that surpasses human levels, and we better hope it loves pets.

“Actually, I discussed one of the problems with technology in my book, where I argued that we could very well end up creating programs for our computers that are equivalent to humans. Of course, they won’t be very human-like, as there are different parameters to consider, but they will be as intelligent as we are. And it’s especially scary.

The problem with artificial intelligence is that these programs are getting smarter, and we are not. So if in 30 years we get human-level AI, then in 40 or 50 years we will have much smarter AI that surpasses human level, and we hope they love pets.

There are other concerns as well, for example several scientists I know are worried about the possibility of a single artificial intelligence achieving superhuman intelligence. Eliezer Yudkowsky’s reasoning states: “A link maximization program that used an AI program should maximize link output while simultaneously making AI smarter and humans smarter. Ultimately, however, we have an artificial intelligence so smart that it can control the world by tricking people into doing what it wants.” So it turns out that a person is not the best way to produce fasteners. Thus, artificial intelligence is replacing people.

In the human body, the existence of consciousness is the first fact that we are aware of. Everything else is filtered through it. This filtering probably cannot be done by simply programming the computer.

So far, people have been impressed with ChatGPT and its descendants. All these programs are not like people at all. They look like smart ways to use human labor, i.e. all text imitates human work. But this is quite different than having the physical being of a person.

“If you mean literally a robotic world, a world in which we have been replaced by artificial intelligence, I assume that they will have the same coordination problems that we have. In an imaginary future where there will be no humans, only robots, these robots will probably trade with each other, make contracts with each other. They can even create robotic companies, since the fundamental and logical problem of recruiting individual actors is that each of them has its own goals that it is trying to achieve, and they need to be coordinated.

Well, I don’t see much of an issue that artificial intelligence raises, as they are just machines. Interesting questions arise when AI becomes human, if ever, and then we have to ask ourselves, how do we integrate them into our legal system? Are they slaves, i.e. is Excel a slave? Will it be the same with the word, which sometimes behaves like a disobedient slave, doing what I do not want? But since they are not people, then maybe they are just slaves. Also, people have had slaves in the past, which is not a very attractive system.

On the other hand, if this human program is really running inside your computer and you can turn it off whenever you want, maybe this “person” should do what you tell him to do.

The more interesting issue is the creation of a world where they are also treated as human, where you have a legal system that somehow takes into account the differences between a computer program and a flesh and blood human.

“We need to think about if robots are people, if artificial intelligence is about people, because a robot basically involves a car with a body. And I may or may not have a body, but if the AI ​​does not reflect the human, then we have an economic system of people using only machines.

Hypothetically, if people have their own system, then we will have an economic system in which people will buy things from AI, and AI will buy things from people. And, you know, people will use computers, and computers will hire people, and because today we have men and women in our system, and we manage to integrate them both into a single economy. Is the computer really that different from my wife? Probably no.