Is it possible for AI to rise up against humanity like in Terminator
Artificial intelligence is becoming a decisive factor in the international technological race. About the same as once the great powers competed, which of them would become the first owner of nuclear weapons. Moreover, there is already a perception that artificial intelligence is even changing the global power dynamics. But is artificial intelligence good for humanity? Adviser to the Minister of Internal Affairs of Russia, retired Major General of Police, Doctor of Law, Honored Lawyer of the Russian Federation, Honorary Officer of the Ministry of Internal Affairs of the Russian Federation, ex-head of the Russian Central Bureau of Interpol, author books on combating organized crime Vladimir Ovchinsky.
We really need artificial intelligence (AI). It just so happened that natural intelligence, that is, the natural human brain, simply cannot cope today even with those elementary tasks that a primitive calculator can easily solve. Yes, a person is able to calculate something in his mind, knowing the multiplication table. But how fast?
This is what Paul Sharre, vice president and director of research for the Center for a New American Security, says. He published an interesting book, which was published in the United States at the very end of February — «Four Battlefields: Power in the Age of Artificial Intelligence.»
– Charré is a strategic planner who worked in office of the US Secretary of Defense. He participated in the development of the use of unmanned and autonomous systems and new weapons technologies, and determined the Pentagon's intelligence policy. Serious specialist.
Charré explores how international competition for the most powerful technology, artificial intelligence, is changing global power dynamics. In his opinion, the outcome of this struggle will determine the next global superpower.
— Very persuasive. He argues that the battle for AI power will revolutionize the world's military and economy. He was recently interviewed in Insider about the issues reflected in his book, which concludes on four key battlefields of global competition: in the field of artificial intelligence or data; computing equipment or computing; human talent; institutions necessary for the successful implementation of AI systems.
– Yes, and he specifically talks about the military conflict in Ukraine. He candidly admits: “I think it’s likely that the US Department of Defense and the intelligence community are using AI to analyze satellite or drone images, and while they don’t publicly admit it, it’s possible that it plays a role in the information that The US is sharing with Ukraine. We have seen other examples of the use of artificial intelligence technologies directly by Ukrainian forces on the ground, in particular by some civilian drone operators. We also see in the war in Ukraine the importance of logistics, for example, and maintenance operations, and this is that the main part of what the military does on a day-to-day basis is moving people and things from point A to point B — this is very similar at Walmart or Amazon. It's a place where the military can improve their logistics, readiness, finances, personnel and maintenance.”
— That's the whole problem — the speed of making the right decision. Yes, a person will find a solution — but when, how quickly? How much time will it take to debate, agree with the authorities, to check certain options? Measure the distance on the same map with a curvimeter? Even just to count something in your mind or in a column? And often moments decide everything.
Charré makes this argument: “As for military achievements in the field of warfare, there are also several examples. In the DARPA Alpha Dog competition, the goal was to create an AI agent that could achieve superhuman performance in a simulation against a human in dogfight. The AI did well, went head-to-head with an experienced Air Force pilot and completely crushed the human pilot 15-0 — the human didn't fire a single shot against the AI.
In particular, head-to-head shots are effectively banned in human pilot training because there is a high risk of collision if the pilot tries to maneuver the aircraft while you are racing each other at hundreds of miles per hour. And in any case, it is extremely difficult to do and requires a superhuman level of precision, but all this was not a problem for the AI agent. He can fire these shots in a fraction of a second while avoiding collision. And the AI agent has learned to do all this on its own — it hasn't been trained to do it. The AI system that won was simulator-trained and had over 30 years of simulated flight time. It was one of the things she just learned on her own from all those years of simulated aerial combat.» That's the whole point of the AI - the universal soldier.
“This is what Sharre says: “There are a lot of places where AI can have a pretty significant impact on economic productivity over time. For example, special applications that can improve finance, medicine, transportation or other industries. Self-driving cars need to be at the stage where they are truly efficient and viable on the roads. The use of AI in medicine, for visualization, such things can be very useful for the whole society. But one of the most exciting things about AI is its ability to improve performance in all kinds of places.”
— A lot. American experts argue that AI is, in fact, not intelligence — it is a prediction. With larger language models, there has been an increase in the machine's ability to predict and execute the desired outcome. Thus, Paul Sharre believes that “AI has great potential to bring benefits in various sectors, including education, healthcare and the fight against climate change. For example, FireAId is an artificial intelligence-based computer system that uses wildfire risk maps. In healthcare, AI is being used to improve patient care through more effective prevention, diagnosis, and treatment. Increased efficiency also reduces health care costs. Moreover, artificial intelligence should radically change – and ideally improve – care for the elderly.”
And what is very important is that artificial intelligence protects people from criminality, that is, it ensures the safety of ordinary people. Charré cites the Chinese experience: “China has half of the 1 billion (!) surveillance cameras in the world, and they are increasingly using artificial intelligence tools such as face recognition or gait recognition to determine who people are based on their models. gaits and how they walk. And in combination with other types of data, such as license plate data, calls or geolocation data on phones and people's shopping behavior, to track Chinese citizens.
China exports most of its technology abroad, so 80 countries world have acquired Chinese surveillance technology.”
— There are such concerns. But this requires not only technological limitations, but also a carefully designed legislative framework. After all, AI, no matter how perfect it is, — albeit self-learning, in itself — is a soulless and devoid of the concept of morality, religion and ideology mechanism. All these properties are invested in it by the developers of the program. That is, hatred of Russia or, say, the United States or China may well be embedded in the electronic brain of AI. And here, alas, all responsibility falls on the human developer.
Agree, it doesn’t matter that the MP-40 submachine gun (it is mistakenly called “Schmeisser”) was once made at a factory in Nazi Germany. What matters is whose hands it was in. It's one thing — in the hands of an SS man who shoots peaceful ghetto hostages. The other is in the hands of a Soviet partisan.
— Writer Zoshchenko once remarked: “In order not to think badly about people, it’s better not to think about them at all.”
< p>
Are you hinting at a Hollywood blockbuster about the terminator? You will not believe it, but I personally do not rule out the so-called “rebellion of the machines”, no matter how fantastic it may seem.
And Paul Charré speaks about this: “Microsoft and Google recently publicly deployed chatbots with artificial intelligence who were not ready. And the problem is not that Bing declares its love for users and says that whoever chats with should leave his wife to be with him — I mean, that's pretty strange.
Agree, it's really strange. Your laptop in all seriousness invites you to live with him instead of his wife. How not to go crazy here?
But here I agree with the same Charré, who comments on this predictable horror: “For me, this is something that should make us think when we think about how we use these AI systems in more real world applications. At present, the risk of a chatbot hurting someone's feelings doesn't stun the world, but as we see AI integrated into applications with greater implications, we want to make sure these systems will do what we get from them. want».
That is, again, it all depends on the person-developer.
— Unlike you, the Americans are not so tragically disposed. Don't take Hollywood products seriously.
Same Charré said cheerfully: “I wouldn't say I'm optimistic about technology. I mean, I'm pretty optimistic about where AI is heading in terms of capabilities. I just think we're seeing tremendous progress and I don't think there are any signs of it slowing down in the near future. But there are many risks associated with AI. I would just say that I am optimistic about society's ability to cope with these risks.”
Whether we like it or not, it is. It is important for us to make AI become not a new weapon, such as nuclear, which would affect the balance of power, but a tool of creation. We already have a means of destroying each other and, in particular, ourselves. Let's think about how to save us all. Maybe artificial intelligence will suggest the correct algorithm?