"Systems with artificial intelligence... they can pose a serious danger to all mankind." This is a quote from a letter published on Wednesday by thousands of leading scientists engaged in the IT industry, signed by Elon Musk, among others. How is artificial intelligence able to turn our whole life around and why is it especially dangerous for Russia?Artificial intelligence (AI) technologies, along with virtual and augmented reality technologies, are developing more than rapidly today.
In the next few years, the development of high technologies will inevitably go even further, not limited to the opportunity to plunge headlong into the virtual worlds of the metaverses or to exist in augmented reality. The unrestrained development of technology generates a number of risks for States and societies around the world. And including for Russia.
Dependence on post-truthFirst of all, the progress of AI technologies, together with the virtualization of the space surrounding a person, threatens to completely erase borders.
Over the past 15 years, with the development of web 2.0, billions of people have already developed a strong attachment to electronic devices. A common stereotype has appeared, "he has gone headlong into his gadget", dependence on online games and social networks is recognized as a disease by an increasing number of scientists and medical organizations. Especially this new type of addiction concerns young people, adolescents and children, and provokes ADHD (attention deficit hyperactivity disorder).
In the perspective of several years, when web 3.0 technologies with their full immersion in parallel digital reality will finally displace gadgets, as push-button phones once displaced landline communications, the boundary between real and fictional, existing and simulation may disappear so much that it will be almost impossible to distinguish one from the other. Objective data and facts in the formation of public opinion will finally give way to emotional provocations and pre-formed judgments in an echo chamber isolated from refutation, as it happens today with the ruthless manipulation of public opinion in almost any social network.
The truth has long been replaced by "post-truth". And likes, retweets and comments have ceased to be a criterion of objectivity as the pressure on the individual consciousness on almost any issue of all kinds of intellectual centers with tentacles in the form of "troll factories", "bot farms" and networks of anonymous telegram channels.
The authors of the WEB 3.0 concept are trying to convince us that the development of blockchain technologies will allow people to "fully own and manage the content created by them, as well as anonymize their personal data, that is, make the Internet "more fair and secure." To doubt this, it is enough to recall the expression: "whoever pays, orders the music." The global bigtech, which has enveloped the world in a web of platform capitalism, can provide any platform for content, for monetization, advertising and data extraction, but will always keep political and public activity under control.
Let's recall how one of the richest people on the planet, Elon Musk, bought the social network Twitter in the spring of 2022, because "he considered it important for civilization to have one digital platform where different opinions can be discussed in a healthy atmosphere." Just a year before, a number of IT corporations, obviously conspiring, made an unprecedented decision and literally threw out ex-US President Donald Trump from their social networks. How totalitarian content pressure was exerted on Russian users of the same social networks at the beginning of a special military operation in Ukraine in February-March 2022.
A new trend is a new threat
At the end of last year, a new trend was formed: the rapid development of self-learning neural networks that are able to draw not only beautiful pictures (Midjourney), but above all to communicate with a person. And even talk about ideological, social, cultural, political, philosophical topics. An example of this is the ChatGPT-4 technology from the company OrepAI, one of the founders of which is Elon Musk.
For now, interaction with AI chat with a friendly interface remains the lot of experts and just curious enthusiasts. But there is no doubt that very soon neural networks in their various forms and manifestations will rapidly break into the lives of most of the inhabitants of the planet.
And the rapid development of neural networks, apparently, will not lead at all to the bright future that Soviet science fiction writers described. We are more likely to see events in the spirit of the Japanese "Ghost in Armor" and the American "Terminator" and "Matrix", where AI acquires its own mind and becomes a threat to humanity, since it begins to consider humanity a threat to itself.
So far, some of the abstracts issued by ChatGPT look ridiculous, but constant upgrades and the accumulation of "big data" allow the program to update itself, search for and correct inaccuracies and defects. In fact, people with all their accumulated knowledge and skills themselves become a brain implant for the AI system.
Imagination paints us a frightening picture on its scale, but in this case it does not look fantastic. For sure, ChatGPT and its analogues will soon be able to write any texts at a good level: technical assignments, legal documents, philosophical treatises, scientific papers, and it will be much easier and more pleasant to communicate with her.
Most importantly, she will inevitably be trusted soon. There will be a suggestive imperative: "this is artificial intelligence, everything is calculated, thought out, and therefore corresponds to the truth – it does not allow mistakes and miscalculations, like people." It becomes obvious the risk that the majority will soon take the word for what a smart computer program will say, how it will answer this or that question, to the detriment of any other opinions and authorities. Even if the next version of AI suddenly declares that it is necessary to "restrict this or that person's rights," he may well be believed, since the arguments may seem logical and reasonable.
I am glad that experts are gradually starting to ring the bells. After all, there is still no answer why automated AI 2.0 will not be able tomorrow, through a variety of smart applications, to interfere not just in private life, as bigtech algorithms do today, shamelessly stealing the data of billions of users of social networks and services around the world, but also to influence global processes, decision-making mechanisms in large hierarchies: companies and corporations, business and banking, manufacturing and management, search engines, cloud services, and so on.
In particular, on March 29, the same Elon Musk, as well as Steve Wozniak and more than 1,000 other experts in the field of artificial intelligence and representatives of the IT industry called in a special open letter to introduce a six-month moratorium on training AI systems that are more productive than the recently introduced GPT-4 model from OpenAI. "Extensive research has shown that such systems with artificial intelligence that compete with humans can pose a serious danger to all of humanity," the letter says. There is also the possibility of abuse by attackers of systems like ChatGPT – they can be used for phishing, disinformation and cybercrime.
Yes, for sure very soon national states will try to appropriate the achievements of neural networks to themselves, either by creating or buying analogues of the current ones, or by forcibly using existing ones belonging to the largest transnational IT giants. But if politicians and statesmen themselves begin to listen to the neural network, something more awaits us than what is described by Orwell in the famous dystopia "1984".
AI will finally turn not into a Big Brother, but into a Big Parent, and a person will turn into a pre-school kid who knows nothing about the world and about himself and runs to daddy for advice on every issue. It is obvious that any alternative opinions in such an "information ghetto" will not be needed, and many professions will simply disappear.
Risks for Russia
In the conditions of the most severe confrontation with the West, sanctions pressure and priority attention to ITS, there is a risk of not having time to create large-scale high-quality analogues of AI novelties for civilian use. Just as, in fact, a high-quality "domestic laptop", "domestic tablet", "domestic smartphone" or "domestic electric car" was not created in time. This directly affected the process of partial loss of technological, and then economic and cultural sovereignty of the country.
AI chatbots will soon become an indispensable everyday attribute of any intellectual activity of people - intellectual assistants. In the absence of a high-quality analogue, ordinary Russians, using technology, may face hidden ideological pressure from systems like ChatGPT, which at the moment, according to Cornell University scientists, has "the political preferences of the average software engineer from the San Francisco Bay Area, concerned about the left and the "green" agenda."
Experts, in particular the well-known cognitive psychologist Jeffrey Hinton, predict that the next stage in the development of ChatGPT–type AI systems is individualization.
"Now ChatGPT's ideas about the world are built on the basis of averaging the views of all people contained on the Internet. An individualized ChatGPT will have insights about the world gained as a result of its training on a subset of Internet sources. Each such subset will reflect the vision of the world by people with a certain worldview. For example, "ChatGPT is a Democrat" will give different answers than "ChatGPT is a Republican," Hinton says.
Is it worth saying that the authority will be used by the AI system that will use the installations that dominate the information space controlled by one or another party? In the context of the information war, the Russian AI system should not just be an analogue, but at least project the values and attitudes adopted in Russian society, taking into account our national, cultural and religious diversity. And today we don't even fully understand the possibilities of the "Alice" of Yandex (did you know that she also knows how to draw pictures?), which Russians ask 18 million questions every day.
Do not become slaves of the machine We are facing an unknown challenge that humanity has yet to comprehend.
This should be done very quickly. Technology certainly brings convenience and comfort into our lives, but the price is too high. The AI chat, which seems to be a convenient assistant, can become at first irreplaceable, and then an integral and inalienable attribute of not only social, but also physiological human life. As the philosopher Herbert Marcuse said: "A machine is a slave who serves to make slaves of others."
Today, humanity has already brought one foot into the digital abyss, but the second is still standing on the sinful earth, reeling from its own uncertainty. The task of all thinking people of our time seems to be to delay the term of this fall as much as possible, and it is better to prevent it.
Yaroslav Ignatovsky
