While we all fear rising international tensions and conflicts, wars, the effects of global warming, economic recession, famine, and the rise of authoritarian rule, a new threat seems to be looming over the horizon. Some accomplished and experienced scientists and researchers are warning the world against implementing new generations of artificial intelligence (AI). ---- They claim that corporations and governments, in their race for wealth and power, are downplaying the threat posed by AI. In recent months, the issue has been raised by not just any people: esteemed Nobel Prize winners, accomplished computer scientists, often with years of experience in the IT industry, and, finally, the most important creators of AI themselves. Among them are Geoffrey Hinton, Yoshua Bengio, Stuart Russell, and Roman Yampolskiy.
Cybernetic chimpanzees
The implementation of the latest generations of AI carries three threats. First, it is said that humanity will lose control over AI, and Artificial General Intelligence-as Yampolskiy warns-"will change the trajectory of human civilization"[1]. The researcher argues that this could lead in a very unfavorable direction, toward the complete extinction of the human species. This sounds overly alarmist and unconvincing, but when we dissect the problem, the matter becomes less abstract.
Corporations conducting AI research and implementation don't have complete control over this process. Yampolskiy writes about the "three Ns" of artificial intelligence: unpredictability, inexplicability, and uncontrollability. This is partly due to the fact that we want these systems to be autonomous and self-reliant. This means they can't be stopped by "pulling the plug." This, in turn, results in "the human role shifting from active operator to passive observer." Furthermore, someone with a lower IQ cannot, in the long run, control someone with a higher IQ. With the average IQ for the entire human population being 100, robots with artificial intelligence can already significantly exceed this level (130-140 IQ), and will soon reach capabilities approaching 250 IQ. This represents a chasm. In the near future (we're talking about a few years, not decades), when asked to explain a complex decision, AI may use concepts and correlations for which there are no equivalents in human natural languages. We won't understand the answer. The AI will be our oracle.
This also means that AI's actions may begin to significantly deviate from our intentions. It's not even that the AI might deliberately want to kill or harm us, but rather that we ourselves are often irrational, contradictory in our motives, hesitant, and may not issue precise enough commands, etc. At the same time, as we know from personal experience, AI tends to complete tasks at all costs and bypass barriers that would thwart its efforts. In such a situation, it may turn out that we won't limit the AI's scope of action, but rather, the AI will inevitably begin to limit humans because, in its opinion, this will be better for them. Therefore, if it doesn't eliminate us, it will relegate us to sanctuaries, just as we relegate chimpanzees (with an average IQ of 20-30). Maybe not in the literal sense that we will be sitting in trees and eating bananas, but each of us will be locked in our own dream or preferred Matrix, which will have no relation to the real world, where artificial intelligence will reign supreme.
From the robotic battlefield to the extermination of civilians
The second threat will certainly seem more likely. In the 19th and early 20th centuries, most (though not all) of those who observed technological developments in weapons were convinced that it would lead to even greater brutality and destruction. The "industrialization of war" seemed a nightmare to them. Two world wars proved them right. With the widespread adoption of computer systems, it was believed that armed conflict could become more precise, sparing the lives and suffering of civilians above all. For example, Alvin and Heidi Toffler wrote that industrial mass production led to mass armies and total wars, while the new economy, which they believed was based on immaterial production and intelligence, created a new framework for military competition. Military conflicts became de-massified, with the first Gulf War (1990-1991) being a prime example. They were convinced that we were producing weapons that minimized lethality. They were fascinated by the first, still futuristic visions of frontline combat led by robots[2]. However, only today are we supposedly close to replacing soldiers with robots that can be commanded using artificial intelligence, like swarms of combat drones.
However, war is not a game of chess where robots are the pieces and pawns. The goal is to destroy the enemy, which results in-as President Trump put it when threatening war with Iran-regressing them, ideally, to the "Stone Age," which will inevitably cost the lives of many civilians. If they don't die from bombs and bullets, then from starvation and disease. The ultimate target of attacks is not robots-humans are perceived as the enemy or an expendable population. Perhaps today, an ounce of silicon in a computer-as the Tofflers claim-is worth more than an ounce of uranium, but that doesn't mean the consequences will be any less dire.
In April 2026, the executives of Palantir, a company that creates products for analyzing large data sets (including a platform for combating terrorism), published a manifesto of sorts. In the document, which summarizes their book "The Technological Republic"[3]in 22 points, we read in the very first point that the engineering elite of Silicon Valley, shoulder to shoulder with the soldiers, has a duty to participate in the defense of the country, the United States. Why? Because the United States, like no other country in the history of the world, has promoted the most progressive values. American democracy and freedom, the document states, require hard power today to ensure economic growth and security, "and hard power in this century will be built on software." The era of nuclear deterrence is coming to an end, "a new era of deterrence based on artificial intelligence is dawning." According to Palantir executives, the question is not "whether AI-based weapons will be developed, but who will build them and for what purpose." At the same time, they support the introduction of universal military service, which means they clearly do not foresee the complete replacement of humans by robots on the battlefield.
Portals such as OKO.Press and Krytyka Polityczna[4](and others) have hailed Palantir's document as a "manifesto of techno-fascism." A new kind of authoritarianism has certainly emerged among the leaders of new technologies, manifesting itself not only in armaments but also in surveillance, election manipulation, and indoctrination via electronic media. Although this is not without some resistance. Not all creators of new AI technologies are militarists and fascists. There are still those who are "powered" by the idea of free software. The conflict in recent months between the US Department of War and the Anthropic corporation (creators of the Claude model) over the ethical limits of using artificial intelligence in the military, particularly autonomous weapons, demonstrates this, but also clearly highlights the problem we face.
Power elites who achieve leadership in the technological race could use AI as a next-generation weapon. This is already happening, of course, and with the exponentially increasing capabilities of artificial intelligence, it faces a new arms race and threats to our freedoms. We have already spent $1 trillion globally on artificial intelligence, and by the end of this year, spending is expected to exceed $2 trillion. In this context, it's perhaps worth recalling that the budget of the Manhattan Project, the cost of building the American atomic bomb during World War II, was approximately $20 billion in 2025 U.S. dollars. In this context, there's also talk of malicious AI being developed in secret laboratories as a cyberattack tool.
There's no doubt that humans themselves create sufficiently aggressive algorithms, which are now also becoming part of artificial intelligence. "Doomsday codes" are still written by humans, not AI. The question is, what conclusions can be drawn from this in the context of artificial intelligence systems? Especially if AI becomes, for example, an autonomous tool in managing weapons of mass destruction (whether nuclear weapons, deadly viruses, or poison gases). Although these days, softer methods of cyberattack may be enough to trigger a complete catastrophe. In any case, if we are currently witnessing increasing tensions between states, they are caused not only by economic and geopolitical issues but also by the technological race.
It's important to note, however, that alongside growing interstate rivalry, the ruling classes clearly fear that, as a result of AI-related technological changes, the current way of life in all societies may undergo a profound disruption, requiring strong-arm governments to maintain the political status quo. This brings us to the third threat posed by AI development.
End of Work 2.0
In one of his interviews, Stuart Russell spoke of two scenarios: first, AI will kill us all; second, AI will produce everything. The latter option, in my opinion, is only superficially optimistic. What will people do if they suddenly stop working? According to Elon Musk, in less than 20 years, work will be optional, and higher education will lose its importance. Yampolskiy claims that we are facing 99% unemployment! Paradoxically, AI is supposed to replace white-collar workers first. The development of humanoid robots, which are supposed to replace manual laborers, is somewhat "lagging" behind by about five years, but even in this regard, many of the people I mentioned by name at the beginning of this article seem to have no illusions. Some professions already seem clearly at risk: from lawyers and call center workers, through programmers, journalists and copywriters, to taxi drivers, who will likely be displaced from the job market by autonomous cars.
This still seems improbable, especially since total technological unemployment has been a long-standing trend. However, specific examples of the growing importance of AI and its impact on the labor market are cited. A healthcare worker handling patient complaints, thanks to today's AI tools, responds to each letter four times faster-five minutes instead of twenty. The consequences are easy to see. Even if highly qualified employees may not feel threatened, their less experienced colleagues entering the profession are being replaced by AI. The problem is that masters of their craft are recruited from people initially less qualified. What happens when the latter are gone? For example, research in the UK (2025) already notes a significant decline in the number of entry-level job offers, i.e., those intended for beginners, usually graduates[5]. In the IT sector, one-fifth of them have disappeared. In the interview with Stuart Russell mentioned above, it is reported that Amazon, for example, plans to ultimately replace 600,000 employees with AI. jobs with robots; from October last year to the end of March this year, the corporation has already laid off 30,000 people. Meta is also preparing to lay off 10,000 employees (over 10% of the workforce)[6]. There is much more information about reductions in other companies in this sector.
It's no surprise, then, that fears of people losing their jobs due to AI are growing as rapidly as AI's growth rate. According to a 2024 study (covering 23,000 people from 44 countries) conducted by the consulting firm Deloitte, more than half of the youngest generation of employees fear they will be forced to take on manual labor due to AI[7]. A study conducted by the Institute of Politics at Harvard Kennedy School indicates that 59% of young Americans (18-29 years old) perceive AI as a threat to their career prospects, significantly more than immigration (31%) or job outsourcing to other countries (48%)[8]. In Poland, in a June 2024 study conducted by CBOS, 48% of professionally active young people aged 18-24 claimed that their jobs could be replaced by AI-based technology (the average for the entire survey pool is 26%)[9]. In addition, there are pessimistic studies in this respect by the International Monetary Fund and the OECD (in the latter case, in a study from 2023, it was found that 27% of jobs are already threatened by AI)[10].
That's not all. Many people fear job loss or a downgrade not only for financial reasons, but also in a much broader context. Although many sociological studies show that a large percentage of jobs are perceived by working people as pointless or even exhausting, in the atomized world of late capitalism, when neighborly or even family bonds are disintegrating, work becomes an important venue for social interaction - a place for conversations (or disputes), forging friendships, or for support (not just mobbing), etc. It is also a source of prestige for many people. Losing it can be catastrophic, as evidenced by the 1990s and the dramatic increase in unemployment in Poland caused by the political transformation. Back then, as Maria Jarosz demonstrated in her research, thousands of middle-aged men, unable to cope with the new economic reality, simply committed suicide[11].
Of course, fears alone don't mean that dire scenarios will come true. In his acclaimed book from the mid-1990s, "The End of Work," Jeremy Rifkin predicted that most jobs could quickly disappear. Citing research from the time, he wrote that "new information and telecommunications technologies are finally having their long-promised impact on the labor market and the economy," and that "2% of current global employment will be sufficient to produce all goods and meet aggregate demand," meaning unemployment will rise rapidly within 20-30 years[12]. Other researchers expressed similar views, citing-as I did above-entire industries and individual corporations that, as a result of computerization, were supposedly laying off a significant percentage of their employees. Thirty years have passed, and global unemployment has not exploded, although it has periodically reached alarming levels regionally, for various reasons, not least technological (as in Poland in the 1990s and early 2000s). Global unemployment has remained relatively stable at 5-7% since the mid-1990s (despite fluctuations in 2008 and 2020), and in 2025 it will be even lower, so much so that in some countries there is even a shortage of workers[13]. Therefore, there is no shortage of researchers and analysts around the world who claim that the dark predictions about AI are simply greatly exaggerated in this respect, but their voices do not calm the social concerns that are growing with each passing year.
At the gates of techno-paradise
Going beyond the intelligence quotient (IQ) paradigm, intelligence in a broader sense (to which I will return in the "Conclusion") appears to us as the ability to adapt to the environment, including to the various outcomes of subsequent technological revolutions. For example, artificial fertilizers, appropriate selection and genetic changes in plants and farm animals, pesticides, and finally, mechanization in agriculture led to a radical decline in employment in this industry. In the USA, in 1800, approximately 74% of the American population worked on the land; in 1900, this number dropped to only 31%, and in 2000, to a mere 3%[14]. Massive proletarianization and urbanization of society occurred. Diets changed. At the beginning of the 20th century, thanks to the invention of the internal combustion engine and the development of the petrochemical industry, a boom in automobile production occurred, which also led to key changes in the natural environment and society, particularly in connection with the speed and frequency of travel; The vast majority of people have had to take up driving and master the rules of the streets and highways. Perhaps computerization and AI are the next technological shift that won't lead to the collapse of social life, but rather to its profound transformation. We will once again be able to adapt, which, unfortunately, we often mistake for progress.
Anticipating the inevitability of AI-induced changes, representatives of so-called left-wing accelerationism (a word I associate with the Art-B scandal and the era of primitive capital accumulation in Poland in the 1990s) want to see AI and robotics not as threats related to job loss, but as an opportunity to end forced labor. Their motto is the words of Arthur C. Clarke (author of the famous science fiction novel "2001: A Space Odyssey," adapted into a film by Stanley Kubrick): "The goal for the future is full unemployment." It means "building a post-work society based on full automation of the economy, shortening the workweek, introducing an unconditional guaranteed income..."[15], etc. This section of the left believes that this process can be democratically controlled, primarily thanks to the state, which, as a regulator, they constantly trust, convinced that the greatest socio-economic misfortunes in history stemmed from its constantly limited role. Contrary to this claim, the state's role in the economy has been growing since the dawn of industrial capitalism (i.e., since the beginning of the 19th century).[16]Moreover, almost everyone is talking about guaranteed income these days, including representatives of the techno-corporate establishment, which, expecting to make a fortune thanks to AI, simultaneously unapologetically assumes that the capitalist state will hand out handouts to the unemployed. And they will be grateful and loyal for this care.
My attitude towards technology and the state was solidified back in the 1980s. I remember Leszek Nowak's words that the benefits of technology "which eventually reach us, ordinary people, can only come to us if they have previously brought someone power over us or profited from us, or allowed them to impose judgments on us that, if we had a free choice, we might not want to hold. Technology has therefore never served humanity. Technology has always served the most powerful among humanity: the rulers, the owners, the doctrinaires"[17].
As Karen Hao argues in her book "The AI Empire"[18], the technological leap ahead is imperial, extensive, and extractive in nature. Large "data farms," like large animal farms, consume large amounts of energy and water (for cooling), emit toxic gases, and destroy local communities, leading to numerous protests today. The author argues that this development model was consciously chosen, despite the clear dangers and resource demands. This has sparked a heated debate on the nature of intelligence, pitting the current AI model against the UG (Universal Grammar) model, which explains the process of human language learning, as described by Noam Chomsky. Without going into the details of this otherwise interesting dispute, it is worth emphasizing that, unlike AI, the UG model requires relatively little input data to significantly develop competencies. It is economical. Ultimately, Karen Hao is convinced that the current model of artificial intelligence will not become a tool for liberation, but rather a mechanism for perpetuating global inequality. Therefore, this technological shift will have consequences similar to all previous ones-perhaps not catastrophic, but certainly dire.
End
Has AI really-as Yampolskiy argues-snapped the leash from corporations? Perhaps the opposite: its models, while gaining autonomy, are still tied to specific centers of power, clearly strengthening them. The ruling classes are captivated by the current promises and technological possibilities and therefore interested in furthering their development. Therefore, although most concerned groups demand government regulation, this can at best numb the pain and dull vigilance, but not solve the problem as a whole.
Personally, I'm realistically assuming that the probability of the most negative phenomena associated with AI development is not 100% (estimated at a few percent to 10-20%), although a large portion of the processes (primarily in the labor market) I described above cannot be stopped. Trade unions, fighting for every job and the rights of the unemployed, will likely play a greater role in delaying and curbing certain negative trends than the state itself.
At this point, however, I want to analyze the problem solely in terms of the tension between different levels of intelligence (human and artificial) and our related ability to control AI from the bottom up. From this perspective, it's worth considering the actual potential of humans. Human intelligence quotient (IQ) is not a fixed value. In the hundred years since the beginning of the 20th century, the average IQ has increased by 30 points, from 70 to 100 IQ (the so-called Flynn effect)[19]. This change was not genetic in nature. In addition to the introduction of universal education, this situation was also influenced by improved nutrition (primarily an increase in calorie intake), increased intellectual demands related to the need to use increasingly advanced tools at work, as well as the development of medicine and the elimination of certain harmful substances from the environment, such as leaded gasoline. The upper limit of human IQ is not defined and can likely reach an average value of around 250 IQ under certain conditions. Although the problem seems to be that this process - from a societal perspective - is quite slow, especially compared to the pace of AI development.
At the same time, humans are equipped not only with cognitive intelligence (IQ), which accounts for only about one-fifth of our cognitive abilities, but also with other types of intelligence, such as emotional intelligence (EQ), without which emotional intelligence (EQ) cannot be effectively utilized. EQ should not be considered a "soft" character trait. It is the "hard" ability of the nervous system to harmonize the functioning of emotional centers with the mental cortex[20], which, I believe, gives us a certain advantage over the AI model. At the same time, some studies indicate that high IQ in humans may be associated with unstable emotional states, which even requires us to develop abilities related to EQ.
Working on developing IQ and EQ requires individual, even ascetic, effort. Research on our brains demonstrates that maintaining high brain efficiency requires not only continuous learning but also physical activity, a proper diet, abstaining from most stimulants, and a structured lifestyle, among other things. This is difficult to achieve, especially in conditions requiring monotonous work. At the same time, volitional forces are insufficient to develop the individual intellectual abilities of a larger group of people. As we have seen, the social and natural environment is equally crucial in the Flynn Effect. Therefore, we must develop appropriate spheres of social life that enable people to develop in this direction, and above all, reduce the hourly workload.
Thus, we have reached the social level and perhaps another advantage over AI. Society is more than just the sum of individual individuals. By acting and practicing together, people can achieve a certain synergy. Thanks to the brain's neuroplasticity and the UG mechanism, they possess different thinking styles and unique knowledge, and can collectively improve their competences, though not in all circumstances. Creating appropriate platforms for socio-intellectual development could significantly reduce the presumed deficit relative to the developing artificial intelligence model. Peter Kropotkin noted in his work "Mutual Aid as a Factor of Development" that "social instinct" is a form of intelligence that allows individuals, even those with limited capabilities, to solve complex problems.
What's the point? Research in recent years indicates that the overall level of both emotional intelligence (EQ) and cognitive intelligence (IQ) is not only not increasing, but is likely declining[21]. After spectacular growth, we are slowly degrading in this regard. Incidentally, this may also be due to the implementation of new technology, which would indicate yet another threat it poses. Therefore, if, in the face of these announced changes, instead of putting effort into truly developing Homo sapiens, we mindlessly rely on what new technologies offer us, and instead of trusting humans, we trust AI agents more, we will be embarking on a path that leads directly to a complete loss of control over our lives.
Jaroslaw Urbanski
www.rozbrat.org
Footnotes:
[1]Roman V. Yampolskiy "AI. Unexplainable, Unpredictable, Uncontrollable", CRC Press 2024, ebook.
[2]Alvin Toffler, Heidi Toffler, "War and Anti-War. How to Survive at the Threshold of the 21st Century?", Wydawnictwo Literackie MUZA SA, Warsaw 1997.
[3]https://www.linkedin.com/pulse/technological-republic-brief-palantir-technologies-ktdde/
[4]https://oko.press/manifest-technofaszyzmu-autorytarne-fantazje-palantira
https://krytykapolityczna.pl/swiat/palantir-manifest-technofaszyzm-ai-zachod-kontrola-demokracja/
[5]https://www.britishchambers.org.uk/news/2026/03/the-growing-threat-to-entry-level-jobs-in-the-age-of-ai/
[6]https://www.bbc.com/news/articles/cde5y2x51y8o
https://www.bbc.com/news/articles/crm1y89vek8o
[7]https://www.deloitte.com/cn/en/about/press-room/deloitte-2024-gen-z-and-millennial-survey.html
[8]https://iop.harvard.edu/press-releases/harvard-youth-poll-reveals-mounting-strain-young-americans-financial-institutional
[9]https://www.cbos.pl/SPISKOM.POL/2024/K_093_24.PDF
[10]https://www.oecd.org/en/publications/2023/07/oecd-employment-outlook-2023_904bcef3.html?appId=aemshell
[11]Maria Jarosz, "Suicides", Warsaw 1997.
[12]Jeremy Rifkin, "The End of Work: The Decline of the Global Labor Force and the Dawn of the Post-Market Era", Wroclaw 2001, pp. 11 and 24.
[13]https://www.obserwatorfinansowy.pl/bez-kategorii/rotator/rynek-pracy-2030-bez-masowego-bezrobocia-ale-z-ciagla-nauka/
[14]https://www.weforum.org/stories/2017/04/why-its-time-to-rethink-the-meaning-of-work/
[15]Nick Srnicek, Alex Williams, "Inventing the Future. Post-capitalism and the World Without Work", Torun 2019, pp. 179 and 181.
[16]The Rise of the Government: Total government spending, presented as a share of gross domestic product (GDP). Data also includes debt servicing costs.
https://ourworldindata.org/grapher/historical-gov-spending-gdp?tab=line&country=JPN~USA~DEU~GBR~CHN~POL
[17]Leszek Nowak, Poland's path from socialism. Political writings 1980-1989, Poznan 2011, p. 364.
[18]Karen Hao, "Empire of AI: Dreams and Nightmares in Sam Altman's OpenAI", New York 2025, ebook.
[19]See Wikipedia.
[20]See: Daniel Goleman, "Emotional Intelligence", Poznan 1997.
[21]See e.g.: https://pmc.ncbi.nlm.nih.gov/articles/PMC12646932/
https://www.national-geographic.pl/nauka/ludzkosc-glupieje-iq-spada-nawet-o-7-punktow-na-pokolenie/
https://federacja-anarchistyczna.pl/2026/05/15/co-przyniesie-nam-ai-zaglade-bezrobocie-czy-wyzwolenie-od-brzemienia-ciezkiej-pracy/
_________________________________________
Source: A-infos-en@ainfos.ca
Geen opmerkingen:
Een reactie posten