AI dangers: top 10 threats to humanity and life

“Whoever succeeds in establishing himself as a leader in artificial intelligence will become the master of the world.”

— Vladimir Putin

The race for artificial intelligence (AI) is on. State superpowers, high-tech companies, dedicated start-ups and other tech-savvy billionaires, all the dominant players in the global economy are investing heavily in this new technology in the hope of crushing the competition. The AI market is growing rapidly. It represented $241 billion in 2023 and could reach $511.3 billion by 2027. Les calls experts and associations calling for a halt or at least regulation of research and uses are ignored, despite the existential threats that AI poses to human societies and life on Earth. Here we have compiled a non-exhaustive list of the dangers of AI.

1) Mass technological unemployment

The mechanization and then the robotization of work have already destroyed many jobs (or profoundly transformed them, humans being reduced to an appendage or a cog in the machine). Coupled with AI, machines will replace even more humans. The novelty with AI is that not only is the machine bound to change, but this time it will no longer need the human operator to know what task to do.

This replacement of humans is possible for very technical construction, assembly or maintenance positions, but doctors, accountants and lawyers are also on hold. In creative and artistic fields, editors, screenwriters, writers, and cartoonists are also under threat. In particular, it is generative AI that threatens artists. Indeed, the first picture painted by an AI has just been sold at auction.

2) Decrease in creativity, impoverishment of social relationships and generalized stupidity

Among the services including AI offered to the population, conversational agents (for example ChatGPT) and other applications are multiplying, diversifying and their performance is constantly increasing. More and more used for writing, research, conversation, conversation, creating images or videos, these applications are sold as assistants in leisure or work allowing a saving of time and quality.

In addition to the obvious loss of diversity and originality of the content (since the results provided are based on what exists), this daily use can cause serious problems for users: loss of self-confidence, impoverishment of expression, learning problems, social isolation when the conversational agent replaces human interactions (in friendly relationships, parents-children, etc.). This is all the more true when the conversational agent used gives simplified, erroneous or made-up answers. The lack of perspective on this technology makes it difficult to fully assess its long-term effects, especially among the youngest. However, the work of the anthropologist and MIT psychologist Sherry Turkle already gives us a good general vision of the mental degeneration of the species accelerated by the development of computer science.

3) Manipulation of opinion: amplification of biases and false information

AI is giving fake news a whole new dimension. It is now within the reach of anyone to create or modify images, documents, and share them to as many people as possible in record time on social networks. Massively used during the countryside In the American election of 2024, false information, supported by almost undetectable falsifications, makes it possible to manipulate opinion and distort reality as desired. In the event of war, this tool allows targeted propaganda in real time. Used in a more personal setting, these fakes can be used for online blackmail or harassment.

AI is not objective. It includes biases. Bias may be due to malfunctioning or to information that the AI intersects to provide an answer and learn. A conversational agent can therefore provide false or non-objective information; an AI with decision-making power can choose the option that is not the best for what is required of it (which is more or less dangerous depending on the field of use).

4) Mass Technosurveillance and Propaganda

Algorithmic video surveillance (VSA), QR codes, GPS tracking, GPS tracking, smartphone tapping, etc. AI coupled with countless public or private connected objects, as well as satellite networks, make it possible to listen to, follow and log anyone, anytime and anywhere.

The most advanced large-scale experiment is the Chinese surveillance and social credit system. Integrating AI into such a system makes it possible to optimize surveillance, but also to establish a predictive system based on each individual's background. A tool of choice for total control of society and the neutralization in the bud of any questioning of the current regime; a control enabled by invasive technologies that destroy privacy, and by extension the very possibility of resisting.

5) Increase in cyber vulnerability

Cyberattacks are among the new tactics in War 2.0, but they are far from affecting only states and big businesses. The digitization of society and human relationships has increased the “attack surface”. Now, all actors in the economy — individuals, VSEs, SMEs, town halls, associations, communities — connected to the Internet network are exposed to this threat that was non-existent only 20 years ago. But the technocrats explain to us that thanks to technological progress, the world is safer today...

With the development of AI, attacks in cyberspace will multiply and become more destructive, resulting in a likely increase in instability locally and internationally. The threat to the globalized and interconnected technological system is all the greater since AI allows large-scale attacks on targets that were previously protected in so-called “hybrid” wars. AI also makes it possible to refine and extend scams, in particular by creating perfect falsehoods and optimizing and automated data processing.

6) Development of autonomous weapons

Even if fully autonomous weapons are not yet in service, the erasure of human decision in favor of arms automation is already underway. A reality. The State of Israel is already making use of softwares artificial intelligence to propose lists of targets to be shot down and autonomous weapons not controlled by humans. Even if the human link in the decision-making chain has not yet disappeared, the weakening of moral decision necessarily leads to a detachment leading to ever more wars. Bloody, where responsibility is increasingly based on the machine.

If autonomous weapons systems became widespread, any convention or treaty on war or weapons would become obsolete. Indeed, the law that regulates conflicts may be ignored since the machine will be held to responsible and no longer the human. The automation of weapon systems could also lead to Blitzz-Wars even to an uncontrollable escalation that would lead to nuclear strikes until mutual annihilation.

7) Destruction of the biosphere and climate

The massive development of AI is accompanied by an increased use of certain resources: metals for connected devices, Electricity and water for the increasingly numerous and powerful data centers. Electronic pollution associated with the development of digital infrastructure is also a problem. AI also generates rebound effects on many other sectors and contributes to increasing their environmental footprint (robotics, aerospace, “smart” cities or Smart City, social networks, etc.).

8) Concentration of power

The development of AI affects all areas of techno-industrial societies: medicine, transport, security, communications, defense, arts, etc. The AI race sees a Handful of firms High-tech dominating the sector. These firms (and their managers, shareholders or associates) therefore have an inordinate power of influence on the policies of a State as well as on the daily lives of its subjects. The risk of authoritarian drift is therefore considerable, all the more important if the State in question is already authoritarian or in the process of becoming authoritarian. AI is a powerful tool for maintaining and strengthening the established order, as we are already seeing in China, India or Russia.

9) Flight forward and techno-solutionism

In current discourse, firms and researchers justify efforts to develop AI by highlighting the solutions that this new technology could provide to contemporary problems. Techno-solutionist, or technocratic, propaganda consists in making the masses believe that technical solutions are effective in solving problems originally created by technical progress. However, the following points should be kept in mind.

First, the major threats to human societies all have the same origin: the techno-industrial system (generalized pollution, civilization diseases, wars, destruction of the biosphere, climate chaos, etc.). AI is the latest cancerous outgrowth, it depends for its existence on a whole tangle of systems and infrastructure networks (datacenters, submarine cables, power plants, electrical networks, electrical networks, mines, electronic waste dumps, etc.) which are themselves problematic for the survival of the human species.

Second, techno-solutionism only offers a few adjustments and optimizations in specific areas. It takes problems one by one and ignores their common cause. As a result, efficiency gains at the local level contribute to increasing the consumption of energy and materials at the global level. The 19th century economist Stanley Jevons highlighted this paradox that is now well documented, a paradox called the “rebound effect”.

Third, while techno-prophets promise that technological leaps will benefit everyone, in fact, these alleged advances will in fact only benefit a tiny portion of the world's already well-off population. The rest of humanity will not only not benefit from the “benefits” extolled by the promoters of AI, but will also have to bear the harmful effects of the development of AI (technological slavery to train AI, climate, biosphere, pollution, pollution, exploitation of resources, neocolonial policies, wars, etc.), further aggravating the inequalities that are already yawning (at all scales).

Finally, AI is and will always remain an insensitive machine, content to swallow, sort and then vomit data without understanding anything. Its added value can be doubted, especially as the reliability of the results is often questioned.

A video popularizing the existential threat that AI represents, very well produced by the Youtuber EGO.

‍

10) Risk of uncontrollable slippage

While the risks mentioned above can currently be more or less identified, quantified and studied, there is still a possibility of Unpredictable slippage with technology of this type. This risk of loss of control is all the more important with the mad rush of state and private actors to be the first to create General Artificial Intelligence (GAI), in other words a superintelligence that would surpass humans in all areas. It is impossible for us to imagine the consequences of the advent of global superintelligence. Sa operating logic would escape any human operator and its computing capabilities would be considerably higher to those of humans, even when assisted by powerful computers.

Conclusion

In the best case scenario, AI offers some progress in very specific areas. But they are obtained at the price of strengthening and maintaining the existing techno-industrial order (social inequalities, destruction of living beings, mass surveillance, aggressive propaganda, inequalities, etc.). In the worst case scenario, AI represents an existential threat to the future of our species and all of life on Earth. That is why it is our duty as human beings to do what we can to stop the development of this deadly technology now.