Capitalism, Not AI, Threatens Humanity

June 12, 2023 2 By Yatharth

M Aseem

The recent remarkable developments in Artificial Intelligence (AI) have generated excitement as well as apprehensions. While AI has demonstrated the potential for revolutionary development of productive forces, apprehensions are also being expressed about the risks posed by it. In this brief writeup, we examine the significance of AI from a materialist point of view. The main apprehension about AI in the nutshell is associated with its superhuman intelligence that might soon become sentient and the AI based machines enslaving humans. The appeal on March 28, with 2,900 signatures including those of Steve Wozniak and Elon Musk to cease training of more powerful AI for 6 months and set regulatory measures for it essentially comes from this idea.[1] And there is also the counter appeal from the CEOs and other experts associated with leading AI companies on CAIS expressing need for regulation to mitigate ‘risk of extinction from AI’ ‘alongside other societal-scale risks such as pandemics and nuclear war.’[2]

Second apprehension is about AI eliminating many jobs. After the introduction of ChatGPT, there has been an uproar that AI will replace a lot of existing jobs. In fact, as the dark clouds of retrenchment and massive unemployment loom over the global capitalist economy, intellectuals of the ruling class are devoting their full energy blaming the technology like robots and AI for such a dire situation. Goldman Sachs predicts that 300 million high-wage, non-manual jobs will be lost.[3] One can even go to the website willrobotstakemyjob.com and input specific occupations to get statistics on the likelihood of such theft. Writers have only a 3.8 percent chance … while machinists face an alarming 65 percent.

First, The Question of Sentience

Let’s first take the question of AI being sentient and intelligent. But whence did consciousness arise? During evolution, there emerged creatures which stood on their hind legs using their forelegs (hands) for other tasks. In the struggle with nature for survival, they realised the need to live collectively. “Mastery over nature began with the development of the hand, with labour, and widened man’s horizon at every new advance. He was continually discovering new, hitherto unknown properties in natural objects. On the other hand, the development of labour necessarily helped to bring the members of society closer together by increasing cases of mutual support and joint activity, and by making clear the advantage of this joint activity to each individual. In short, men in the making arrived at the point where they had something to say to each other… Comparison with animals proves that this explanation of the origin of language from and in the process of labour is the only correct one.” (Friedrich Engels, ‘The role of labor in the transformation of apes into men’)[4]

It was through this collective interaction with nature that consciousness developed through the reflection, abstraction, and generalization of greater understanding of the properties of nature in the human brain, the organ controlling the working hands, and the language developed to communicate this to each other. Without language there is no independent existence of consciousness and thought. Thus, human consciousness and thought developed based on practical social activity directed to satisfy human needs and is a special property acquired by one organ of the human body – the brain – in the course of evolution. We form ideas that express the relationships between things, and we understand what is useful and important in these relationships, because we need to understand them to survive in the world.

Consciousness is not something independent existing outside the human brain that can be implanted elsewhere by an external force. We cannot implant this consciousness in any machine made in the form of tools to enhance the capacity of human labour. Computers and the software that runs on them are also tools to enhance human capabilities. Machines are tools that can perform repetitive actions of humans in a continuous, tireless, automatic manner with greater accuracy and intensity through motion provided by external force. The tools that perform the repetitive actions of the brain in a continuous, tireless, automatic manner with more accuracy and intensity through the motion provided by external force are the computing (both hardware and software) machines. AI is but an advanced form of such machines.

What is AI? Can it Become Conscious?

The fear of AI becoming conscious is based on an erroneous conception of consciousness according to which the only difference between a computer and a thinking person is that the human brain is more powerful and sophisticated than a computer. This leads to the conclusion that as soon as powerful computers exceed the capabilities of the human brain, they will become conscious. But the way AI processes information is quite different from that of the human brain. Even the most advanced AI does not have the ability to think like humans. AI has fantastically high level of capabilities but only it’s only one part of what is called intelligence – it can identify patterns existing in data by passively collecting huge amount of it without understanding the context or the real purpose of the given task. But these patterns do not explain the interrelationships between things and their necessity. The AI has no idea that these data represent real objects that are related to each other, that have objective properties. AI has no idea why these specific patterns exist in the data or what they mean.

AI using machine or deep learning technique based ‘neural networks’ (so called because they reflect a few features of human neurons) are trained on datasets containing billions or trillions of pieces of data. Some data is pre-classified (labelled), some it clusters based on statistical probability (what is found with what and how often). This is how tokens (word or phrases in the context of large language models like ChatGPT) are created. The dataset that ChatGPT was trained on has over 1 trillion tokens. Based on statistical probability, which word/phrase is added before or after what and how many times, the AI generates the pattern. Depending on the question that is put to it, it creates new patterns based on the existing patterns and generates linguistic answers. That is why it is called Generative AI. Actually. ChatGPT name expands to Chat Generative Pre-Trained Transformer. It generates linguistic answers based on the principle of statistical probability, not the meaning or context of the pattern. Hence mistakes are natural to it. Due to its very nature, the possibilities of mistakes are reduced in the formulaic areas of coding, mathematics, natural sciences, etc., unlike more verbose areas.[5]

Clearly there is nothing either artificial or intelligent in this Artificial Intelligence. It is not artificial because it is not creating anything afresh. It just mixes and matches the humongous body of knowledge created by humans – workers, technicians, writers, artists, scientists, doctors, musicians, programmers, etc. – and presents it to us. It has no sense of context and expressions in this mix and match. As far as intelligence is concerned, this word originated not in the sense of being intelligent, but in the sense of espionage, because the first application of pattern matching AI was there – based on patterns in the satellite images, finding out which ship belonging to who were moving where in the sea, or what weapons and forces were deployed, etc. We can see a lot of statements and news related to war and foreign policy nowadays, which is based on such intelligence.[6]

This can be easily demonstrated by asking the image or text generating AI questions that require an understanding of the whole and its parts and their interrelationships. Neuroscience professor Gary Marcus, who is an ‘AI sceptic’, asked the AI to create image of an astronaut riding a horse, which it did very well. He then asked it for the image of horse riding an astronaut. It produced that too as it does not understand the various relationships between these parts. It simply creates images based on the words that are associated with these words. It doesn’t know what an astronaut is, and why it is absurd for him to ride a horse (much less for a horse to ride an astronaut). But if we ask a school child to do the same, she will immediately grasp the absurdity of it and laugh at us.[7]

True, the latest AIs are more efficient than humans at many tasks. But on deeper examination, these achievements are precisely the result of the fact that AI has no consciousness. AlphaGo AI on Google’s DeepMind supercomputer, which beat the best human Go player in 2016, took 30 million games to be trained to reach this level. That is more games than any human can play in her lifetime. A human could never play so many games, not just because life is limited, but because humans are intelligent and sensitive. These machines are so powerful precisely because they are not sentient. They can be programmed to test things by performing tasks repeatedly and read large amounts of text, so that they can identify useful patterns or ways of doing things. Obviously, it’s not intelligence, but AI’s ability to calculate non-stop which is superior to humans. Similarly, a person can’t run thousands of kilometers without stopping but a man-made car can. The car and AI both enhance the human capabilities, neither being conscious.

The relationship between concepts is an important part of consciousness. But AI does not ‘think’ in terms of concepts. It only recognizes specific patterns of datasets. AI can accomplish complex tasks with ‘understanding’ gained by training on huge datasets. But it does not have the ability to transfer that understanding to something only slightly different, a natural feature of human thought. An AI was trained to play a video game. It was able to do so better than any human being. But when parts of the game were redesigned with only a few pixels changed, the AI had to be retrained on the changed game. AlphaGo’s victory was widely reported in 2016, but it was outsmarted by an amateur player in 2023, defeating it in 14 out of 15 games when one of its weaknesses was revealed. AI could not correct its mistake, because it does not think.[8]

Clearly, the hypothetical debate that AI is conscious, or will be in the future, simply obscures the fact that what is being developed is merely a tool to enhance the capabilities of humans. That AI is superior to the abilities of humans in some areas is proof that it is an unconscious tool, not super-intelligent. The very job of machines is to be more powerful, more accurate, and faster than humans at certain tasks. Hence AI has nothing to do with conscious understanding. It is not able to and does not wish to rule and oppress humanity. It is in fact free from any desire.

True, AI can be used with enormous speed and on humongous scale for disinformation, vile propaganda, rumours for instigating murderous riots, massacres and wars, etc. But it is not AI which is responsible for it. It is the ruling class in an exploitative capitalist society, especially one which is now increasingly showing fascist tendencies world over, which uses AI and all other techniques for their venomous design. But which tool and technique in history has not been used for that purpose by the ruling classes? It is the ruling class and exploitative capitalism which need to be fought against instead of being trapped in the liberal attempt to divert attention from the real culprits.

AI’s Revolutionary Potential for Development of Productive Forces

The chief capability of AI is to identify patterns in large datasets with high accuracy. Training on large amounts of information gives them the ability to recognize patterns that humans either cannot or would take too long. Based on this pattern recognition AI can make predictions. This capability of AI can be applied to all types of activities to recognize complex patterns and find more efficient ways of organizing production.

This potential of AI to solve complex problems is real. Knowing the structure, function and behaviour of proteins has become easier using Google’s DeepMind. Nuclear fusion, known in principle to be the method of producing vast amounts of clean energy, is another big challenge where AI is already helping. DeepMind has also worked to discover hidden biological patterns in the eyes that may indicate the likelihood of vision problems later in life. This helps doctors to treat diseases before they even appear. Large amounts of energy can be saved by analysing energy use patterns in buildings and discovering more efficient modes of operation. The design of complex things like satellites and airplanes can become more efficient. If AI is systematically implemented in every sector of the economy and public services, energy savings can be increased enormously.[9]

AI can help grow human creativity by recognizing complex patterns and making predictions where some data is missing. An example of this is automated translation. It can translate any text reasonably accurately very quickly because AI identifies correlations between words and sentences in different languages and reliably predicts which words or sentences mean the same thing in the other language. This is making it possible for instant audio translations, so that someone wearing earphones can hear a live translation of a foreign language being spoken. Microsoft has already developed a device that enables blind people to learn about the world through an app. This will enable blind persons to perform various tasks independently. DeepMind has also been able to help archaeologists decipher ancient scripts in which parts of the text were missing or otherwise incomprehensible.[10] Anything on which AI training is possible with enough data can be solved by AI’s power to uncover hidden patterns.

Undoubtedly the possibilities opened by AI in terms of aiding and enhancing human creativity are fascinating. Based on the vast amount of visual data and written language available on the Internet, these AIs can create new images instantly at a prompt. High quality images are obviously very helpful for those who need to build prototypes or proofs of concept. Text generator AIs can help anyone quickly draft relevant text for various needs. AI will also make it possible to write code even for those who have no training in coding. All they must do is write in natural language what they want their website/software to do and what it should look like.

Shackled to Capitalism, AI Will Increase Monopoly & Inequality

What then is the real significance of AI? What will be its real impact on the society? For this, we need to know the character of the social system in which AI is to be used, its class configuration, the mutual relations of classes, the production and distribution system in which it is to be used – who owns it, rather than the character of AI. For what purpose does it will be used? What will be its impact on the interests of different classes of that society? To understand AI’s significance these are some of the important questions that must be considered.

Marx pointed out that every new social system in history paves the way for faster development of productive forces. But at a certain stage of their development the productive forces outgrow the relations of production in which they must operate, and these relations of production turn into fetters to further development. The capitalist mode of production also freed the development of productive forces to an immensely higher level than that of feudal society. But it has now become parasitic and a shackle for development. Investment and productivity growth are very low now despite new technologies. AI, and other digital technologies such as the Internet, represent means of production that are ‘too-advanced’ to be of reasonable use in capitalism as it is a system of production for private profit. If profit cannot be made from a potential investment, that investment will not be made. And profit can be made only by exploiting the labour power of the workers and selling the commodities produced by this labor in the market.

Technologies like the Internet and AI put a question mark on this profit-making by automating many processes to a very high degree. For example, the Internet enabled the copying and sharing of large amounts of information with little manual labor. A movie or song can be shared with countless people with negligible cost. So, the Internet made one of the key components of the music and film industries – the copying and distribution of recordings – redundant overnight. This presented a huge problem – how to make a profit when anyone can get a copy of the album for free with almost zero wage labour involved? Capitalists solved this problem by criminalizing online ‘peer-to-peer’ sharing and setting up streaming services with monopoly ownership of the content to extract rent from viewers/listeners. This solution has been effective from the perspective of protecting corporate profits. But it is an irrational limit on the production and distribution of creative works and prevents us from making full use of the potential of technology.

Marx pointed out that machinery lowers the value of commodities by increasing labor productivity in a capitalist economy.[11] AI does the same. For example, if writing and images for publications can be generated by AI, and TV-movie writers can rapidly develop plot ideas, the value of their work will be greatly reduced. If the training and skills of workers required to produce goods were to be reduced to mere typing signals, the value of their labor-power would also be greatly diminished. Gavin Mueller describes how tech companies make use of this low value labour power,

“Getting AI systems to function smoothly requires astonishing amounts of “ghost work”: tasks performed by human workers who are kept away from the eyes of users, and off the company books. Ghost work is “taskified”—broken down into small discrete activities, “digital piecework” that can be performed by anyone, anywhere for a tiny fee. The labor pool for ghost work is truly global, and tech companies have been eager to exploit it…. jobs can be sliced down to microtasks, paid as piecework, or “gamified” so they aren’t paid at all. This potential future of work has nothing to do with expanding leisure from “full automation.” Quite the contrary: in this future, work seeps into every nook and cranny of human existence via capitalist technologies, accompanied by the erosion of wages and free time.”[12]

However, in a socialist society which will need to develop productive forces to a very high level to fulfil collective social needs of all citizens, this would not be a bad thing at all. Such a society would need great innovations to increase labour productivity. Advanced technology will assist workers for this purpose by enhancing their capabilities. For example, the artist will have no fear of AI’s powers to produce ‘artwork’ at a moment’s notice, as art’s relationship to private property will be severed. It will not be produced as a means of profit or livelihood, but for the self and the society. Therefore, AI will not be a threat. It will become a helpful tool for the artist instead.

Like all technology, AI under capitalism will only fuel its inherent tendency towards monopoly and inequality rather than liberating humanity. The most powerful AI for will be developed by monopolies like Google and Microsoft because of huge capital owned by them. They would use their monopoly to make a profit, and technology would be used to speed up and cheapen production, lay off workers and reduce wages. It is already being used as a means of speeding up labour, that is, increasing the rate of exploitation. Cameras and other sensors can cheaply and effectively monitor and discipline the labor process of thousands of workers so that they produce more for the same amount of pay. Amazon is notorious for this.[13] As automated surveillance becomes more advanced and affordable; it will be implemented throughout the economy. This will increase the stress and alienation of the workers.

All technology like AI whose true potential is to rationalize production and enhance the creative powers of humanity, is always used by capitalism to further regiment and discipline the workers and make them redundant. The effect, therefore, would not be to bring stability and abundance to society, but greater uncertainty, greater concentration of power in the hands of giant corporates, and increased inequality. By further monopolizing the economy, lowering wages even further, and concentrating capital in fewer hands, AI will add to the anarchy under capitalism.

Global capitalism is already facing an almost permanent economic crisis because growth of value creation has become stalled, and rate of profit from commodity production is falling. Ever larger share of total profits now come from financial speculation and manipulation instead of commodity production turning capitalism parasitic. By taking level of automation to a new high, AI and other such techniques will further reduce the variable capital or labour content in the commodity production. This will further deepen the crisis of global capital.

Therefore, AI generates apprehension and fear, despite its amazing potential for the benefit of humanity. This fear of AI reveals the irrational and anarchic contradictions of capitalist production. Under capitalism, the highest achievements of human thought, the most wondrous technologies capable of eradicating the evils of poverty and ignorance, end up creating more poverty for teeming masses while at the same time concentrating wealth and power in the hands of few capitalists. We fear being enslaved by an impersonal and insensitive AI, since we are already subject to the impersonal, blind, unconscious, and anarchic capitalist production relations.

But the apprehensions expressed by two groups of financial capitalists, CEOs, managers, technocrats, other ruling class intellectuals and experts referred to the in the introductory para are not in the above category. Those two are attempts to serve the interests of two groups of the industry, one which has taken the lead in competition to develop and market AI techniques and the other one which has for now been left behind but wishes to challenge the incumbent leaders. Hence the differing views on moratorium and regulation.

AI In Service of a Planned Socialist Economy

But under socialism AI will be able to assist in planning for a complex economy to meet collective social needs. It will help resolving many of the problems faced by socialist planners earlier. With modern technology like sensors, it is possible to automate logistics, as Amazon has shown. In its massive warehouses, Amazon efficiently uses AI and robots to decide which items to move where and in what quantities by using a software-directed system that coordinates human laborers, who select individual goods, with robots, who move large shelves. There is no reason why sensors cannot be integrated into the economy, providing real-time data on what is being produced and consumed, in what proportion and where, what is to be supplied in what quantity and where, what equipment is vulnerable to breakage and requires maintenance.

AI will be very helpful in planning for the economy in a future communist society run by the Coordination Committees in which production will be organised on the principle of ‘to each according to need’ thereby requiring abundant production of human necessities. In such a society AI will help maximize the efficiency of meeting the collective needs of humanity, so that no one must live in fear of losing a job or of hunger or homelessness. This will enable the working hours to be reduced by eliminating the vast amount of redundant work. In the rest of the free time, the citizens will voluntarily contribute to human knowledge – science, art, music etc. for the all-round enrichment and prosperity of themselves and humanity. Obviously, AI would then be a tool of service in the collective oversight of society. The insights provided by it into the pattern of the economy and suggestions of the best way to bring down the cost of production will be crucial. We have in our hands the technology to harmonize production to eliminate the excesses, greed, and irrationality of the capitalist system. We can use it to give all of humanity what it needs to live well. This would make the construction of a socialist society free from deprivation and class divisions quicker and less onerous.

This revolution of the imminent arrival of ‘super-intelligent’ AI has also been welcomed by some ‘leftists’ who believe that this miraculous technology has the potential to solve all our problems. They hope that new technology will also ‘automate’ the need to overthrow capitalism, and that with its help we will be able to reach ‘fully automated’ communism.[14] But so long as we are subject to capitalism, it is capitalism and not the technology, that will determine how AI is developed and its vast potential is used. Therefore, the prediction of freedom from the exploitation and anarchy of capitalism through automation is an illusion. AI, no matter how advanced, cannot serve to liberate humanity from capitalism. The only force that can combat capitalism, and that has an interest in doing so, is the proletariat. Only the working class is interested in achieving socialism. Only when we finally overthrow capitalism and subject the economy to conscious, rational planning can AI and other technologies flourish to their full potential as the most wondrous and common tools of human development.

Are AI & Automation Responsible for Unemployment in Capitalism?

Clearly technologies such as AI, Internet of Things, data driven decision making, intelligent transportation hold the potential for fundamental transformation in the production process. Many old skills will no longer be required. On this basis, these techniques are being blamed for large increase in unemployment. But the threat of automation is as old as the Industrial Revolution. While humanity dreamed of the ability to break free from back-breaking labor since the advent of mechanized production, the fear of being replaced and made useless by machines is equally old. This myth of large-scale loss of employment because of machines persists since the beginning of mechanization two centuries ago. Fear of mechanization also has a history of resistance to it. There was a movement called Luddites in Britain in the 1820s, in which workers used to attack factories and break machines. The unions in India also launched a movement against computerization in 1980s.

Technologies like AI are also machines that automate some mental work. Therefore, their possible effect should also be studied according to the effect of mechanization. Does the history of mechanization prove that unemployment has increased because of it? And if mechanization and automation are not responsible for unemployment, then what is the real reason responsible for rising unemployment? History tells us that despite mechanization and automation, the need for labor in industries has increased instead of decreasing, and to meet this need, a large population has migrated from rural agriculture and handicrafts to urban industrial areas. It is true that a smaller number of workers are needed for the same level of production after automation. But the keyword here is, ‘for the same level of production’ because the development of machines and technology not only creates new areas of production but also expands the needs of the society.

Therefore, there is a need to increase the level of production to meet the social needs. In the process, older forms of employment are phased out and new jobs are created. Overall, the need for workers does not decline. It rather keeps on increasing. The important question to ask here is whether the production of goods/services has increased to such an extent and their availability has reached such a level that there is no longer a need for expansion of production? It has also to be examined whether the workload on the workers has reduced so much that they have become redundant. If both the above conditions are fulfilled, then surely, with the introduction of AI and other advanced technologies, no other employment will be available for the labor force which will become surplus for the current level of production. Does anyone believe that these conditions have been fulfilled?

Let us take two examples of the impact of mechanization on the economy. Computerization was expected to cause large scale unemployment in India. But on the contrary a whole new industry of information technology has been established, in which many jobs are available today. Similarly, 30-40 years back sale of ice for domestic use was a big retail business with shops on almost every street. Many workers were engaged in ice factories and its transportation. When the domestic ice making became automatic, of course most of these jobs disappeared. But as soon as it became automated, it ceased to be an independent industry and became a part of the new industry of freezers and refrigerators, which has created many new jobs. Many industries existing today were not there half a century ago. Mechanization has destroyed old jobs, but jobs created in these new industries took their place and the total number of workers has in fact increased instead of decreasing.

In fact, the only source of profit in capitalist production relations is the appropriation of surplus value produced by labor power. If an industry is fully automated, then there can be no surplus value and profit in it. Therefore, the existence of a fully automated industry is impossible under capitalism. Rather, the history of capitalism shows that it continues to give rise to new industries in which, in the beginning, the need for labor is high and the organic composition of capital is relatively low. The rate of profit is high in these industries. The emergence of such industries also acts as a counteracting factor in the tendency of the rate of profit to fall in the capitalist system.

Therefore, it is a lame ‘argument’ that jobs have been lost due to reduced work due to automation and technology. This is the sophistry of those who defend the exploitation of capitalists. If there is no work left due to advanced technology, why have the working hours in every sector increased to 10-12 hours a day across industries? Why are efforts being made to legalize a 12-hour working day instead of the 8 hours one? Why even the lunch and toilet breaks being cut short? Why are the workers being forced to speed up the work? Why is the speed, stress and hours of work increasing everywhere instead of workload getting reduced after automation? In fact, workers are being forced to work more hours and faster for less wages by spreading a fear of the loss of jobs due to mechanization.

If the media and intellectuals of the ruling class are to be believed, the machines do all the work. But the reality is that contrary to the principle of 8 hours of work, 8 hours of rest, 8 hours of entertainment which was established by the glorious labor struggles starting from the 19th century, most workers are compelled to work more than 12-14 hours these days. Why so, if the workers have become redundant, and there will be no work in future because of AI and automation?

Thus, while emphasizing only one aspect of the reality of advanced technology and mechanization, attempt is being made to cover up the truth that it is only under capitalist production relations that automation puts people’s livelihoods at risk. Therefore, instead of blaming automation, the capitalist system must become our target. In the absence of capitalism, technological advancements in robotics and artificial intelligence will reduce working hours and make people’s lives easier than ever before.

In fact, whatever commodities we look at from food grains, pulses, fruits, vegetables, milk, meat, clothing, shoes, housing, cleaning materials, water, electricity, irrigation, transport, schools, colleges, hospitals, cultural requirements, hardly 10% of the population can fulfil their needs. If a plan is made to meet the needs of all the citizens, there will, in fact, arise a shortage of labor for the required rapid expansion of production. In that scenario, the need for labour will not be fulfilled even after making labor compulsory for all men and women. There will be a need for further development and automation of technology. Why should a person keep sweeping the road, why not a machine that can clean quickly? Why should an inhuman system like caste-based division of labour exist for such tasks? Why should any person toil like a beast day and night in the fields and mines? Why not increase mechanization to a level so that all men and women can handle the responsibility of such tasks without anyone having to do back-breaking labour.

Therefore, not only is there enough work, but there are also immense possibilities of its expansion. The real hindrance is the private ownership by a few people of the means of production produced by the collective labor of the whole society and the compulsion of the rest to sell their labor power to them. The stolen part of the value produced by labour power is the profit of owners of capital, having accumulated which they have become immensely rich. The less labor power they use to produce, the more profit they see. But where is this product sold? The purchasing power of the workers is low. They cannot buy even when they are in need. For the capitalists themselves, there is already a surplus piled up. Hence there is ‘over-production’. Industries are operating at 60-70% of their installed capacity. So, no new investment is forthcoming. The main objective in capitalism is not to increase production, but to get it done with fewer workers to get higher profit. Automation, therefore, results in layoffs and unemployment.

This is the basic contradiction of today’s society, the solution of which is not the basic income by the government, but the collective ownership of all the means of production. By planning production to meet the collective social needs, not only will employment be created for those who want it, but work will have to be made mandatory for all. The reason for unemployment is the capitalist system of production for profit. If production is organized for the abundance of essential goods/services to fulfil social needs, there will be shortage of workers, and technology will have to be further advanced. In that scenario, the fullest use of the potential of AI and other technologies will be possible in the interest of humanity.

Capitalism, not AI, Poses Risk of War, Pandemics & Extinction

It is not AI which can enslave or extinct human race. Disasters like war, pandemics and destruction of environment are the result of rapacious and parasitic capitalism-imperialism. Since the capitalist crisis of 1970s we can observe constant increase of share of financial profit within the aggregate profit in advanced economies giving rise to the tendencies of financialisation and parasitism. The economy of the whole world is being financialised and made subservient to financial interests. This results in not only the fall of rate of profit but also the fall of the total profits from production of new value in commodity production. While the financial magnates are getting enriched as the quantum of fictitious capital from the manipulation of share prices and amount of derivatives has become humongous, the small investors are becoming insolvent. It is causing immense unemployment and destitution and fall in real wages.

Along with the stalling of labour productivity growth and slowing development of means of production, the link between labour productivity and wage growth has broken. The fall in wages and quality of employment in addition to growing unemployment and jobless growth has created an immense problem of expanding indigence. The poverty is increasing. The gap between rich and poor is widening. This is pushing ever larger share of profits towards investment in the financial sector.

This results in constantly increasing dominance of arch-reactionary finance capital on the entire economy and social life. This finance capital is greedy to the extent of risking extinction of humanity. This is not an indication of just the emergence of an economic crisis but the very signal of it becoming irreversible. This is an almost permanent crisis with no solution in sight except a disastrous war on the ruins of which another edition of the capitalist ‘golden era’ might be built.

In brief, it is the crisis of over-development of capitalism making it necessary for it to reach the zenith of parasitism, unbelievable fall, and recession unless humanity goes forward beyond the limit and horizon set by capitalism. Obviously, the most reactionary and counter revolutionary section of the big monopoly finance capitalists sees no other alternative but to opt for fascism, moving towards suppressing democracy and seeking a way out to free themselves from all constraints by a fascist dictatorship so that they can persevere in continuing their maximum loot and profits.

The long persisting economic crisis is not only giving rise to fascism, but it has also quickened the possibility of a World War between the imperialist camps, Ukraine, Syria, and Taiwan being its chief flashpoints. The main reason for this is the emergence of a strong Russia-China-Iran axis to challenge the global hegemony of the Anglo-American imperialism led NATO, and the new global polarisation and balance of power emerging from this. AI emerges in this scenario with its revolutionary potential to develop the productive forces and automate many of the production processes. By reducing labour content in many production processes, it will further squeeze the quantum of surplus value that can be appropriated from commodity production inducing a further fall in rate of profit. Hence AI is going to accentuate the capitalist crisis which has already become almost unending and irreversible, and is propelling capitalism-imperialism towards fascism, war, pandemics, and environmental disasters.


[1] https://www.marca.com/en/technology/2023/03/29/64249d4b268e3e851c8b4591.html

[2] https://www.safe.ai/

[3] https://www.zerohedge.com/markets/ibm-stop-hiring-roles-can-be-replaced-ai-nearly-8000-workers-be-replaced-automation

[4] https://www.marxists.org/archive/marx/works/1876/part-played-labour/index.htm

[5] Amitabh Guha, https://indianexpress.com/article/opinion/columns/what-technology-leaders-asking-for-a-six-month-halt-on-ai-dont-want-you-to-know-8534917/

[6] Evegeny Morozov,  https://www.theguardian.com/commentisfree/2023/mar/30/artificial-intelligence-chatgpt-human-mind

[7] G Marcus, E Davis, Rebooting AI: Building Artificial Intelligence We Can Trust, Pantheon Books, 2019

[8] https://arstechnica.com/information-technology/2023/02/man-beats-machine-at-go-in-human-victory-over-ai/

[9] Daniel Morley, http://www.marxist.com/artificial-intelligence-doomsday-for-humanity-or-for-capitalism.htm

[10] Y Assael, T Sommerschield, B Shillingford, N de Freitas, “Predicting the past with Ithaca”, Deepmind, 9 March 2022

[11] [11] Marx, Capital, Volume One, Ch 15.

[12] https://futuresofwork.co.uk/2020/12/02/the-full-automation-fallacy/

[13] N Dyer-Witheford, A Mikkola Kjosen, J Steinhoff, Inhuman Power: Artificial Intelligence and the Future of Capitalism, Pluto Press, 2019, pg 93

[14] Aaron Bastani, Fully Automated Luxury Communism: A Manifesto, Verso Books, 2019