By MARCO SCHNEIDER, WILLIAM FRANCE & LUIZ CLAUDIO LATGÉ*
The history, development and implications of artificial intelligence, both in technical, social and economic terms
Artificial intelligence
If the history of technology can to some extent be adequately illustrated with the help of the well-known image by Marshall McLuhan (1969), according to which the media and other techniques would be extensions of the human body, the extensions become extractions under the regime of capital, through two modalities of alienation.
Firstly, when direct producers are alienated from their means of production – land, workshops and tools, causing peasants expelled from the fields and artisans, whose workshops cannot compete with manufacturers, to become wage workers, in what Marx (1985) calls the formal subsumption of labor to capital.
Secondly, and this is the point that interests us here in particular, when they are alienated from their labor knowledge, which will be subdivided and embodied in independent and combined work units, to be later incorporated into the machinery of industry, a process that constitutes the real subsumption of labor to capital (Marx, 1985).
The real subsumption of labor under capital is one of the ways of converting living labor into dead labor. The other is simply the production of commodities under a wage-earning regime, already present in formal subsumption (Marx, 2002).
The process of alienation of knowledge called the real subsumption of labor to capital began in the early days of modern industry and continues to this day, despite the complexity of the new international division of labor. It begins with the formal and then real subsumption to capital of so-called manual labor and continues with intellectual labor (Antunes, 2006).
Initially, and to some extent even today, the subsumption of intellectual work occurs only in formal terms, as in any situation of intellectual work under a salaried regime – teachers, journalists, scriptwriters, etc. Gradually, it also occurs in real terms, at least since the calculating machine – the first form of Artificial Intelligence, we dare say –, whose new forms have been enchanting and astonishing the world.
We therefore have that Artificial Intelligence, in a first approximation, from the perspective of the critique of political economy, is the result of a new stage of the formal and real subsumption of intellectual work to capital.
Furthermore, Artificial Intelligence is a generic term that designates roughly the ability of machines to emulate or surpass so-called intelligent human actions. Intelligent actions involve understanding, adaptation (adapting to the situation or adapting the situation to oneself), invention, reproduction, transformation. They are assimilative, adaptive and inventive capabilities, dependent on processes called, in the case of artificial intelligence, machine learning.
This learning requires, in addition to algorithm programmers – whose intellectual work is subsumed to capital in formal and real terms –, immense amounts of data, alienated from all users of digital platforms, through processes of surveillance, espionage and capture by the big tech, which operate under an oligopolistic regime. These extremely wealthy corporations are mostly American and whenever possible place themselves above the law of the rest of the world, and even the US, in some cases.
Thus, in a second assessment, Artificial Intelligence is also the result of a new modulation of the old imperialism, Postgraduate Course , as it generates stratospheric profits and strengthens the geopolitical power of these companies and the American State, respectively, at the expense of the world's population, including the American population. These costs involve semi-slave labor in gold and niobium mines in Brazil and Africa, the uberization of work (Bezerra, 2024), unemployment and invasion of privacy around the world.
Artificial Intelligence, therefore, is the composite result of the alienated knowledge of algorithm programmers and the big data, this produced by the navigation of all people connected to digital platforms, in their free time or at work, whether they work for the platforms or not.
The latter case includes the work of those directly responsible for the production of knowledge and entertainment – scientists, journalists, scriptwriters, artists, etc. – knowledge whose production (training, investigation, research, creation) and circulation (publication, broadcasting, distribution and exhibition) costs are considerably high, meaning that its alienation has generated disputes between old and emerging sectors of the dominant classes linked to the production, reproduction and circulation of information and communication, such as journalistic companies and platforms.
Conventional media follow a well-defined logic of production, distribution, marketing and business, which establishes an entire process, rules and codes of values. This is how newspapers, radio, TV and now digital media are differentiated.
It is possible to analyze the media within a hypothetical spreadsheet, which can be conceived in its simplest form, despite the complexity of each activity: product, technology, geography, production system, business model, metrics and prices.
In the case of printed newspapers, the product is the daily newspaper, a collection of pages full of photographs and news. The news must be attractive to the reader, which will allow the newspaper to sell copies and advertisements. To be produced, it depends on a technology, a printer. Production requires the hiring of reporters, photographers and editors, and a production time. Newspapers used to have morning and afternoon editions. Culturally, a daily edition has become the reference.
The product has a limited reach, since the newspaper needs to be transported to the reader. This defines its geography. And also the concept of news, of what happens more or less nearby and can have an impact on the citizen's life.
The production process also defines the size of the newspaper. How many pages can it have? Professionals are needed to produce the news, which has a cost. The equation then becomes how many news stories can I produce, how many pages can I deliver, so that the newspaper reaches the reader on time and at an acceptable cost. The business logic then needs to balance production time, production cost, number of copies and sales price so that the company can survive and make a profit.
All variables are quantified: audience, copies, cover price, advertising value. The newspaper will have two sections, twenty pages, will have 30 reporters and photographers, will have five news vans and a circulation of 50 thousand copies. It is an accounting exercise. The newspaper then establishes that it will cost R$5 and that it will sell an ad page for R$10 thousand. A spreadsheet that will be revised over time according to fluctuations in factors. Paper becomes more expensive, labor costs increase, sales fall, any change requires adjustments. But it does not change the business logic. Offer information to create an audience and sell subscriptions and audience.
TV has other unique features, but it follows the same business logic. While newspaper technology was the printer, TV needs antennas, editing suites, transmission equipment and receivers. Just like newspapers, this technology and its production costs will be important for defining the business. But the logic is the same: to build an audience by producing audiovisual content, news, sports and entertainment, which will be consumed by a specific audience in the area of influence of the electronic signal, its geography.
Thus, open TV sells advertisements. TV Globo, Bandeirantes, All time lap record, all in the same way, on their business scale. The production cost, as in newspapers, will adjust to the marketing capacity. It is necessary to invest in technology, personnel, cameras, etc. But this cost must fit within the commercial budget. TV creates its own language, audience metrics and marketing tables. If newspapers sell centimeters, TV sells seconds, to the audience it is able to reach in a given geographic area.
The model can be applied to any media, but it begins to break down with digital media. Technology breaks down several limitations of traditional media. It popularizes the production and circulation of images, sound and text, the sum of all other media. It is interactive. And it dismantles any geography, and can be accessed virtually anywhere in the world. In this way, it imposes a scale that is unattainable for other media – although each of them can take advantage of new technologies to update their products and deliveries.
So far, we have been talking about technology – and that is how platforms like to present themselves, as technology companies, not media. However, in the same way as those connected to networks. This is where the differences begin, because the essence of the business is not in arousing the interest of the public through the circulation of professional content, the production or acquisition of which has a cost, but through the permanent circulation of free information, which keeps the consumer connected. After some erratic attempts, in the early days of the internet and during the ten years that followed until the popularization of smartphones, digital platforms established their business model: connecting the public, offering an incessant flow of information.
Universal connectivity and a permanent flow of information are what will make this new business viable. With the knowledge extracted from billions of connected people, it will be possible to offer programmatic advertising, whose degree of precision is superior to advertising in conventional media.[I]
The cost, then, shifts from the production of information to the surveillance and data processing systems, the algorithms and search tools, and the consumer (re)cognition. Since the cost of producing constant information would be prohibitive and the platforms demand an infinite flow of production, they cannot remunerate this production as newspapers or TV did, paying professionals and broadcasting rights. How can an infinite flow of content be paid for?
In an elementary math exercise, the issue becomes clear. Paying a penny per post would already be prohibitive. A penny multiplied by infinity would generate an unpayable expense of infinite reais. It would also not be feasible to reduce the content to the payable value, because this would interrupt the flow necessary to maintain the connection. Thus, the value of information on the platforms is very close to what the platforms spend to display it, very close to nothing. Tik Tok, Whatsapp, Facebook, Instagram, all compete with the The New York Times and Nature, with the same value as a click.
It is true that all content travels through the platforms, including qualified content from newspapers and other media outlets, which are very expensive, and no matter how hard they try, they will not be compensated for it. Eventually, the platforms may agree to some kind of agreement. But they have already shown that if the charge is significant, they prefer to take the news off the air. Besides, the hypothetical spreadsheet we presented still applies: the platforms have built their business logic, their metrics, their sales tables and compete for market share with traditional media outlets, with a huge advantage.
The impact of this change is greater than the change in the media market. It establishes a new information regime (BEZERRA, 2023), by confusing transmitter and receiver, experts and amateurs, devaluing the quality of information, under the appearance of a democratization of communication. In reality, digital platforms favor narratives and define an entire cultural melting pot that shapes our time. Among the values, freedom of expression stands out. A fundamental right of the citizen and a value historically defended by the press.
It seems to be the same cause, but it is not: what is being defended is the possibility of equating consistent information with opinion, even if based on false data and intentionally manipulated to generate clicks. The narrative is that every citizen has the right to express themselves, even if it means conspiring against democratic governments, destroying reputations, or selling fake products to the sick. The narrative will establish elective affinities, which have already been clearly demonstrated by the growth of extremist groups, especially the far right, as we have seen in the United States, Hungary, Brazil, and now also in Argentina.
These affinities will make platforms align themselves, preferentially, with these groups, to generate engagement (audience) and protection. The best example can be taken from Brazil with the action of platforms in the election of a disinformation bloc, which was elected by spreading fake news. The articulation became evident when the government tried to debate in Congress the regulation of digital platforms and media and Google, among others, distributed content presenting the project as the Censorship Law. With the support of the disinformation bloc.
Not that traditional media was or is free from bias and manipulation. Far from it. There was, and is, however, some degree of regulation, a ethos professional journalists, a more or less true tradition of commitment to the public interest and to the factuality of news, a reputation of credibility to uphold. In addition, journalistic and broadcasting companies have an address, CNPJ, are responsible, and are subject to national legislation. Thus, regardless of the ideological bias and commercial interest they may have, there are checks and balances, while the algorithmic mediation of platforms, in addition to being opaque and apparently ownerless, has as its ethos unique, so to speak, the profit margin of programmatic advertising.
Therefore, they can digital platforms be considered neutral tools, pure technology for connection and free, undirected navigation? The way they employ and encourage the use of artificial intelligence, for example in scandals such as Cambridge Analytica, Brexit, the 2016 American elections and the 2018 Brazilian elections, among many other well-known cases, suggests that it does not.
In purely descriptive terms, Artificial Intelligence is an area of Computer Science. The term was first used by John McCarthy during a conference at the Dartmouth College, in 1956 (IBM).
Basically, although it emulates human intelligence, Artificial Intelligence performs tasks that humans cannot perform at the same speed, increasing productivity and optimizing processes (Bengio, 2023; He and Degtyarev, 2023). It also performs tasks that humans are perfectly capable of performing, with greater efficiency than Artificial Intelligence – as in the case of automatic response services by telephone or WhatsApp –, but at much lower costs.
Although it may seem new, artificial intelligence began to be developed in the first half of the 2022th century and is now used in various sectors of society, both civil and military. It mediates everything from interpersonal relationships to economic and political relationships, including digital disinformation on the network (Schneider, XNUMX).
Artificial intelligence is not limited to computer software, but applies equally to hardware. The so-called Internet of Things (IOT), self-driving cars, programmatic advertising (microtargeting) and biometric systems (fingerprint, facial recognition, voice recognition, etc.) are some examples of Artificial Intelligence activity, among a myriad of other possible ones.
Machine learning (machine learning algorithm ), deep learning (deep learning) to big data are key terms when talking about Artificial Intelligence. This is because this learning, which is what ultimately allows Artificial Intelligence to function, is only possible with a considerable volume of data that provides the algorithms, another inseparable element of this set. Based on the data analyzed and organized by these latter (the algorithms), the machine learns.
The term “learning machines” first appeared in the article “Computing Machines and Intelligence”, by Alan M. Turing, in October 1950. Guided by the question “Can machines think?”, Turing proposed a test to verify whether a computer, when in a competition situation with a human being, would have satisfactory performance.
Turing (1950) concluded that, depending on the storage, processing and programming capacity:
Parts of modern machines that can be considered analogous to nerve cells work about a thousand times faster than these [nerve cells, read “neurons”]. This should provide a “safety margin” that can cover speed losses that may arise in various ways. (Turing, 1950, p. 455, our commentary).
To better understand the machine learning process, it is worth introducing the equally important concept of “neural networks”. Turing (1950) mentions these networks in his conclusions, but it is Hopfield (1982) who focuses strictly on this topic, explaining, from the perspective of chemical engineering, the analogy between neurons and “physical systems with computational capabilities”, chips.
Artificial intelligence, therefore, has imitated human cognitive capacity since its inception. It has developed exponentially in the last decade and has become a topic of expectations and concerns for populations and governments around the world, due to its known and imagined benefits and harms. Among the known harms, we have the spread of misinformation in elections, sometimes with the use of deep fakes, also used to simulate pornography, mixing faces, bodies and voices; the uberization of labor relations (Bezerra, 2024); the replacement of human labor by Artificial Intelligence, promoting unemployment, etc. Among the unknowns, there is speculation about Artificial Intelligence getting out of control and turning against people, in an update of Frankenstein, Golem and other fears reactive to radical technological changes.
We will not discuss the unknown here. In the Final Considerations, however, we will return to the critique of the well-known harmful effects of Artificial Intelligence, with the aim of problematizing the notion itself in its ideologically fallacious aspects, suggesting alternative uses. To support the argument, we will incorporate into the debate a brief review of Marx's critical appropriation of Hegelian dialectics and Feuerbach's materialism, in order to discuss the very notions of intelligence, reason and cunning.
Dialectical reason
Something intelligent must be rational. But what exactly does being rational mean? In common sense, it means not acting in an absurd, incoherent way that causes harm (to whom?). It also means calculating accurately.
Intelligence involves Logos e metis (Capurro, 2020), usually translated as reason and cunning.
Logos essentially concerns the notion of truth; metis, to that of effectiveness.
Something is true in the sense of the maximum equivalence between the understanding and the thing understood. Reason means among other things the subjective faculty of establishing this equivalence, with the mediation of the senses, of language or of both. But this thing understood can be a more or less indifferent object of contemplation or the result of a planned action. It can even be the goal of a vital action. In the case of planned action, its truth lies in the efficacy of its result in relation to the goal. But who planned it? How does the result affect those who did not plan it? For whom is it reasonable?
According to Herbert Marcuse, the French Revolution introduced into history the idea of a social order governed by reason. The main German philosophers of the period, with Germany politically and economically behind France and England, explored this issue in greater depth, divided between admiration for the advances of the Revolution, the achievements of science and technology, and growing individual freedom, on the one hand, and rejection of the Terror, the misery of the masses, the fraying of social ties, and what Max Weber would later call the “disenchantment of the world.”
German idealism has been regarded as the theory of the French Revolution. This does not mean that Kant, Fichte, Schelling, and Hegel developed a theoretical interpretation of the French Revolution, but that they largely wrote their philosophies in response to the challenge from France to reorganize state and society on a rational basis, so that social and political institutions would be adjusted to the freedom and interests of the individual. Despite their severe criticism of the Terror, the German idealists unanimously welcomed the Revolution as the dawn of a new era, and without exception associated their basic philosophical principles with the ideals it promoted. (Marcuse, 1978, p. 17)
But what would be a rational basis for the State and society? The world should become an order of reason.
The ideals of the French Revolution found support in the processes of industrial capitalism. Napoleon's empire had eliminated the radical tendencies of the Revolution while consolidating its economic consequences. French philosophers of that period associated the achievement of reason with the expansion of industry. Growing industrial production seemed capable of providing all the means necessary to satisfy man's needs. Thus, at the time when Hegel was developing his system, Saint-Simon, in France, extolled industry as the only power capable of leading men to a free and rational society. The economic process appeared as the foundation of reason. (Marcuse, 1978, p. 18)
How far we are today from these dawning horizons, to use a term dear to Bloch. Today, there is fear that artificial intelligence, the great-great-granddaughter of the steam engine and the mechanical loom, will dominate or destroy humanity. At the very least, that it will be used for very bad purposes from the point of view of democracy and human rights. And no one has faith in the capitalist economy in terms of the general good anymore. At best, it is claimed that there is no better option. And the idea that the economic process would be the foundation of reason seems bizarre, absurd, except to the rich or ultra-rich.
Shortly after Napoleon liquidated “the radical tendencies of the Revolution”, Auguste Comte set out to liquidate the radical tendencies of the French Enlightenment, formulating a notion of positivist reason, which deliberately dulls the critical edge of reason, that is, its property of denying what is irrational in a given state of affairs – limiting itself to denying what was already dead, the Ancien Régime. Thus, with the triumph of the bourgeoisie and the subordination of the popular classes in capitalist molds, a new dominant rationality is established (Marcuse, 1978, p.309-325), crowned by the idealistic and conservative binomial “order and progress”, inscribed on our national flag and whose authoritarian vocation we know so well in Brazil.
Confronting the rationality of the positivist type and its basis, we have dialectical reason: “In Hegel’s view […] man will reach certain conceptions that reveal that reason is in conflict with the existing state of things. He will come to realize that history is a constant struggle for freedom, that the individuality of man, in order to be realized, requires that he possess some property, and that all men have an equal right to develop the faculties that are theirs. In reality, however, servitude and inequality prevail; many men have no freedom at all and are deprived of the last crumb of property. Consequently, the “non-rational” reality must be modified until it conforms to reason.” (Marcuse, 1978, p. 19)
There is, however, a difficulty: “[…] What men think to be true, right and good must be realized in the actual organization of their social and individual life. But thought varies from individual to individual, and the resulting diversity of individual opinions cannot provide a guiding principle for the common organization of life. Unless man possesses concepts and principles of thought that designate universally valid norms and conditions, his thought cannot claim to govern reality. In line with the tradition of Western philosophy, Hegel believes in the existence of such objective concepts and principles, and he calls their totality reason.” (Marcuse, 1978, p. 19-20)
Reason, then, is the totality of objective concepts and principles of thought that denote conditions and norms of universal validity, which must be implemented, because: “For Hegel […] reason cannot govern reality unless reality has become rational in itself. This rationality is possible through the irruption of the subject into the very content of nature and history.” (Marcuse, 1978, p. 21)
For positivism, the end of Ancien Régime should carry with it the end of thought as a negation of the irrationality of the existing. Progress would come within order. Hegel, however, throughout his life lived the tension between the rational negation of the irrational existing and a refusal of fanciful idealisms.
In any case, how can we make reality rational if our era intends to give up universal conditions and norms – such as the end of the exploitation of man (and woman) by man – and is marked, on the contrary, by the dispute of the most varied particularisms, some progressive and libertarian, others reactionary and authoritarian, or even opportunistic, but all resistant to the very idea of reason, accused by some of being totalitarian, by others of being blasphemous?
It is this dispute that has educated, so to speak, Artificial Intelligence, with the second group gaining ground.
Except in situations where corporate interests are threatened by science, as in the case of polluting fossil fuel or pesticide industries, positivist reason does not tend to confront the status quo, whether of the liberal or authoritarian type. This remains the task of dialectical reason, which developed, from Marx onwards, as a critique of contemplative knowledge, a critique synthesized in his 11 Theses on Feuerbach.
It is known that famous people Theses These are Marx’s personal notes, later published by Engels. They are brief and powerful postulates, which served as guidance to Marx himself in the elaboration of his system. However: “[…] short sentences sometimes seem to be able to be comprehended more quickly than they actually are. And sometimes it is typical of famous sentences, quite against their will, that they no longer provoke reflection or that they are swallowed still very raw. […] what exactly is meant by thesis 11? How should it be understood in Marx’s always precise philosophical sense? It should not be understood, or rather, misused, in any mixture with pragmatism.” (Bloch, 2005, p. 271)
Ernst Bloch clarifies that there was much debate about the ordering of Theses. According to him, the unsystematic nature of personal notes indicates that their order is random, and an underlying structure can be identified that can be ordered in terms of an epistemological group, which discusses contemplation and activity (theses 5, 1 and 3); a historical-anthropological group, which deals with self-alienation and true materialism (theses 4, 6, 7, 9 and 10); a theory-praxis group, which discusses the issue of proof and validation of knowledge (theses 2 and 8); the 11th being the crowning, the conclusion.
There is no space here to discuss in detail the Blochian exegesis of Theses. However, we will discuss the question of contemplative knowledge, as well as draw attention to the criticism that Bloch directs at misreadings of the Theses, particularly the 11th, which lead to irrationalist and potentially reactionary theoretical, epistemological and political confusions. Our aim is not to add anything new to studies on Marx, but to advocate the relevance of Ernst Bloch's critique of pragmatism and practicalism to the contemporary debate on Artificial Intelligence.
By pragmatists Ernst Bloch refers to a certain voluntarist and anti-intellectualist tendency, present in his time and in ours: “As far as the “practitioners” of the socialist movement are concerned, it is obvious that morally they certainly have nothing in common with the pragmatists; their will is transparent, their intention revolutionary, their aim humanitarian. However, when they put their heads aside, nothing less than the entire richness of Marxist theory, together with the critical appropriation of the cultural legacy made by it, ends up emerging, on the occasion of the “trial and error method”, of dilettantism, of “practicism”, that cruel falsification of thesis 11, which methodologically resembles pragmatism. […] The “practicists”, who at most give short-term credit to theory, […] introduce into the luminous essence of Marxism the darkness of their own private ignorance and of the resentment that is so easily associated with ignorance. […] the schematism of lack of reflection also lives on the inactive anti-philosophy itself. In this way, however, one can refer even less to the precious thesis on Feuerbach; the misunderstanding then turns into blasphemy. For this reason, it must be continually emphasized that, in Marx, a thought is not true because it is useful, but useful because it is true”. (Bloch, 2005, p. 273)
Ernst Bloch is here attacking a mistaken reading of the 11th Thesis, according to which, according to Marx, philosophy had already fulfilled its role and only revolutionary action was valid, as if this could do without revolutionary theory, as expressed in Vladimir Lenin's famous sentence. In Bloch's own words: “As much as the anti-pragmatism of the greatest thinkers of praxis [… provides open doors, these can repeatedly be closed by a self-interested misinterpretation of thesis 11. By an interpretation that, in a grotesque way, believes it can detect in the ultimate triumph of philosophy – which occurs in thesis 11 – an abdication of philosophy, precisely a type of non-bourgeois pragmatism.” (Bloch, 2005, p. 273-4)
That is, if praxis requires metis, also requires Logos: “[…] if the destruction of reason causes us to sink again into the barbaric irrational, the ignorance of reason causes us to sink into the imbecile irrational; and the latter does not end up spilling blood, but ruins Marxism. Thus, banality is also counter-revolutionary in relation to Marxism itself; for it is the concretization (not the Americanization) of humanity’s most advanced ideas.” (Bloch, 2005, p. 274)
A 11th Thesis on Feuerbach is not a rejection of philosophy. It is both a critique of the merely contemplative element of almost all the philosophy that preceded it, and an invitation to a permanent updating of philosophy and critical thinking as a whole through praxis, that is, the feedback of theory and practice, in a transformative perspective, tributary both to the history of human struggles for justice and freedom and to its best theoretical expression in philosophy: “[…] against the preceding philosophers the accusation is raised, or rather: it is identified in them, as a class barrier, the fact that they merely interpreted the world differently, not the fact that they philosophized. Interpretation, however, is similar to contemplation and derives from it; non-contemplative knowledge, therefore, is now distinguished as the banner that truly leads to victory. However, as the banner of knowledge, the same banner that Marx stamped – of course with action, not with contemplative tranquility – in his main work of erudite investigation. This main work is pure instruction for action; however, it is called The capital, and not a Guide to Success or even Propaganda in favor of the act; it is not a recipe for heroic deeds […], but is situated […] in the careful analysis, in the philosophical investigation of the interrelations within the most complicated reality, taking the path of understood obligation, of knowledge of the dialectical laws of the development of nature and society as a whole. […] Without a doubt, Marx uttered cutting words against philosophy, but he did not do so against contemplative philosophy pure and simple, whenever it was a question of a relevant philosophy of important periods. He did so precisely against a certain type of contemplative philosophy, that is, that of Hegel's epigones of his time, which was rather a non-philosophy”. (Bloch, 2005, p. 274-5)
The notion of contemplative philosophy to which Bloch refers concerns, according to Bloch himself, practically the entire history of philosophy prior to Marx, in a more general sense, and Marx's critique of Feuerbach, more specifically.
Marx would have been the first serious thinker to place transformation, or rather, the articulation between transformative theory and practice, at the center of his system. But the very idea of transformation is not valid in itself. Not every transformation leads to improvement. And what should guide transformation was postulated by the best of philosophy, Marx (2005) recognizes in Introduction to the Critique of Hegel's Philosophy of Right, where he praises the great achievements of philosophy, particularly classical German philosophy.
Marcuse, in turn, by situating classical German philosophy in its clash with British empiricism, helps us to better understand the ways in which Marx is both an heir and a critic of both currents: “German idealism defended philosophy from the attacks of English empiricism, and the struggle between the two schools did not simply mean the clash between two different philosophies, but a struggle in which philosophy as such was at stake.” (Marcuse, 1978, p. 28)
It was also an ethical-political conflict. It is true that the materialist character of empiricism was important to Marx. The problem was its contemplative aspect: “If experience and habit were the only sources of knowledge and faith, how could man act against habit, how could he act in accordance with ideas and principles not yet accepted and established? Truth could not differ from the established order, nor could reason contradict it. This resulted not only in skepticism but also in conformity. Empiricism, by limiting human nature to the knowledge of the “given”, eliminated the desire to transcend it […]” (MARCUSE, 1941, p. 31-2)
Marx's critical appropriation of British political economy, as well as French socialism, is well known. We will focus here on his relationship with German idealism, which has also been widely studied, but with a particular focus, mediated by Marcuse's study of Hegel and his critical fortune, particularly in Marx's thought, together with Ernst Bloch's exegesis of 11 Theses. The objective of this approach, it is perhaps worth remembering at this point in the exhibition, is to explore elements that allow us to discuss Artificial Intelligence, as a notion and phenomenon, in light of the theoretical framework arising from this critical tradition.
Hegelian dialectics and its attention to the contradiction inherent in historical reality itself as the driving force behind its transformation occupy a prominent place. However, it is also worth highlighting Kant's categorical imperative that subjects cannot be treated as objects, linked to Feuerbach's critique of religion, as can be read in Marx's praise for the radicalism of the "German theory", which concludes the critique of religion "with the doctrine that man is the supreme being for man himself", leading to the "categorical imperative to overthrow all conditions in which man appears as a humiliated, enslaved, abandoned, despicable being". (Marx, 2005, p. 151)
Ludwig Feuerbach's critique of religion denounces the inversion between creator and creature in the relationship between human beings and divinity, but remains in the abstract and contemplative sphere:
Feuerbach's “anthropological critique of religion” derived the transcendent sphere as a whole from the fantasy of desire: the gods are the desires of the heart transformed into real beings. At the same time, through this hypostasis of desire, a duplication of the world into an imaginary and a real world emerges, with man transferring his best essence from the here and now to a supra-earthly beyond. (Bloch, 2005, p. 259)
The most famous of the Theses, the 11th, is precisely the one that claims that philosophers have limited themselves to contemplating the world and that the time has come to go further and transform it. This would be both the end and the fulfillment of philosophy, as the realization of its greatest achievements. And what would these be? To conclude that the world must become an order of reason, that the essence of reason is freedom and that this must imperatively have universal validity, that is, apply to everyone:
[…] Marx’s continuation of Feuerbach’s anthropology as a critique of religious self-alienation is not only a consequence but a renewed disenchantment of Feuerbach himself or of the ultimate, anthropological fetishization. Thus Marx leads the ideal-generic man, via mere individuals, to the ground of real humanity and the possible humanitarian stance.
To do this, it was necessary to look at the processes that are really at the root of alienation. Men duplicate their world not only because they have a torn, desiring consciousness. This consciousness, together with its religious reflection, originates, rather, from a much closer division, namely, the social division. Social relations themselves are torn and divided, they reveal a below and an above, they show struggles between these two classes and nebulous ideologies of the above, of which the religious one is only one among many. For Marx, the work that still remained to be done was precisely to find this closest to the worldly foundation – itself an immanent in relation to Feuerbach's abstract-anthropological immanent. (Bloch, 2005, p. 261-2)
The proletariat, as a universal class, subordinated to the irrationality of exploitation, deprived of property and freedom beyond the letter of the law – formal freedom –, would be the social subject responsible for transforming an irrational social order – despite the achievements of the French Revolution, since it was still based on the exploitation of man by man – into a rational order, in which individual and collective freedom would not be in contradiction, but would be mutually conditioning. This necessarily involves a transformation in the property regime, with the end of private ownership of the means of production, etc.
“[…] without the partisanship of the revolutionary class position there is only retrograde idealism in place of forward praxis. Without the primacy of the head to the end there remain only the mysteries of dissolution in place of the dissolution of mysteries. Thus, in the ethical conclusion of Feuerbach’s philosophy of the future both philosophy and the future are absent; Marx’s theory, in function of praxis, put both into operation and ethics finally became flesh.” (Bloch, 2005, p. 270)
However: “[…] what was it that the starting point of the Eleven Theses, that is, the incipient philosophy of revolution, discovered? It is not only the new task of the proletariat, however decisively it has moved away from contemplation, however refusing to accept or even eternalize things as they are. Nor is it only the critical-creative legacy received from German philosophy, English political economy, and French socialism, however necessary these three ferments, especially Hegel’s dialectic and Feuerbach’s renewed materialism, were for the formation of Marxism. What definitively led to the Archimedean point and thus to theory-praxis has not yet appeared in any philosophy […] Until now, all knowledge has essentially referred to what has passed, since only this is contemplable. Thus, the new has remained outside its understanding; the present, where the becoming of the new has its front line, constitutes an embarrassment.” (278-9)
We have attempted above to summarize the transition from a critical but still contemplative rationality to the notion of praxis, with the aim of contributing to the debate surrounding the critique of the artifices of cunning that make up so-called Artificial Intelligence and that make it essentially irrational, and therefore not intelligent, from the point of view of its victims, direct victims of fraud, exploited workers, monitored users, deceived crowds, diffuse rights attacked, democracies threatened.
Final considerations
The notion of intelligence needs to be thought of in teleological terms, related to means and ends. As a first approximation, an action should be considered intelligent if the means employed favor or ensure the achievement of the desired end. This is unquestionable, but it does not resolve the problem of how intelligent the ends deserve to be classified.
Nothing is or can be purely artificial, because nothing exists beyond nature, except processes mediated by human action from nature. From chipped stone and bonfire to algorithms. Artificial is, therefore, an indication that there was human interference in that result, or rather, it means something that would not have been without that interference. It is not, therefore, something outside of nature, but the fruit of human mediation.
In the case of Artificial Intelligence, this involves engineers, programmers, producers and consumers, without forgetting the owners and shareholders of the platforms, whose telos DM-D' remains the decisive mediation amidst the complex of intervening mediations in action.
“Only man has the power of self-realization, the power to be a self-determining subject in all processes of becoming, for only he has an understanding of what potentialities are, and knowledge of “concepts”. His very existence is the process of actualizing his potentialities, of adapting his life to the ideas of reason. Here we find the most important category of reason, namely, freedom. Reason presupposes freedom, the power to act in accordance with knowledge of the truth, the power to adjust reality to potentialities. […] Freedom, in turn, presupposes reason, for only comprehensive knowledge enables the subject to conquer and exercise this power.” (Marcuse, 1978, p. 22)
Freedom is then understood as the power to act in accordance with rational knowledge of the truth and the power to shape reality according to its potential. Logos e metis they mix here, mediating freedom and being mediated by it.
However, the freedom of shareholders and platform owners to subsume the world's intellectual work and leisure for free to enrich themselves, indifferent to the consequences of their turbocharged Frankenstein, is pure metis it's too little Logos. At the same time, would flat-earthers or more dangerous illusionists be effectively exercising their freedom of expression by spreading nonsense on digital networks? Or would they be led by opportunists, most often followers of irrationalism, if not outright fascists?
Is freedom conceivable without truth, desire, courage, fear, pain, and pleasure? Is intelligence conceivable without these things, unless it be an abstract, purely formal, or merely instrumental intelligence?
A machine, essentially and forever incapable of these things, because it is not organic, because it is not alive, cannot possess concrete intelligence and freedom, nor even cunning. It can only act within the complex of social mediations in order to meet the most influential demands.
Certainly, the mechanical processes sweetened as Artificial Intelligence are the result of human action on natural powers and things. Who are these humans? When and where do they act? In what ways? With what results? For whom?
Who: gold, columbite and tantalite miners; engineers; mine owners; politicians who legislate property and labor relations in relation to mining activities, processing, circulation, purchase, sale and application of mining results, which well before and at the same time as it is a source of daods, is “[…] valuable minerals, such as coltan and gold, for the electronics industry. Coltan – a mixture of two minerals, columbite (from which niobium is extracted, which has superconductor properties) and tantalite (from which tantalum is extracted, used in the manufacture of small capacitors) – is a metallic ore used in most electronic devices, such as smartphones, notebooks and other computers. […] Since Brazilian law, until 2023, was based on the seller’s declaration of good faith to legitimize the sale of Brazilian gold on the market, it is difficult to determine the percentage of gold illegally extracted from indigenous reserves (such as the Yanomami) that is contained in each smartphone”. (Bezerra, 2024, p. 49-50)
When: in a time of crisis of American hegemony, strengthening of China and rise of the extreme right around the world. Where: Africa, Yanomami lands. In what ways: under semi-slave labor and even well-paid labor, obtaining astronomical profits from the overwork of the two previous agents, miners and programmers, distributing dividends to the next group of agents in the bourgeois states, political lackeys of capital, and to those responsible for circulation, which in turn involves a complex division of labor and property all its own.
Artificial intelligence, or rather, the artifices of cunning are distributed unequally, as is the chain of command and execution of the actions necessary for its existence, without forgetting the more or less gratifying or catastrophic results for the various agents involved, including the users of the systems.
Among the catastrophic ones, insufficiently discussed in the public sphere, is a socio-technical infrastructure composed of extremely expensive and ecologically destructive data centers, which consume enormous quantities of electricity and water. How much intelligence, how many tricks! Can there be an irrational intelligence? Can there be a stupid reason, that is, a dumb and brutal one? Intelligence is irrational and reason is stupid from the point of view of its victims: no one considers that being deceived, tricked, exploited, crippled, massacred is an intelligent or rational result.
Artificial intelligence is a new type of artifice used by the clever and the astute. Or, in more formal terms, artificial intelligence, as we have seen, is the most recent result of the historical tendency to subsume labor under capital, both formal and real. This phenomenon, together with the tendency for constant capital (dead labor) to grow in relation to variable capital (living labor) in the organic composition of capital, simultaneously generates an increase in productivity, unemployment, and a fall in the rate of profit, as it reduces the active presence of the only source of added value in the production process, whether material or symbolic: variable capital, living labor.
This trend, however, is not a fatality of a mystical or cosmological order, like the apocalypse or the explosion of the sun, but of a sociohistorical system that began, grew amidst so many crises, is experiencing yet another moment of contradiction between the development of productive forces and the current relations of production, and will be overcome by a better one. Or not. It depends in part on us, who discuss these things.
The framework presented presents legal challenges, related to the regulatory debate on platforms and Artificial Intelligence itself, which touch on sensitive issues such as freedom of expression, which in turn confronts the notions of individual and collective freedom, especially in the face of growing waves of misinformation or mass fraud.
There are related economic challenges, which involve the demonetization and possible imputability of disinformation agents on a scale that poses a risk to democracies and diffuse minority rights, given the concentration of power of the platforms, with their armies of lawyers and representatives in conservative congresses.
There are narrative challenges, because what is at stake is historical truth itself, along with the credibility of institutions recognized since modernity as cognitive authorities: the press, science, the rule of law, which obviously should not be shielded from criticism, but criticism is not the same as slander, defamation, sabotage. It is important, therefore, to also face the problem of truth and freedom, without losing sight of Heller's (2004) advice about not confusing the necessary prudence in dealing with the truth, due to the lack of absolute certainty, with a surrender to relativism.
Facing this set of challenges requires seeking synergies between the virtuous but insular actions currently underway by academics, governments and activists; denouncing and politically pressuring for regulation of digital platforms that profit from disinformation in its most pernicious forms: racism, misogyny, LGBTQIA+phobia, denialism, revisionism, all of which come together in the neo-fascist bundle; promoting critical competence in information and communication on a large scale; the establishment of national and popular digital sovereignties, through investment in public platforms.[ii]
* Marco Schneider He is a professor at the Department of Communication at the Fluminense Federal University (UFF). Author, among other books, of The dialectic of taste: information, music and politics (Circuit).
*William France is a journalist and PhD student in Information Science at Ibcit-UFRJ.
*Luiz Claudio Latgé is a journalist and has a master's degree in Media and Daily Life from UFF.
Expanded version of an article originally published as a book chapter at the Instituto de Filosofia de Cuba.
References
ANTUNES, Ricardo. The Senses of Work. Essay on the Affirmation and Denial of Work. São Paulo: Boitempo, 2006. 261 p.
BEZERRA, Arthur Coelho. Information regime and class struggles: reconstruction of a concept in the light of the critique of political economy. In: Ci. Inf. Rev., Maceió, v. 10, n. 1/3, p. 1-14, 2023. Available here.
BEZERRA, Arthur Coelho. Technology and precarious work: a critique of the political economy of digital capitalism. In: The Social in Question – Year XXVII – nº 58 – Jan to Apr. /2024, p. 37-56. Available here
BLOCH, Ernst. The Hope Principle. Volume 1. Rio de Janeiro: EdUERJ; Counterpoint, 2005.
BLOCH, Ernst. Commentary on Theses on Feuerbach. In: The Principle of Hope. No date. Available here.
CAPURRO, Rafael. Pseudangelia – Pseudangelos. On False Messages and Messengers in Ancient Greece. Contribution to the virtual Seminar with Prof. Maria Bottis (Professor of Information Ethics and Law) and colleagues at the Ionian University (Corfu, Greece), April 21, 2020. (pdf). Students questions and tentative answers (pdf). In: Information 25(1), 2020, pp. 106-131. http://www.capurro.de/pseudangelia_english.html.
HE, J. & DEGTYAREV, N. AI and atoms: How artificial intelligence is revolutionizing nuclear material. Bullet of the Atomic Scientists, 2023. Available here.
HELLER, Agnes. Everyday life and history. São Paulo: Paz e Terra, 2004.
HOPFIELD, J. J. Neural networks and physical systems with emergent collective computational abilities. Proceedings of the National Academy of Sciences, v. 79, n. 8, April, 1982, p. 2554-2558. Available here.
IBM. What is Artificial Intelligence (AI)? Available here.
KANT, Immanuel. Foundation of the metaphysics of morals. New York: Routledge, 2009.
MARCUSE, Herbert. Reason and revolution. Hegel and the rise of social theory. London: Routledge & Kegan Paul Ltd, 1941.
MARCUSE, Herbert. Reason and Revolution: Hegel and the Advent of Social Theory. Rio de Janeiro: Peace and Land, 1978.
MARX, Carl. Chapter VI Unpublished from Capital. Results of the process of immediate production. New York: Routledge, 1985. 169 p.
MARX, Carl. Critique of Hegel's Philosophy of Right. Sao Paulo: Boitempo, 2005.
MARX, Carl. The capital. Book I, v. 1. Rio de Janeiro: Brazilian Civilization, 2002. 576 p.
MARX, Carl. The capital. Book I, v. 2. Rio de Janeiro: Brazilian Civilization, 2003. Pp. 577- 929.
MLUHAN, Marshall. The Means of ccommunication cas extensions of man. Sao Paulo: Cultrix, 1969.
SCHNEIDER, Marco. The Age of Disinformation. Post-truth, fake news and other traps. Rio de Janeiro: Garamond, 2022.
TURING, A.M. Computing Machinery and Intelligence. All, Volume LIX, Issue 236, October 1950, p. 433–460. Available at here.
Notes
[I] There are other business models for digital platforms, such as Uber and Airbnb. We are talking here about the models of companies like Amazon, Google and Meta, which can be considered new media.
[ii] We would like to thank Faperj, CNPq and Capes for the research support grants.
the earth is round there is thanks to our readers and supporters.
Help us keep this idea going.
CONTRIBUTE
