The artificial man
This article was published in a reduced version on the Truth of January 31, 2020 with the title "Artificial intelligence does not exist but serves to make us like machines".
Not a day goes by without we come across the announcement of new and increasingly daring applications of artificial intelligence : that of the indicative future that will drive cars, diagnose diseases, manage savings, write books, settle disputes, demonstrate unsolved theorems. He will do everything and do it better, so that the writer imagines the coming times in which man will become "obsolete" and will be progressively replaced by machines, until he proclaims with concealed orgasm the advent of an apocalyptic "government of robots". However, this talk of new things is not new. The phantatechnical projection has enchanted the public for about two centuries, since that is when "the religion of technicality" has meant that "all technical progress [appeared to the masses of the industrialized West] as a perfecting of the human being himself" (Carl Schmitt, Die Einheit der Welt ) and, in anchoring this improvement to what is not human, has given him the illusion of an unstoppable and glorious movement. Like all religions, even that of "technicality" produces as a corollary of the "sacred texts" of officiant-technicians an apocryphal counter-song of popular legends in which the hopes and fears of the assembly of devotees are transfigured. Legends do not need to investigate plausibility, but the meaning.
By artificial intelligence (AI) we mean technologies that can simulate the abilities, reasoning and behavior of human beings. It is therefore difficult to understand from what point onwards the AI stands out, for example, from a small calculator that performs an activity proper to the human mind (the calculation, in fact), or from a personal computer that already simulates many skills of the reductionist man, that is, breaking them down into countable entities. The concept of AI therefore seems to be more optative than technical . It does not introduce any revolution but rather identifies, under a captivating label and of dubious epistemic solidity, the effort and the desire to develop increasingly sophisticated and powerful IT techniques. That these techniques always end up replicating, enhancing them, some functions of the human mind is obvious in definition, having been conceived and created precisely by that mind and precisely with that objective, from the beginning.
What fascinates the most recent applications of AI (i.e. the computer) is the growing ability to process inputs that are not rigidly formalized, such as photographic shots, somatic traits, inconsistent databases and – above all – language. The latter, a free and creative expression that is constantly regenerating (Noam Chomsky), is in fact the most important test case. To be fully deciphered requires not only the correct understanding of the complex syntactic rules, but also that of cultural, symbolic and emotional subtext and contexts (semantic understanding). Much more than an instrument, language is the embodiment of intelligence which in language is (re) created, translates the infinite streams of individual and social experience and communicates to others. The cybernetic assault on this impervious mountain, which is so reminiscent of the Babelic end in the chaos of languages, is only at its timid beginnings and has so far produced more or less promising mathematical metaphors to get closer to the mysteries of the mind. But however far we can travel in this direction, we would still remain ontologically far from the goal.
Intelligence is not only functional, that is, it does not limit itself to solving problems but poses them, formulates them and arranges them according to hierarchies . In this it is both conditioned and finalized by the subject that expresses it, it is also defined etymologically because it is an indissoluble and direct expression of its fines , of the limits that trace its unrepeatable and indivisible identity: desires, preferences, fears, affects, education, empathy and social relationships, faith in transcendence, corporeality, death and much more. If the logical-mathematical competence is common ground for all men and all machines, its exercise is instead enslaved by the gradations and changing conditions of each. A machine cannot reason like a man simply because it is not a man, just like a child it does not reason like an adult, a rich man as a poor man, a healthy man as a sick person, an atheist as a Christian, an aboriginal as a European etc. We must therefore ask ourselves the reason for this fiction, to deny the natural complementary relationship between the two domains with the pretense that they may, for some indeed have to, overlap until they merge and replace each other.
***
Here I venture two answers. If the intelligent subject looks inside ( intŭs lĕgit ) his condition in the world to formulate the objectives to be subjected to the logical and computational processes possibly delegated to an algorithm, that is, if he operates a "pre-analytical choice" (Mario Giampietro) that precedes and informs those processes , remains open the question of who would dictate ex multis goals the machines so that you can call "intelligent." Like Mario Draghi's "autopilot", AI will drive by itself and brilliantly overcome any obstacles, but towards what goal? Excluding the apocalyptic hypothesis (the one in which you would give it alone), it will inevitably be the goal registered in the code by its clients , who governing the code will enjoy the privilege of imposing their own ethical, political and existential models on everyone, wherever there is a processor and a network card. From the tangle of technical sophistications a more linear dynamic of man's domination over man would emerge, where the aforementioned fiction would be nothing more than a variant of the technocratic claim, of encapsulating the interests and motives of a class in a self-styled aseptic, unalterable procedure and necessary, thus removing them from the resistance of the other social forces. For those who let themselves be put in the strains by the "iron laws" of the economy (that is, by someone's priorities, according to their premises and their vision of the world) and by "science says it" (ditto), it will not be difficult to accept that the best solution is that created by the ventriloquists of the cybernetic and "intelligent" puppet.
The second hypothesis calls into question the limit of man, that is, his definition. Numerous indications suggest that, in common sentiment, the reduction of the subjective and plural set of human intelligences in an acephalous subgroup of erga omnes procedures is not intended as an impoverishment, but as a salutary overcoming of the swarming and unpredictable complexity of thoughts, behaviors and motives of the human anthill, and therefore of the "dangers" that would lurk there. The machine (one thinks) does not "keep family" and has nothing to lose or gain and therefore (one thinks) it can only do "the right thing" for everyone. From the exquisitely adamitic and Gnostic temptation to separate the weeds from the wheat prematurely arises the illusion of distilling infallible cognitive and decision-making processes – or in any case the best possible ones – by deactivating everything that can generate the "error": fragility, affections, inclinations, willful intent, but also and ultimately the irrefutable free will, the freedom of each. However, it has been seen that the indissoluble unity of intelligence and subject renders this illusion vain, the only result of which can be to move the arbitrariness in a few powerful hands, approving the rest. But it doesn't matter. The disgust and fear of the indisciplinable unknown man is stronger , the desire to unleash his arms by chaining and denying it in its distinctive essence, the thinking one. This desire for the non-living, to extinguish the dissonant chorus of the intelligences to reduce them to the zombie monody, is not measured only by the dreams – absurd even technically – to checkmate scam and corruption thanks to certified electronic transactions, to « eliminate (sic ) the mafias "with virtual money or frauds with voting machines , but even more directly from the moral eugenics of those who would like to expel" hatred "," fear "and other" bad "feelings (starting, ça va sans dire , from the most tender age, in extreme cases up to the ideological or physical seizure of childhood), silence the health, climate and economic specialists who do not parrot a thesis or put on top of the values politicians " honesty ", that is the demented execution, sicut ac machina , of a written law, thus imagining to program humans.
Let's look at reality. In practice, almost everything that is emblazoned on the magazines and parliaments of the AI label today – that is, digitization, in any way or measure it is applied – is very far from the requirement to bring the machine in the modus cogitandi et operandi of beings. humans to put themselves at their service. On the contrary, its applications imply the need or even the obligation for men to adapt to the machine procedures and to serve it. For example, if we really were dealing with a silicon humanoid intelligence that discreetly integrates into our mental structure, what need would we have to complain about the lack of "digital culture"? Shouldn't the burden of absorbing our culture affect the computer? And what good is it to teach "coding", the language of computers, to all children? To greet him (boom!) As " the new Latin "? Weren't robots supposed to speak our language? And why bother with electronic procedures, online forms, telephone assistants, PEC, apps, PINs, SPIDs, electronic registers etc. and upset the way we work and think in order to serve the "ready meals" to be digested by the computer? Why work twice as hard to send him our bills in the only format he can understand, when a mediocre accounting student would have been able to decipher them in every formal variant? And why spend time, money and nervous health to learn all these things? Wasn't "deep learning" a prerogative of the new algorithms? In short, one has the impression that the celebrated humanization of the machine is resolving precisely in its opposite: in a machinization of man . That the impossibility – we repeat it: ontological – of bringing the circuits into our ranks is producing the inverse result of flexing us, whatever the cost, to the rigid blindness of their law.
Of course, we can tell ourselves that these are only transitory paradoxes that serve to refine and instruct the AI so that the promised flight will soon emerge. But the truth is different and is there for all to see. It's that AI is our intelligence, AI is us . It does not speak to us of the progress of engineering and science, but of a desired progress of man called to strip himself of his defects – that is, of himself – to clothe himself with the foolish obedience, predictability and governability of electronic devices. If in the first phase this transition was imposed with the seduction of its advantages, from the personal computer in each home to free internet services up to mobile connectivity, in the following phase it must force the hand magnifying its benefits and making them in any case mandatory with some painful pretext: simplification, savings, progress-that-can't-stop. It is the phase in which we find ourselves today: that of 5G, of household appliances and cars on the net, of telephones that never turn off , of Kafkaes' telematization of public services and, together, of the stomach aches of those who worry, resist and he doubts, also because the promises of social improvement that accompanied the previous wave have all been miserably disregarded (that we are talking about the crisis since we speak of the "digital revolution" is a detail that not everyone has overlooked). In the meantime, someone, made bold by the innovator-coercion state, discovers the cards and prepares the third and final phase in which human beings will have to accept the machines also in their own body and no longer only in their thoughts, with the installation of circuits and processors connected to organs or directly to the brain . With many greetings to computers that become intelligent, intelligence will become a computer and man "will then be harnessed with prostheses before he himself becomes an artifact, sold in series to consumers who in turn have become artifacts. Then, having become useless to his creations, he will disappear "(Jacques Attali, Une brève histoire de avenir ).
***
This reflection would not be complete without asking: why ? What is the meaning of this process and of its being hailed as a holy hand, or at least as a challenge that must not be avoided? Undoubtedly someone will not mind the idea of tracing, controlling and conditioning every action or thought of each individual, anywhere and at any time. Nor to subject people to automatic processes and processors that leave no way out, devoid of reflection and empathy and therefore inexorably faithful to the mandate, were even the most atrocious. But even this dream or nightmare would not be new. The psychopathology of omnipotence and the will to dominate have always existed. Sadder, on the other hand, is the assent of the guinea pigs who lend themselves to such an experiment of subhumanism: from politicians who indulge global fashions and impose them on citizens, on the citizens themselves who imagine themselves pioneers of an arrogant age of silicon. There is, of course, a perception problem that cannot be just an effect of propaganda. A civilization that wishes to surpass the human can only be deeply discontented with itself. It is a disappointed and trapped civilization, unable to achieve the goals it has set itself but equally unable to reject them and recognize them as hostile to its own need for prosperity and justice. He cannot imagine an alternative and then imagines that the rotten link in the chain is precisely its members: the weak and irrational men, unworthy of the goal. Umso schlimmer für die Menschen! From this, from the creeping perception of an epochal failure, the illusion of saving oneself by chaining passengers to the seats and suppressing their safeguards to atone for the "Promethean shame" (Günther Anders) of not being up to their creatures, even political . To understand the roots of this despair it is therefore useless to question the engineers. Technologies, whether intelligent or not, are only the pretext of an escape from oneself that should be tackled at least by abandoning the childish temptation of "perfect" solutions and therefore extraneous to the irreducible mystery of a humanity in which "dust and divinity are mixed" (Fritjof Schuon), who lives in quantity while aspiring to the innumerable and disseminating its provisional truths in billions of souls. The compromise of a life that is certainly not geometric and reassuring like a video game will remain, but precisely for this reason it is possible, perhaps also worthy of being lived.
This is a machine translation from Italian language of a post published on Il Pedante at the URL http://ilpedante.org/post/l-uomo-artificiale on Fri, 31 Jan 2020 06:26:28 PST.