“Homo Deus”(Yuval Noah Harari) Critique and Commentary from a Philosophical Perspective(Aristotle, Kant, and Wittgenstein) Part Eight: Action, AI and Totalitarianism

Hits: 281

The Liberal order, according to Harari, is defined in terms of individualism human rights, democracy, and the free market and is also a form of religion. Human rights as we learned earlier are figments of the imagination and Humanism, a term traditionally closely associated with liberalism, is also more or less defined in terms of a romantic solipsistic individualism that does not have very much in common with our traditional notion of an ethical humanism steered by law, and reason. Ever since Kant associated ethics and human rights with freedom, freedom also became a more systematically characterized concept than it was when it was referred to by Hobbes as that which citizens have to partly abandon in order for the Leviathan or the commonwealth to provide security for nation states citizens.

As was the case with the difficulty of seeing the principle of the soul through a microscope or via imaging techniques we are now also informed that the free will cannot be detected with scientific instruments either. Looking for freedom with such instruments is of course what a philosopher would call a category mistake. Harari claims that the attribution of free will to humans is a fact and that is true, but some facts are categorial such as “all men are mortal”: no observation would ever reveal the counterfactual that a man is immortal and trying to base this conceptual fact on purely observational grounds is failing to appreciate the logical and categorical nature of conceptual truths. Hume once tried this line of reasoning with the self and pointed out that we are not able to observationally detect the “self” and therefore the self-did not exist. Kant pointed out a number of objections to this line of empiricist thinking, amongst which was the self or soul being an idea of reason(a principle) rather than a phenomenal thing to be encountered in the phenomenal world. He pointed out, in other words, that the initial premise that we cannot observe the self is ambiguous. If the self is a principle that cannot be observed because it is a condition of what it as that we are observing then it cannot warrant the conclusion that there is no such principle or condition. Similarly, if freedom is an idea or principle of ethical activity then claiming as the author does, people are free because they “feel free” is incoherent. When people say they feel free it is a negative judgment which is in focus: the judgment namely that no one is preventing them from doing what they wish to, which in turn focuses attention on the fact that freedom is an idea or condition qualifying action and can not qualify sensation:actions and sensations are different psychological entities and even if sensations might peripherally be associated with action there is no logical connection between these logically different entities(See O’Shaughnessy’s “The Will: a dual aspect theory). There is even a difference in the causal origins of these two entities with action having an immediate psychological cause and sensation or feeling having an immediate physical cause. This discussion also accords well with the Kantian ontological distinction between “what happens to man” and “what man makes of himself”: therefore, all talk about there not being a self “present” in the stream of consciousness is an unnecessary spiritualist reification of the concept. One of the immediate psychological causes of action is, for example, desire which certainly must be an occupant of the so-called stream of consciousness, but one should not succumb to the scientific temptation of reducing action to mere bodily movement that can be observed, for two reasons: firstly there are mental actions such as trying to remember a name and secondly, there is trying to sustain a mental image. Both are mental actions sharing the logical structure of being active (one can ask people to stop doing them) and both share the relation to the psychological realm rather than to physical origins. Secondly, action is a concept transcending the ontological realms of thought and existence and thereby is what O´Shaughnessy calls an apriori metaphysical concept that only a self-conscious consciousness or language user is capable of comprehending. We know behaviourism is behind much of the confusion related to consciousness and the understanding of language. Having being forced to abandon experiments with humans because of the difficulty of controlling and measuring the variables, experiments with animals only succeeded in producing the limited results they did because of the presence of an animal form of consciousness and the consequent existence of teleological intentional behaviour. Attempting to generalize these results to humans failed because self-conscious consciousness and language use is a higher form of life. O Shaughnessy points out how the way in which we linguistically demarcate the concepts of perception, action, and consciousness includes an ineluctable first person identification of the occupants of our stream of consciousness:

“While action and perception and consciousness have no tendency to cause any single phenomenon-or set-of-phenomena in one setting, they nevertheless have a characteristic “outer face” in the following sense. The simpler the organism in which they occur, the more they figure in causal transactions from inner to outer that are readily interpretable to a third person other. For example, nearby perception by an insect of its natural prey or predator will very often cause movement: and this movement in this situation is readily interpretable. Then I would suggest that such an “outer face”, call it C-phi must have been that via which the psychological item in question first came to the consciousness of some third person other.Yet it is one thing to know of and notice such psychological items in another via C-phi, it is quite another to relate to them as does a self-conscious consciousness. That is, to be in a position to notice and know of them/in oneself and another/ under linguistically demarcated concepts. This huge development depends on C-phi- but also on the internal psychological setting of the outer phenomena, call it C-psy. For example, physical action Phi gets thus conceptually demarcated via a C phi which includes “say” the presence of a quarry, and a C-psy which includes desire. Now C phi and C-psy are interdependent: indeed C-psy rationalizes C-phi…In short, the development of self-conscious conceptualized knowledge of the physical act is made possible by C-phi and C-psy , and ultimately by C-psy.”(Volume 1 p80)

C-psy is inaccessible to the scientific method. Pretending that the firing of neurones is logically equivalent to the desire for the quarry is a mind-brain identification fallacy but it is the only move available if one does not want to deny the existence of the internal phenomenon of desire. The science of Aristotle and Kant it should be pointed out had no difficulty in accommodating transcendental and metaphysical truths and did not need to resort to the spiritualization of phenomena that cannot then causally interact with the physical world.

One of the major arguments used against the notion of the self is the following:

“Liberals believe that we have a single and indivisible self. To be an individual means that I am in-dividual(indivisible). Yet my body is made up of approximately 37 trillion cells and each day both my body and my mind go through countless permutations and transformations…..For liberalism to make sense, I must have one–and only one–true self, for if I had more than one authentic voice how would I know which voice to heed in the polling station, in the supermarket, and in the marriage market…However, over the last few decades the life sciences have reached the conclusion that this liberal story is pure mythology…if I look really deep within myself the seeming unity which I take for granted dissolves into a cacophony of conflicting voices, none of which is “my true self”. Humans aren’t individuals. They are dividuals.”

The use and abuse of the term “liberal” in this work is similar to the use and abuse of the term “humanist”. If one connects liberalism to liberty and the positive concept of freedom then Kant must be the liberal par excellence in virtue of the fact that his reflections on the concept of freedom are the most systematic account of the concept we have. Kant, of course, reflects on the parts of the self which for him are sensibility, understanding, and reason. For him, it would be perfectly consistent to maintain that someone could steal something in spite of hearing the voice of reason within telling the individual not to do the deed. When the individual finds himself in court and is asked why he stole the item he may, like a good scientist, cite a number of causes: he needed to pay his rent, his mother left him when he was 6 months old, his father became an alcoholic, he fell in with a gang of thieves etc. Now if the legal system were based on the causal principle of science, the judge would have no alternative but to release the man. One cannot be blamed for one’s choices if one has no control over them. But the criminal can be blamed: he heard the voice of reason and if he did not, he ought to have developed his reasoning about his social existence to the extent that the voice did protest the deed. The criminal is blamed on the grounds of one of the key concepts of social science, namely freedom. The judge in sentencing him refers to the principle that the criminal could have chosen not to steal the item. He is sentenced, that is, for not using his freedom and reasoning powers. Now let us assume, incidentally, that our criminal is a scientist and as a Parthian shot shouts out that it was not he that committed the crime but some other person, the person he was six months ago is not the person he is today, literally millions of cells and chemicals are different in his body, therefore, he is different. This narrative indicates the irrelevance of scientific concepts in the realm of social science and law. We would not find Kant the liberal denying the validity of the assumption of the legal system that the criminal is the same person that committed the theft. The Sensible part of the mind may have been behind the action but the criminal will understand the reasoning of the legal system if he is not mentally ill. If he is mentally ill he may continue to insist that the voice telling him to commit the theft was not his. He may even complain about a cacophony of voices. There is a contradiction in insisting that the number of cells changing is logically relevant to the claim that there is not an individual self and many materialists have themselves attempted to patch up their faulty reasoning by reference to the functional continuity of the organs of the body that may or may not be sufficient to claim the unity of the self. For Aristotle, the hylomorphic philosopher who acknowledges the truths of materialism, the functional unity of these organs were a partial explanation of the unity of the self, i.e. in modern terminology, they were perhaps a necessary but not sufficient condition of the unity of the self. Functional unity, however, commits one to teleology: the organ’s function is for the sake of the individual and many physical scientists do not want to be associated with this kind of “backward causation” as they incorrectly call this type of explanation. A teleological explanation is however very much tied up with our conceptualization and linguistic demarcation of action. O’Shaughnessy points to instances of actions in which observation based descriptions of instrumental action would be confined to just the movements of the body. So when in walking to the shop in order to buy some milk, I say so upon being asked, and the scientist obsessed with observation claims that to be an illegitimate description of my action because all I can scientifically be said to be doing is walking. This, after all, is all that can be captured on his video of me on the way to the shop before being asked about the matter.

After discussing the respective different functional roles of the right and left hemispheres Harari claims that Daniel Kahneman’s experiments prove the existence of an experiencing self that remembers nothing(a moment to moment consciousness) and a narrating self that remembers selectively and is duration blind. It is this narrating self, the author argues that is behind the liberal belief that we have a single enduring self. This narrating self also spins a plot from its stream of consciousness that might be fictional, containing omissions, and changing continually to such an extent that one plot may be contradicting another.

The author goes on further to suggest that the self is a result of an imaginary story that we tell ourselves that selects items which give me images of who I am and what I am doing with my life. Lives can be lived comically tragically or religiously but in the end, it is argued, everything is an imaginatively based story which the author seems to be able to distinguish from the truth in some fashion in a way that he thinks the people he is talking about do not possess the capacity to do. He assumes that the language located in the left hemisphere is one without a truth function that subjects putative facts to observational tests, testimony from other witnesses, different kinds of reasoning processes and different theories. If Aristotle is a liberal as he ought to be, given his suggestion of the importance of an educated middle class to the fate of the political community, why would he reject the above account so categorically? Here are some of the grounds he would use to reject the above account: Reason uses arguments and theories to determine the truth of putative beliefs that can occur uncontested in a narrative: Experience can be organized by narratives and also organized at a higher level by knowledge: My life is the true story of who I am where I came from what I am doing–it characterizes my epistemic relation to myself. To the extent that I am capable of telling this true story is the extent to which I “know myself”.

Harari concludes this section with the claim that:

“The free individual is just a fictional tale concocted by an assembly of biochemical algorithms.”(P354)

…thus dispelling the myth that a story is told by one free individual to another which needs to be done if one is to accept the totalitarian story of the future which we will be treated to in the next section.

The arguments of Kant pertaining to the dignity and worth of the individual have made no appearance in either of Harari’s works thus far. If Kantian liberalism has survived the totalitarianism of the last century then it must be because of these arguments. The author claims, however, that: liberalism was successful because it makes a lot of sense from a variety of different perspectives to treat the individual as having a value(that appears largely instrumental). Militarily, of course, the value is quantitative and instrumental: hands are needed to pull triggers and push buttons.

This makes liberalism a utilitarian Philosophy that resonates with the political philosophies of Hobbes and the Mill’s. We are all familiar with the futile attempts of utilitarian ethics to explain the worth of an individual and his freedom. Attempts that at the political level justify the unjust tyrannizing of the minority by the majority however insane the policies of the majority are. At the individual level, the liberalism of the Mills and Hobbes will allow the individual who is mentally ill the freedom to destroy himself. Utilitarianism was a political philosophy inspired by the science of the 16th and 17th century. It relied on a method of resolving wholes into their parts and attempting to re-compose them back into the whole(the former process being of course much easier than the latter) and it also relied on an assumption of linear causation applied best to the billiard balls of Hume cannoning off each other. Utilitarianism was science “applied” to social phenomena with a teleological twist: man pursues happiness it was claimed. The test of whether something was good or not was a consequential test and Bentham’s sovereign principles of pleasure and pain were the key indicators of a man’s happiness. If a man was happy, irrespective of whether he met the Kantian condition of deserving his happiness, this was a sufficient test of the ethical good. What are we to say of the happy man who then asks himself the Kantian question and upon realizing he does not deserve to be happy is now unhappy? This unhappiness is, after all, a consequence of a consequence and if consequences determine what is good why should one particular consequence be preferable to any other? This is, of course, a variation of Aquinas’ double effect theory that is a standard objection to any consequentialist ethical theory.

Ethics is, of course, a figment of the imagination for Harari as is human rights which are tied not to man’s imagined happiness but to his actual dignity and worth. In his reasoning about the utility of man for the economic or political system, he notes that artificial intelligence is decoupling intelligence from consciousness and that there is no guarantee that this will not lead to man becoming superfluous at the point at which all occupations can be performed more efficiently by computerized robots. This is a serious prediction. Hannah Arendt pointed to the consequences of the industrial revolution when large numbers of men became superfluous in Europe and created the economic and political conditions for two world wars. It should, however, be pointed out that it was precisely the political and economical utilitarian value of these men which contributed to their alienation. Even advanced Educational systems did not suffice to convince the masses of unemployed that they possessed a value in being human.

Sometimes there is something of the air of a science fiction film hanging over these reflections. Firstly, the reflections ignore the variable of freedom and consent. Would people consent to have AI based teachers and doctors and politicians? Secondly, it is not clear that algorithms will be able to capture the essential elements of teaching to take one example. Everyone in education knows how difficult it is to change one small component of the educational system: there just is no hope of agreement over what constitutes a good education. There must literally be hundreds of thousands of algorithms involved in the educational system and we look forward to seeing the conference which will discuss where to start. Let us, however, explore this totalitarian vision to its logical conclusion. Algorithms are going to take over the world, they will own things and employ people in the way gods did over 5000 years ago.

The author appears to believe that if we provided Google free access to our biometric devices, our DNA scans, and our medical records it would provide a better solution to the problems in our lives than the narrating self with its “cooked up stories. The Google algorithm will help us make choices in the supermarket, in the polling booth and the marriage market:

“The new technologies of the 21st century my thus reverse the humanist revolution, stripping humans of their authority, and empowering non-human algorithms instead.”(p401)

All of this will apparently be based on the life sciences conclusion that a living organism is just a collection of algorithms. This in its turn will transform medicine from an ethical project of healing the sick to an elitist project of upgrading the lives of the healthy. Elites will want these services, it is maintained and they will behave no differently to the elites of history in relation to focusing upon the needs of the poor: they will focus on themselves.

These super-humans will, it is argued, abandon their “liberal” roots and treat humans “the way 19th century Europeans treated Africans”.

It is difficult to treat some of these reflections academically because of their “imaginative” science fiction-like character against the background of the absence of philosophy, ethics, philosophical psychology, and law, but the purpose of this review thus far has been to present this absent background. With this in mind it is difficult to believe that given our natures and the knowledge we have of what happened in history the last time we encountered the above-mentioned phenomena when “science” was left to dominate the field of humanistic explanation, we will travel this totalitarian road again without recognizing its landmarks. We are free to choose not to go down this yellow brick road.