Towards the beginning of Homo Deus: A Brief History of Tomorrow, Yuval Noah Harari discusses what he expects to be the three great projects for the human future: the quest for immortality, the quest to re-engineer human beings to be happy all the time, and the quest for god-like powers. It sounds like we are about to be exposed to standard futurist fare, with bold predictions of an eye-popping future. The reader quickly sees, however, that for better and for worse, that is not Harari’s intention.
He admits, for example, that he is skeptical that the project for immortality will succeed any time soon. Yet almost immediately he notes, “The scientists who cry immortality are like the boy who cried wolf: sooner or later, the wolf actually comes. Hence even if we don’t achieve immortality in our lifetime, the war against death is still likely to be the flagship project of the coming century.” He acknowledges that immortality could have dark sides and discusses a few of them, but given our fear of death or our belief in the sanctity of life, he concludes that the war against death will have “irresistible momentum” and “seems to be inevitable.”
It looks as if Harari is predicting an immortality project but withholding his approval. But is he really predicting it? He resists this characterization, writing that his description of future projects is not a “prophecy,” but rather “a way of discussing our present choices.” Discussion makes sense if we have a choice about the future. Yet it was Harari himself who set up the whole quest for immortality in such a way as to make it appear to be “inevitable” or to have “irresistible momentum.” Are we free to reject it, or not?
Throughout the book, on this topic and others, Harari’s ability to play tug of war with himself can be confusing. But it also illuminates the dilemmas posed by the particular forms of scientific materialism and the fact/value distinction that ground his argument.
It turns out that the title of the book, Homo Deus: A Brief History of Tomorrow, is misleading. While the book uses a future-oriented rhetoric, substantively it is actually less about the future than it is about arguing for a central thesis concerning the nature of life itself. As Harari tells the story, the great insight on which contemporary biological sciences are founded, the insight that has swept the field of all other understandings, is that living things are bundles of algorithms. He writes: “An algorithm is a methodical set of steps that can be used to make calculations, resolve problems and reach decisions.” The biochemical processes that give life to living things and keep them alive are the material instantiation of algorithms, themselves products of natural selection, which itself is a decision-making algorithm of sorts. A virus, a gnat, a pig, and a human being are all the same in this respect and, but for the materials that make them up, not so very different from vending machines.
A better title for the book might have been Algorithmic Man. This contemporary take on a materialistic understanding of life and human life has not yet cracked the problem of consciousness and self-consciousness, Harari admits. But that may not turn out to be any kind of serious problem, because even if consciousness is not algorithmic, intelligence is. Intelligence is the key to the future, and intelligence does not require consciousness.
In a characteristically confusing formulation, Harari presents a “thought worth thinking, even if it is not true:” consciousness is “a kind of mental pollution produced by the firing of complex neural networks. It doesn’t do anything. It is just there.” It probably does not have to be said that there is also no room for souls in Harari’s brief history of tomorrow. And it has no room for a self, either, if by that we mean anything more than the sum total at any given moment of the bundles of often conflicting algorithms that constitute us. If there is no self, then the question whether the self is free or not is obviously moot.
Revising Religion: The Rise of “Dataism”
To make a long story short, this understanding, which Harari stresses repeatedly is nothing more than the orthodoxy of contemporary life sciences, is to his mind completely incompatible with philosophical and political liberalism and, indeed, all forms of humanism—as Harari presents it, the idea that it is free, choosing humans who endow the world with meaning. Humanism is important because in pursing the project of modern natural science, which Harari stresses is unable on its own terms to make any value statements, we have made a bargain. In order to gain power over the world through science and technology, we gave up the idea that there is any meaning already embodied in the world, leaving meaning-making up to us. Hence the “religion” of humanism. Notwithstanding his odd use of the term “religion” to apply to all normative discourse, Harari argues, reasonably enough, that this bargain has been pretty spectacularly successful at creating the conditions for longer, happier, healthier, and more prosperous human lives such that we now can contemplate even grander projects like immortality.
But there will need to be a “different kind of deal” now that science has exposed the assumption behind humanism as false. Actually, it looks as if Harari might better have simply said, a different kind of “religion,” for the deal remains much the same: we need something to give meaning to a world that science has stripped of meaning. While he considers alternatives, Harari appears to settle on “dataism” as the appropriate new religion that will give us the values that we need.
Dataism turns out to be nothing more than the algorithmic view of life that Harari himself had previously articulated, but with the added normative propositions that more complex information structures are better than less complex ones, and that information wants to be free. What this means in practice is that the meaning of life is the accumulation and manipulation of data. The data that represent my life-world need to be free so that artificially intelligent algorithms can learn from them how to direct me to lead my life better than I could unaided. They need to be free so that data about my health can be used to find the most effective cures for me and others. The data that I embody as a collection of algorithms need to be shared so that that “bundle” that I think of as “me” can be given virtual instantiation, with all the consequent possibilities of manipulation, alteration, transmission, enhancement, etc.
By this path we reach a kind of immortality, at least so long as my virtual being is sufficiently backed up. And in this virtual world, algorithms could make us happy all the time, and grant us godlike powers. So dataism brings us back to where Harari began.
A Timid Book
Decades ago, roboticist Hans Moravec offered a similar but better thought-out vision of this future, imagining how the immortality of virtual beings would probably end up looking very different from that imagined by biological beings who seek immortality. He also foresaw the competition for computational resources that would likely produce a great Borg-like wave of intelligence flowing though the universe, assimilating all matter to its own purposes, which would be incomprehensible to beings such as we are.
But rather than present such a straightforward vision, however speculative, Harari once again confuses. His book ends with three big questions that he hopes “will stick in your mind long after you have finished this book.” The first and third are most germane to the issues taken up in this essay. First: “Are organisms really just algorithms, and is life really just data processing?” And third: “What will happen to society, politics and daily life when non-conscious but highly intelligent algorithms know us better than we know ourselves?”
How is the first question even a question, given that he has provided no serious reasons for us to think that what he claims to be a scientific consensus is anything but true? And why, if the book is not a prediction of things that will happen but a discussion of things that could happen, is the third question posed as “what will happen to society when” rather than “would we want to live in a society that”?
Despite its bold title, Harari could be accused of having written a rather timid book. He is too cautious to make a big prediction like Moravec’s, despite the fact that his three projects are arguably already well underway. Despite his advocacy, his closing question suggests that he is unwilling to assert that in reaching the algorithmic understanding the life sciences have come across something that is simply true. And by speaking of the “values” of dataism as a religion, he distances himself from those values; they may be consistent with science, but they do not follow rationally from science. Such values may be necessary to structure societies, but societies have been known to choose the wrong values. He may be more willing to highlight the irrationality of the values of biblical religions than the irrationality of dataism, but that does not mean he is freed from the corrosive consequences of his particular deployment of the fact/value distinction for his own ability to justify rationally the values of the data religion. Where he stands is unclear, probably because he has left himself very little ground to stand on.
Indeed, if there is no self, who has authored his book? Did he not experience his writing as a combination of self-consciousness and intelligence? But what he could make of such a combination is far from clear. If the book proceeds from intelligence, he is committed to the view that his opus stands as the net result of various and sundry conflicting algorithms that, insofar as they might be modeled, could produce the same result. Any role for consciousness would reintroduce the “thought worth thinking, even if it is not true:” his own book, like all others, would be a “a kind of mental pollution produced by the firing of complex neural networks.”
What a great marketing opportunity and teaching moment it would have been to give authorial credit to “The Algorithms formerly known as Yuval Noah Harari.” But, bless his soul, he missed it.