I recently finished a book, Blindsight by Peter Watts. I have some critiques even of the writing (it feels, at times, like he binge-watched a bunch of TED talks which he comments on throughout the book), but that’s not the most interesting part. If you want to see what will soon be a popular argument in a clearest and developed state, I think it’s worth a read.
Watts’s main thesis is that consciousness, self-awareness, and even empathy, are evolutionary quirks that at best slow down our technological development and at worst harm us. Mr Watts may claim that this was a thought experiment, that he doesn’t particularly like the results, or that the book has an unreliable narrator. The last point is certainly true, and the narrator’s unreliability is hardly subtle. However, he repeatedly argues that conscious thought is unuseful and even harmful. This is repeatedly moved from subtext to text. In a particularly obvious section:
So what if your lessons are all learned consciously? Do you think that proves there’s no other way? Heuristic software’s been learning from experience for over a hundred years. Machines master chess, cars learn to drive themselves, statistical programs face problems and design the experiments to solve them and you think that the only path to learning leads through sentience? You’re Stone Age nomads, eking out some marginal existence on the veldt—denying even the possibility of agriculture, because hunting and gathering was good enough for your parents. Do you want to know what consciousness is for? Do you want to know the only real purpose it serves? Training wheels.
The two main components of Watts’s thought experiment are his aliens and vampires. First are the alien scramblers, which are able to read thoughts, solve complex math problems, and learn and converse in English without ever having any semantic understanding or consciousness. (With respect to language, they just spit out contextually appropriate strings which get the other side to do things they would like, without any semantics.)
Vampires are somewhat closer to humans. They are an invented extinct hominid which must eat homo sapiens to survive, brought back through genetic engineering. They are more intelligent than humans but without empathy or any inner life. Repeatedly these creatures (as well as our narrator, whose brain is damaged to the point where he can see patterns subconsciously without understanding why) are shown as more intelligent and capable than “standard” humans. Toward the end, when our narrator is on his way back to Earth, it is implied that vampires have taken over human society:
Because we Humans were never meant to inherit the Earth. Vampires were. They must have been sentient to some degree, but that semi-aware dreamstate would have been a rudimentary thing next to our own self-obsession. They were weeding it out. It was just a phase. They were on their way.
…
Why wouldn’t they reclaim their birthright? Not genocide. Just the righting of an ancient wrong.
Lest you think I have unfairly taken his fiction writing out of context, here is a blog post by Mr Watts in which he advocates for sociopaths’ rights. After all, they participate in society, and may even be more successful in business and life than us mere baseline humans. Their inability to feel or reflect may be an evolutionary advantage, and we should accept this. It’s worth reading the post.
There’s nothing particularly new about the idea that sociopathy and unthinking efficiency are more valuable attributes than reflection or empathy. This line of thought was famously popularized by Ayn Rand, probably the 20th century’s most influential author. It ultimately derives from two assumptions:
- Instrumental reasoning is the only valuable application of the reasoning faculty.
- There is no objective moral value present in the universe.
To this, Watts adds a third:
- The only thing like a moral imperative is the evolutionary one to propagate oneself (survival) and one’s lineage (reproduction). This determines what survives and shapes the future, and should be our guiding force.
Assumptions one and two are really about the concept of telos, or the end to which something drives, and removing it from one’s worldview. Assumption three is the reassertion of teleological thinking, once other forms of telos have been removed. It is only made possible, chronologically, after the first two have been accepted.
Instrumental reasoning is the application of reason to solve particular problems, rather than the application of reason to understand things in and of themselves, or to assess value. It is a difference in the telos of reasoning. Is reasoning done for its own sake, to understand the world, moral behavior, or one’s self? Or is it used in order to accomplish something particular? There is no judgment implied in this distinction. Instrumental reason can mean applying reasoning faculties to design life-saving drugs or a deadlier bomb. Both processes share the use of reason to achieve some immediate goal. As long as humans have had the capacity for reason, instrumental reasoning has been present. The novel thing about assumption 1 is that instrumental reasoning is the only thing reason is good for, and the rest is worthless.
The second assumption about the absence of moral value is relatively new, or at least new as a widespread doctrine. The belief is that people might claim that they like things to be a certain way, that they have their own values which are related to their own well-being and that of people they care about. However, this is only defining a preference, and not any non-subjective right or wrong. A person who accepts this reasoning will typically include something about the importance of people helping each other, but on the basis that this may help the self. It is a social contract based on each individual working to shape society such that it leaves him maximally unencumbered.
Once you have the first two assumptions, they can then combine to reach their end-point: since morality does not exist, the only real imperatives in life are biological, those inevitably encoded into our natures. We are designed to survive and pass on genes, like all life. We find various things pleasurable that evolved in order to perpetuate the organism human, but there is no meaning or morality derivable from any of this. We can simply maximize this function, and indeed we have no other choice in the matter. I think the popularization of this viewpoint owes much to Richard Dawkins.
The religious person may be tempted to call these things symptoms of atheism, but I think this is a misunderstanding. There have been atheists for quite a long time, and most of them did in fact have some idea of a telos going beyond the biological: reason is a tool for discovering truths, morality is about transcending animal nature to become beings who are kind, loving, and so on. This is visible in the philosophies of midcentury atheists like Carl Sagan or Roger Penrose. The scientists in a previous era (although I think this was gradually lost in a process spanning much of the 19th and 20th centuries) certainly seem to have thought, on the whole, that there was a purpose to things and that it was possible to be good or bad (at least in principle) in an objective way.
But this goes back quite a bit further. Socrates was at least accused of atheism, although in Plato’s dialogs he doesn’t seem very interested in the question of gods, considering it not to be all that relevant to his inquiries. Although Aristotle believed in a God in a very abstract way, this was likely not recognizable to people at the time as theism: The difference between Aristotle’s views on religion and a contemporary atheist would likely have been rather subtle. This applies even to Spinoza, who eschewed “good and evil” as merely good and bad for the beings called human. But this presupposes that there is such a thing as good and bad for humans, and one should follow the good! I understand him (to the extent I do) as basically a stoic: humans should apply their reason to detach from their immediate interest and attain a state of moral enlightenment that is no longer tied to his passive emotions. That is, even godless (or pantheist) Spinoza understands human beings as having a moral telos, guided by reason.
The anti-humanism seen in Peter Watts’s philosophy is thus dependent on assumptions which come from the rejection of any sort of telos for human beings. Into the vacuum opened up by the rejection of earlier telic thinking, biological imperatives fill the void.
This position is incoherent on its own terms, but I first want to explain why I think this argument is going to become increasingly popular. The valuation of instrumental reasoning above value reasoning pays dividends in an industrial capitalist society. Learning how to make a product cheaper or how to make it sell more than your competitor will reward you in a way that refining just war theory will not (unless, perhaps, you tell a powerful person what he wants to hear about his war). The rejection of morality altogether is compatible with corporate reward in exactly the same way. The person who asks, “Will this product be good for others?” is less likely to become wealthy and influential than the person who asks, “How do I increase the profit margin, no matter how it is done?” To people living within this social system, rejecting assumptions one and two above can expose one to real material risks, and so there is every proximate reason not to do so. And so from the dismantling of other sources of value, we arrive at survival and reproduction as the teleology for life. This is again compatible with the reasoning of a corporation. The logic for success in business – that is, the survival and reproduction of the business – can seamlessly be translated onto the logic for living a human life. (And this reproduction need not be in the obvious way of having children – it could also mean reproducing one’s ideas, selling more one of one’s product. This brings oneself even more into line with corporate reasoning.)
Peter Watts takes these theorems all the way to the end. Conscious thought? Not necessary and maybe harmful. Empathy? Merely a local maximum.
His argument about conscious thought misses the mark entirely of what self-reflection can do even from an instrumental perspective. Without the ability to reflect on one’s own behavior and think through it awarely, it is impossible to respond to something radically new in an adaptive way. Machine learning has not actually solved this problem, and may not be able to until and unless it is capable of self-reflection. Algorithms like neural networks are extremely useful and impressive. However, they are also notoriously domain-specific. You train it up (over an enormous amount of data) to function at a task, and it does very well. In order to move it to another task, you have to train a new neural network, again over huge quantities of data. This is better than humans only if we restrict ourselves to the task of correlating large amounts of data and looking in the future only at other similar data. It is possible that some breakthrough in AI will create a truly adaptable algorithm in a way that RNNs currently are not, but for now it looks like this kind of adaptability is exactly the case in which self-awareness is useful. It’s why you do things automatically you have learned well, but have to work to adapt yourself to a new skill. We use metaphors. Translating a skill from one area to another is better and less data-hungry if you have some ability to reflect on it.
Watts also misses, in the way instrumental reasoning tends to, how significant scientific advances are made. Paradigm-changing insights like General Relativity or the Atomic Model are not discovered through application of known methods, but through reflection on problems where known methods fail. Often there is a period where various wrong solutions take hold before being dismantled (such as the aether theory of light). But to get from one understanding of fundamental concepts to a better one has always required self-aware thinking, so that the old system is dismantled and a new one put together creatively. Implemented, straightforward and unconscious adaptive responses can only be understood as useful within a particular paradigm: They can’t break out of it and restructure one’s understanding of the world into a more complete model.
It would be foolish to fully criticize Watts from within his own paradigm. These beliefs aren’t just wrong because they are unuseful. They are wrong because they create a morally bad world. People from within a paradigm like his may argue there’s no such thing as morality, but they do not actually mean this. This can be seen if we simplify morality down to its most basic, first steps, just looking at people. There are things which are good for human beings (health, shelter, peace of mind), and things which promote what is good for human beings (peaceful and mutually supportive communities, access to resources and knowledge for self-preservation). There are on the other hand things which are bad for human beings (disease, exposure, insanity), and things which promote what is bad for them (isolation, ignorance, poverty). To promote the good is to behave morally; to promote the bad is to behave immorally. There’s quite a lot of disagreement about how that is done, what the things are that are truly good or bad for people, but many of the basics are widely agreed upon.
Now Watts (or anyone else) may say, “But I am not interested in promoting what is good for people. Why should we care about this one little group of beings among the cosmos?” And that’s fine; one can say that. But that just amounts to a rejection of morality, not dissolving the existence of such a thing. For even after fully accepting this rejection, there still remains such a thing as what is good for people and what is bad. These people have simply decided that they don’t care. There is nothing particularly profound or insightful here. It is just a rejection of human goodness. It is simply misanthropic.
As I said, I think people will begin to argue this kind of view more openly in the future. This will likely start with elites (who adopt such reasoning for the purposes of maximizing value at their company) and trickle down. It’s worth recognizing it, and formulating reasons why it is incorrect. The rejection of all reason except instrumental reason is not even good on its own terms. The rejection of morality is not the erasure of such a thing, only a stated disinterest in it. The replacement of human empathy and morality with biological imperatives removes the only thing about us that makes life worthwhile.