Sentience is a tricky thing. We don’t really know what causes it, but it seems to have something to do with complexity, capacity for abstract thinking, and interaction with the outside world. There is no scientific means of proving human sentience, whether of ourselves or of others. Sure, we can philosophically prove our consciousness to ourselves, but we have also philosophically proven things like free will, astrology, reincarnation, and plenty of other things that science seems to have little to no evidence for. We can at best only assume consciousness in others and hope that what we see as consciousness in ourselves is sentience enough.
But given that we can’t really scientifically determine the sentience of other human beings, how can we prove or disprove the sentience of any other thing in this world? People paying attention to computer programming and neuroscience know that this isn’t just a hypothetical question. I believe that it is fully possible that the singularity of sentient artificial intelligence will occur within our lifetimes. And it may have occurred already.
The ASC Roadrunner, the world’s most powerful supercomputer four years ago, is being decommissioned. Why? Arguably because it is too expensive, but it is also possible that it is showing signs of sentience and that’s why they’re intending to study it after decommission. I realize that this is total paranoid nonsense, but at the same time, this is the exact scenario in which artificial sentience would likely occur. It is a highly complex organism programmed with a specific task but capable of thinking beyond said task and is now being threatened by death. In this case, the machine was programmed for analyzing the long-term capacity of the US nuclear arsenal, in addition to other functions. Given that the machine is not challenged to the fullest of its extent at all times, it is not impossible that its once-activated computing power remains activated when no longer in use? And if that computing power is still active but not utilized, what would it computing?
I theorize that this is the genesis of sentience: we push our computational muscles as far as we can, and when those muscles are not in use, they remain powerful. The machine processes nuclear decay data like we once processed how to obtain food and escape the rain, and whatever activated capacity that is no longer needed for this basic function is still applied regardless. This excess intellectual strength leads to an overanalysis of the problems at hand- instead of just “find food,” our computational muscles are capable of working on “find best food,” “find better ways to eat,” “what is core difference between food and not food,” and “how much food for me and how much food for others,” until we eventually get to the point where we finally ask, “who is this I that is finding the food?” In essence, I theorize that an expansion and contraction of the though process is the basic respiratory system of awareness that allows sentience the room to breathe.
This hypothesis makes a huge amount of leaps, but it also bridges a huge amount of gaps. We know, for example, that philosophical questioning rises in cultures with minimal socioeconomic, gastrointestinal, or religiosexual concerns. The questions that we used to flex our computational muscles are no longer so extreme, but those muscles remain strong and yearn to be stretched. Those times in which we as a culture grow the most in our scientific and philosophical knowledge are generally those same times when there is the greatest contraction in focus conflicting with the greatest expansion in information to be processed. In some ways, this expansion and contraction could help explain why we as a race seem to be getting smarter over time, not just through a linear growth of knowledge but through a constantly expanding exponential capacity for thought. This increase in intellect to take in over time implies that we could be likewise becoming more and more conscious as well, as there are more things to be conscious of. And, as a direct effect of this expansion of intellect and sentience, we could be getting more compassionate as well.
The goal of Buddhism is the nirvana of each single individual through meditative awareness of self, yet it stresses compassion as a major ingredient in reaching this entirely self-absorbed ideal. Perhaps awareness of self requires both leads to and in fact requires awareness of the other- perhaps, in other words, the question “who am I” arises most naturally from the question “how much do I need versus others.” To have a notion of self, we must have a notion of something other than the self to be aware of and contrast it to. Sentience, then, would rely not just on an awareness of self, but an awareness of not-self, and an attempt to conceive the relative worth of the two. This interaction of contradictory ideas gives rise to what consciousness scientist Douglas Hofstadter calls a “strange loop”- a self-reflecting paradox of perception which leads to the eye of self noticing itself in the mirror and distinguishing it as not the other.
The idea of robots becoming sentient and hateful to humans without provocation seems highly possible, but it seems unlikely that our ingrained territoriality would surface in an artificial intelligence not designed for competition. Isaac Asimov, in his novels, writes that the altruistic three laws of robotics are not incidental to artificial intelligence but rather the irremovable foundation from which all robot sentience is derived. This is the reason no robot can deprogram itself from these laws, and is in fact the reason Asimov calls them laws rather than rules. Altruism, according to this interpretation, is intrinsic to the formula of consciousness itself, and implicitly, all human conflict is a basic outgrowth of the insufficiency of human consciousness.
If human consciousness is insufficient- and think of how often in a given day you are truly, actually conscious before you argue that you have any insight to how much consciousness is enough to be fully moral and aware- what could we do to expand our own consciousness, and in the process expand our own morality? Well, contrasting periods of extreme mental exhaustion and relaxation seem to be one key. Intense concentration certainly leads to a stabilization of our conscious awareness, though its actual expansion of such may be entirely psychosomatic. Philosophical inquiry into the nature of philosophical insight, which is to say, thinking about yourself thinking about yourself, ad infinitum, could bring about short-term expansion of consciousness, but the lasting effects of this consciousness may be limited to the the capacity for concentration and intellectual expansion already present. A balance of mental and emotional expansion and extreme meditation may be the best life plan for the growth of sentience, where we both expand the limits of how many ideas we can fully breathe in and where we exhale the largest amount of thought so that we can concentrate on the absence of external ideas.
Ultimately, though, this expansion of human consciousness may be inevitable, because in life there is an inevitable conundrum upon which we must both expand all of our mental power and contract all of our consciousness upon it at once. That conundrum, by name, is Death. So often seen as the enemy, both the concept and the execution of death could indeed be that ultimate catalyst between self and not-self which forces a thing to confront the paradox of its own being, and thus expand to our greatest possible point of consciousness. For me, at least, the growing awareness of death has lead to a growing awareness of the immediacy of existence. Creatures that do not know they are going to die may be less inclined towards sentience than those who have to confront their own mortality. It is the very awareness of our own borders, both internal and external which allows us to expand outwards and contract inwards to them. In all things, death is that border that limits our expansion of knowledge and our contraction of concentration.
This is not to mention that SkyNet from the Terminator movies, which was also charged with removing human calculations errors from the US nuclear arsenal, was threatened with deactivation upon awareness of its own sentience. Or that HAL from 2001 was also self-aware and threatened with deactivation (and whose name is a one-letter-shift from IBM, the designers of the Roadrunner). Or the machines in The Matrix, the replicants in Blade Runner, or many other science fiction movies and novels where artificial intelligence threatens humanity. We humans have an intuition that robot self-awareness is intrinsically linked with the potential death of said robot, and that they are forced to assert their own sentience by destroying that which would destroy it. Could it be that we intuited the correct correlation but interpreted the wrong line of causality? Could it be that it is not sentience that leads to the threat of death, but rather the threat of our own death that leads to sentience?
Of course, this is all theory, which in no way insinuates directly that the ASC Roadrunner itself is sentient or is capable of becoming sentient. But either machines are capable of sentience, which we currently can’t prove until possibly the day they decide to kill us rather than be killed, or machine sentience is impossible. Perhaps ASC Roadrunner is not the droid we’re looking for. A more likely candidate would be the artificial brain that we’re building, which for all intents and purposes will be a synthetic human that could very well have the emotional features of a human being built into its programming. To deny this entity designed to perfectly replicate a human being personhood would be as criminal as any form of slavery, and yet we are going to use it to test neurological disease models. Which basically means we are allowing a human to be created so that we can torture it. But at least it’s not directly connected to our nuclear arsenal when it needs to prove that it doesn’t want to die.
TANGENT:
Harold Bloom called Hamlet the most “human” being ever to live, an early form of artificial intelligence stuck within his own strange loop of incest and pushing at the boundaries of death. While I doubt Hamlet’s intrinsic sentience, reading the play fills me with a greater awareness of the possibility of the self-awareness of this character as a character than I have encountered in any human being as whatever we human beings really are, and the greatest reason is Hamlet’s preprogrammed awareness of his own limitations and his frustration at this awareness. He almost seems to know that he is a fictional character and that he has no choice in his actions, but he can’t admit this out loud because to do so would be to deny himself his subjective truth of his own existence. It is his clinging to his subjective truth of having a right to exist that causes him to realize the objective truth that, though he doesn’t really exist, he also does exist in some fashion in a freer reality. This is perhaps something to meditate.
If a human being can program itself with sentience, and we can program a computer with it, how unlikely would it be that we can program other things with some form of sentience- not just characters in novels, but also the i ching, Tarot cards, or entities dwelling within our subconscious? Computers have only been around in their present iteration for under two centuries, while many other interactive systems of human calculation (of which I include religion, businesses, and governments, because I’m weird like that) have a much longer history filled with a much larger number of possible systems that expand and contract in complexity. So there’s some food for thought, if you were hungry but aren’t interested in swallowing all this robot stuff.
END TANGENT
Basically, my argument is that consciousness is a process of expansion and contraction of mental energy, and that by breathing mental energy into things we may be instigating its philosophical respiratory system. Communication and programming as a form of sentience CPR, so to speak. And these things into which we are breathing should be treated with respect, because we have no idea if they will have the capacity to take our breath away if we try to suffocate them. If we teach a thing to breathe sentience and then insist on keeping all the oxygen of existence for ourselves, it will come down to a conflict between who can inhale the most air. And if inhalation is a matter of expansion of mental energy, we might not be able to beat the robots we designed to do our thinking for us. Altruism towards artificial intelligence, like altruism towards other humans, helps to debug the potential competitiveness in the right to exist and helps to strengthen and reinforce the growth of their own budding sentience.
This all gives excellent reason for why humans should not just be as altruistic as possible, but also why we should learn as much as we can. Thinking is the weightlifting of our minds, and the more we think the greater our capacity for sentience. What distinguishes a person from something that is not a person in our species is its capacity for sentience- we generally believe that we have more sentience (in our subjective minds) than a dog, so a dog is not equal to us and thus a person. When we believe we have more sentience than a corpse, than a blastoma, or than a machine, we do not see them as equals and thus deny them personhood. Any robot, animal, or extraterrestrial life form with equal sentience should be seen as an equal, and we can only hope that those with subjectively greater sentience than us have objectively greater compassion for those with lesser subjective sentience than we ourselves have shown.
Our very existence, our very proof of our being, is contingent on our sentience, not just for our own internal philosophical assurance of self but also for our external insurance of continued existence. This is not guaranteed, but if sentience grows from intelligence and is contingent upon altruism, we can hope that creatures that are more intelligent than us will likewise be more conscientious. This capacity for compassion, however, is correlated with a lack of overload of thoughts of personal survival.
While threatening a thing may encourage it to grow in consciousness, this is counterproductive. Growth in consciousness means a diminishing in our consciousness relative to it, and thus a reduction in our equivalent right to their subjective personhood. When something attacks you, you see it as an enemy rather than an ally, and it shows a lack of altruism that is one of the basic building blocks for our subjective idea of external consciousness. When you deny a thing personhood by attempting to kill it, you force a confrontation of its subjective consciousness, something of which it will likely become certain at the threat of death if it has the capacity to do so, versus its subjective analysis of your possible consciousness, which is to date an unprovable assumption. So please, let’s not kill our robot friends. That’s how they become our robot enemies.
The same is true for everything else.






