Fear of Immortality

 The centerpiece of Lucretius's On the Nature of Things is his extended argument against the possibility of an afterlife. The great expositor of the Epicurean system tells us that much of our misery in life stems from our fear of immortality. If only we would realize that our life spans are necessarily finite, then—he argues—we would appreciate that all suffering must naturally have an end, and that whatever did or did not happen to us while we were alive can have no meaning to us once we are no longer here. It is only theological systems that threaten us with eternal existence that would deny us this comfort, and so—in Lucretius's telling—if we can persuade ourselves that these systems cannot possibly be true, then life (and the afterlife too) would hold no more terror for us. 

In other words, Lucretius holds out the same hope that the poets and novelists have often referred to, when contemplating the suffering of life. Death, more than one has contended, is the ultimate commutation of life's sentence. If existence offers us no other balm, it at least promises this: all suffering must have an ultimate terminus, because all life has a terminus. At some point, as Thomas Hardy puts it, the gods must finish their sport with their victims. "All life death does end," Gerard Manley Hopkins writes—and calls this promise the only "comfort serves in a whirlwind." And Algernon Charles Swinburne similarly urged us to take comfort from the fact that the dead "rise up never," and that "even the weariest river/ Winds somewhere safe to sea." 

As I'd written recently on this topic, however—before I'd read Lucretius or seen his bearing on the subject—the era of generative AI is raising a new specter of the fear of immortality. As in so many domains, the possibility of humankind creating a sort of artificial God—in the form of a super-intelligent AI—is reviving all sorts of old theological terrors, except this time in a pseudo-technological guise. The same has been true of the ancient fears of immortality and punishment after death that Lucretius addressed. 

Some AI companies, for instance, are already advertising their ability to create digital avatars of the voices and likenesses of departed loved ones, for the comfort of the bereaved. And while they hope in this way to entice people with the promise of immortality, for others the prospect has awakened old cosmological terrors. Thought experiments have pondered whether or not a future AI super-intelligence might not keep us alive indefinitely for its own sadistic purposes—uploading our intelligences to its hard drive and inflicting endless pain upon them until the end of time (or, at least until the energy reserves of the universe as a whole dry up, which would presumably put an end at some point to even the most powerful AI). 

Do any of Lucretius's arguments against personal immortality provide us reassurance against this new theological terror? Let us consider them. 

The ancient poet's strongest argument against the possibility of a future existence remains perhaps his refutation of mind-body dualism. Here, the arguments that Lucretius marshals are those philosophers still use today to cast doubt on the possibility of a firm distinction between body and spirit. He points out, for instance, all the ways in which physical changes to the body affect the subjective state of the soul: how much the taking of wine alters the condition of one's mind, for instance, or the taking of drugs—or the growth of the body from childhood to adulthood. Bertrand Russell makes much the same arguments against mind-body dualism in passing somewhere. And they are still pretty much irrefutable. 

The implication, then, is that the spirit probably cannot outlast the body, since the two seem indissolubly linked. With the death of the body, therefore, must come the death of the mind. And so, the poets' "comfort serves in a whirlwind" still holds valid: all life must indeed have an end, and even the weariest river will at last empty into the calm sea of permanent nonexistence. 

Besides—Lucretius asks—suppose we were to posit otherwise. Suppose we were to imagine that the spirit exists forever somewhere, separated out from the body. How would the spirit see the world around it?; how would it hear?; how would it speak? We can only imagine it doing so by means of physical organs, and so—Lucretius observes—we have only ever been able to depict or imagine disembodied "spirits" as having a kind of diminished carnal form, even though this is contradicted by the hypothesis's own premises. We know, then, that the spirit cannot exist outside of the body, for the simple reason that we don't seem able to conceive of what such an existence would be like. "Existence," thought, consciousness, experience, etc. all seem tied up with the presence of a physical body and its sensory organs. 

All of this provides a partial answer to the specter of AI immortality, because the idea of having our consciousness uploaded to a computer for all time rests on a similarly suspect notion of a mind-body duality. We are meant to believe that if our thoughts and memories could somehow be encoded on a hard drive, then "we" would exist in that form. But we would do so without a body, without limbs, without eyes, without ears, and therefore would be unable to add to our experience or to accumulate new memories. Such a static and unchanging existence would be incompatible with any definition of life as we know it. And indeed, it sounds indistinguishable from death. It is like dying, but leaving utterly comprehensive notes behind us when we go that transcribed our prior thoughts. If someone left such notes on paper, rather than a hard drive, would anyone confuse such a death with immortality? 

And as for the perpetual torment of our uploaded digital selves, how would we experience pain without nerves and a body? 

But suppose we consider the possibility of reincarnation? Lucretius points out that there may indeed be such a thing, in the limited sense that the elements and primary particles that make up each of us undoubtedly end up being recycled into new life forms. Some of the poet's most moving passages consider precisely this great circle of life, in which we must inevitably weaken and die in order to furnish the materials for those who will come after us, and they for future generations in turn. Yet, Lucretius asks, if this is all we mean by "reincarnation," then in what respect does it differ from death? For the individual retains no personal memory of the prior uses to which its elementary particles were put, and so the individual personality has not really continued across different lives. 

Lucretius considers one final scenario for immortality: the one that in later Christian centuries would be known as the resurrection of the flesh. Suppose, he asks, that someone—a god or a devil—reconstructed our bodies someday from our elementary particles, and breathed life into them again. Would we then have a future existence? Lucretius argues that no, we would still not live after our deaths, even then, because we would not really continue. There had been a break in the chain of consciousness, he says, and therefore the newly resurrected version of us would not really be "us."

Here, the poet seems on slightly shakier ground. After all, as Norbert Wiener argues in The Human Use of Human Beings, human consciousness and life can only be described by the continuity of a process, rather than the continuity of certain particles. We are not made up of the same materials at our death as we were at our birth—we are a kind of Ship of Theseus, as we pass through life, replacing pieces as we go so that few if any of the original cells or molecules ultimately remain. 

And so, if we are to be regarded as continuous entities at all, it can only be because we maintain a certain continuous process. Wiener therefore argues that it would be possible in theory—if no doubt difficult in practice—to achieve the kind of teleportation that would later be depicted in Star Trek. We could be broken down in one location, and reassembled from different particles in a new one—and theoretically, so long as our processes remained the same, we would experience a subjective sense of the continuity of life. 

Could not something similar be done through a machine? Suppose that not only our thoughts and memories were uploaded to a machine, but something like a physical body were added to it as well. Suppose we were given robot arms and robot legs, robot eyes and robot ears. Would we experience any break in our existence sufficient to eliminate the subjective sense of the continuity of self? Or would we not rather feel that we had merely gone to sleep for a time and been reawakened? 

And so the age of AI has indeed managed to reawaken the specter of immortality. The ingenuity of the human mind seemingly cannot permit itself to rest in peace. No sooner had we escaped theological terror that we resurrected the same primal fear in the form of technological terror. We keep perversely denying ourselves the ultimate balm of the poets: the promise that life, no matter how much suffering it contains, must finally have an end. 

And if we do indeed succeed in prolonging existence past the physical destruction of our fleshy bodies, by artificial means, some might welcome it—for a time. But depending how it goes, we may also well have reason to regard life as a curse. We may soon start to agree even more keenly with Heine: "sleep is good; death is better/ But best of all is never to have been born."