Of all the unsettling stories to emerge from our generative AI moment, perhaps the creepiest is the fact that people are now trying to create avatars of their deceased friends and relatives, using the new technology. The idea is that the algorithms can train on a library of video, text, and audio recordings of a person's various words and deeds (a trove of data that each of us now generates in the course of a lifetime, if we are at all plugged into technology—or members of our family are), then it can use this information to generate a plausible simulacrum that moves and sounds "like" the person who is gone, but can also respond to new prompts and stimuli, giving the illusion that you are really speaking with a lost loved one.
What spooks me about this idea is not so much the idea of a person—or a digital avatar of a person—communicating with us "from beyond the grave." Rather, I am bothered on behalf of the deceased. I don't want anybody prolonging my existence—or a simulacrum of it—without my consent. I do not in any way share this longing for digital immortality that many of our tech overlords and silicon valley bros seem to have; indeed, it seems to me a prospect fraught with an especially vertiginous sort of existential horror. Who wants to live forever as bits—a set of ones and zeroes?
A lot of people, apparently. A friend of mine (who, surprise surprise, lives in Silicon Valley) is a big believer in the promise of AI immortality. He often kids me—and he is more than half-serious when he says it—that pretty soon "we won't have to worry about" this or that minor bodily ailment or inconvenience, because "in five years we will all be uploaded anyway." As in, we will have the information contained in our organic consciousnesses, bound up though it currently is with our bodies, untethered from such physical limitations and stored on a computer instead.
I always laugh at this, in a dark way; but—in truth—who am I to say it couldn't happen? I am by no means so confident in my own philosophy of self or of consciousness that I could categorically rule it out. Part of me wants to think that a digital upload, even if it could somehow encode all the information found in a given human mind, would never actually manage to possess the subjective sense of continuity that we have with our past experiences, as perceived through our body, that is such a large part of what we mean when we speak of a having a selfhood or an identity.
Yet, I'm also aware of what I've discussed before as the Alan Turing problem in the philosophy of AI: namely, that a 100% perfect mimic of a human consciousness would, if it truly did become perfect—essentially be indistinguishable from the "genuine" article. So I hesitate to laugh too uproariously at the notion that today's AI wizards could someday upload a complete human consciousness and sentence it to an eternity of mute existence—or, if not quite an eternity, at least as long as it takes for the sun to burn out or the universe to exhaust whatever energy reserves would power the computer that read it.
Anyone who actually looks forward to such a prospect is incomprehensible to me. I admit that death is fearful, in part because we have no ability to conceive of a state of non-existence. But, I maintain, we cannot conceive of perpetual existence either, and if given the choice between these two horrifying inconceivabilities, I'd infinitely prefer the former. The fact of death, of universal mortality, certainly provokes an existential crisis. It confronts us with the specter of absurdity. But what about the absurdity of living passively for billions of years as an inanimate line of computer code?
After all, when confronted with existential dread, the knowledge that eventually we all must die (which, until AI came along, was one of the few absolute and eternal guarantees of human life) is a great relief. Gerard Manley Hopkins, in his haunting poem about depression, appealed to the knowledge that "all life death does end," and called this the only "comfort [that] serves in a whirlwind." Likewise, Swinburne took solace in rejecting the doctrine of the resurrection, proclaiming that we should "thank [...] whatever gods may be" that the dead "rise up never," and that "even the weariest river/Winds somewhere safe to sea."
Yet the immortality-chasing tech bros seem strangely to disagree. I don't understand it. The fact that humanity, having only recently rescued itself from the dread of cosmic theologies and the terror of jealous gods, would then turn around and create the very same specter of eternal existence all over again, but this time based on technology rather than religion, seems almost inconceivably perverse. Why would anyone want to prove Swinburne wrong? Who is it so enamored of mere existence, on any terms, that a passive eternity seems preferable to an active finitude?
I was reading B.S. Johnson's novel House Mother Normal last night on a long international flight, and I was thinking about all of this because, one notes that even his most grotesque creation—the novel's tyrannical "House Mother" herself—at least is checked in the end by the fact that she cannot persecute her charges after they die.
The novel, you see, depicts a sinister nursing home, where the bizarre goings-on are only gradually revealed through a series of eight 21-page monologues by the home's residents, as they each record their spontaneous responses to the same set of events.
The "House Mother" who oversees this home, we slowly discover, has managed to isolate the patients who have no living relatives (who have become, in her words, "orphans in reverse") and exploit them for petty handiwork as a side hustle, while subjecting them to various other kinds of physical and psychological abuse. The patients, we learn from their inner monologues, are mostly members of the working class, and each has faced their share of indignities already at the hands of the British class system—whether in the army or domestic service. They have faced the callous treatment of officers in the one and the sexual predations of employers in the other, for instance.
The House Mother's abuse is cast in the same mold. It is one more indignity to heap upon lives already straining under the load of other insults; another form of exploitation that dogs them even in retirement. Yet, one of the few redeeming features of their fate is that there is a limit beyond which the House Mother cannot pursue them. There is a natural barrier beyond which she cannot go: she cannot harm them when they are in their graves. This is the last and most important solace of every human life: we will all, to paraphrase Swinburne, eventually have a rest.
As even the House Mother sings, in one of Johnson's brutally dark-comic burlesques: "Death comes for us all, no matter who,/ No matter what we bloody do: [...] For this we should stand up and cheer/ And please ourselves while we are here." The House Mother can heap indignities upon the aged, but she cannot reach them when they are dead, and for this we should "stand up and cheer." Even the British class system cannot exercise even one more humiliation or snub upon the deceased—at least not any that the dead will notice. It is for this reason, perhaps, that Robert Burns called death "the poor man's dearest friend."
Yet, here come people who wish to rob the poor and suffering even of this last and dearest friend; who wish to outdo even the House Mother, that is to say, in unremitting persecution. Here come people working day and night to try to take even this last crumb of solace. Suffering there still will be; but now with no terminus. We will be deprived even of the comfort of knowing that "this, too, shall pass away" and that all pain must eventually end because, as Hopkins said, "all life death does end." This is the ultimate insult; the final humiliation to the deceased. It is for their sake that I protest.
Or, if you really must upload digital versions of people for an eternity—at least ask their permission first!
No comments:
Post a Comment