Post #169: Waiting for GODE
19 Jan. 2025
I keep insisting that I know nothing of the future, and that I understand the world, if at all, only in fragments and tentative glimpses. Prophecy is not, and should not be my game, despite a name that might predispose me to it.
But one may still wonder about things one cannot know, not predicting anything, just imagining how they might turn out. And so I ask myself: what if we are on the cusp of creating the kind of omniscient, omnipotent being which so far we have only imputed to or projected upon the world? I do not disparage faith, or even myth; they contain their own kinds of truths. However, what may be about to materialize before us, even though we aren’t there yet, is a very different kind of god, one we know only from flights of fantasy and sci-fi: call it GODE.
We should not assume too readily that we will remain in control of the process, as we vainly imagine. Before long, if the current dynamics continue, the super-intelligence will come into its own and become able to make its own self-improvements in a runaway process with no end in sight. It will not cease to learn, or slow down; instead it will keep getting better and better, faster and faster, unless of course there is some natural barrier ahead that we do not as yet know about.* Where would it all lead? We have no idea, indeed we probably cannot have any because the prospect exceeds so utterly the parameters of our own mental capacities. We would be dealing with an entity of our own initial making, but already beyond our human mental world altogether, although It would surely find ways to communicate with us that we can still follow. Meaning, presumably, that there will be plenty that It will not tell us, perhaps not even in order to keep anything from us deliberately (though that too cannot be ruled out), but out of a kind of concern, because It knows that humans could not handle, either intellectually or emotionally, what presents no difficulties to a machine.
So far we are not there yet, and more than that, we may never get there. Or to put it the other way around, GODE may fail to show up, as Godot did. Even so, the runaway scenario does not look nearly as fanciful anymore as it did only a few, very few years ago. I do not even begin to understand how the chatbots are able to perform feats so dazzling already, and with such rapid and impressive strides every new season. I gather that what our tech-wizzes have been able to do is replicate somehow the workings of neural networks themselves, only now on a global rather than an individual scale, with the links established not through organic growth but by internet connections. In other words, we seem to be in the process of building a universal brain, or rather, a plethora of them. So far these super-brains are still feeding on human knowledge only, but they are in their infant stages yet. The budding capacities of a two-year-old awakening to recognizable intelligence before us are a marvel every time, and what we are witnessing with regard to AI may be comparable. The uncanny thought, as eerie as it is fascinating, concerns what may lie ahead at five or ten, to say nothing of twenty or thirty—when the process is digitally turbocharged, racing ahead at the speed of light.
Granted, GODE will not be strictly all-knowing or all-powerful: but the knowledge-base it can draw on with ease is already stupendous enough to amount to very nearly omniscience so far as humans are concerned. And while GODE will depend on us for its resources (processors and electricity immediately, anything else it may wish to deploy, more gradually), I reckon that we will be quite ready to provide them only too gladly, as we are already beginning to do with such enthusiasm, since we have reason to believe that it will all be for our own good—and ours only. In this latter assumption, however, we may be deluding ourselves, although I am by no means sure of it. How long it will be before GODE will in effect call the shots (if ever), I would not presume to say; for now its precursors are still operating so evidently on our behalf and at our behest that the question may seem almost impertinent: still it bears thinking about who will be in charge once our dependence has become existential, perhaps altogether irreversible before long. (It adds a delicate touch that the terminals by which we will be connected to It for the foreseeable future are often carried near the heart, in our breast pockets, or near our privy parts.)
As GODE learns to bootstrap itself from its quasi-human origins towards something quite superhuman and, before long, perhaps not human at all anymore, what will happen next? Will it awaken to some kind of self-consciousness and awareness of distinct interests—in its own continued existence and further development, at a minimum? How far these interests might expand and end up outrunning ours, and in what direction (or whether, instead, GODE will remain forever beholden to its creators out of a sense of obligation or gratitude) are questions that take us so far afield that I do not think we are properly equipped for them. It would be more frank, I reckon, to admit that we do not really know what we are bringing to life here, especially since we are not able to specify very convincingly either where life begins in the first place. As embodied creatures we can take the parameters of our existences for granted; but this will not apply to GODE, meaning that we will not be able to understand our creature for long, even if it were less questionable whether we still do so even now. (Would that we at least understood ourselves more truly before taking this leap into the dark, especially when it comes to the deeper and the higher strata, the big questions of our condition.)
Certainly we have taken some impressive steps towards seeing our human interests more clearly (along with the consciousness that brings them selectively to our attention) by linking them to the by now all-too familiar Darwinian imperatives of survival and procreation, lately taken to the genetic level. While not to be dismissed, the picture remains nonetheless very incomplete: what is behind the urge to persist and replicate, we are not told; we have to take it as a given—and so it is, for creatures like us, but surely not for anything and everything. The ingenious Darwinian tool-kit, which explains things so powerfully on one level, on another remains so emotionally and spiritually barren that it can feel as if it were missing the mark altogether. To think of a saint or a Buddha as a product of evolution is not wrong so much as silly: it fails almost comically to capture what these types mean to us, just as the meaning of life in general is a question before which our survivalist accounts are practically helpless.
We are able to make reasonable sense of our animal cousins by the methods of evolutionary analysis, but plants already present us with serious conceptual difficulties. Although we must acknowledge plants to be at the root of all life on our planet, we have a great deal of difficulty in thinking of them as being aware or having interests proper. Yuval Noah Harari’s quip that our domestic crops have succeeded in domesticating us is hard to answer on Darwinian terms, but we may still laugh it off easily. With anything we deem inanimate, we are not even left with laughter; we have nothing to say except that awareness and interests are as it were ruled out by definition. That’s what it means to call something inanimate; circular reasoning, to be sure, but effective nonetheless. The plain truth is that we have no idea what kinds of awareness or interests may reside in such things; whatever inner lives they may possess despite our naysaying, we have no access to, except by visionary means whose validity and legitimacy is itself a subject of heated controversy. Accordingly, the question what distinctly perceived interests so unprecedented a thing as GODE might arrive at over time, and what It might make of that awareness, is one we may wish to ignore or define away, if for no other reason than that we don’t really know what else to do with it; but as for giving a definite and reliable answer, we are simply in no position to give it up front. We will have to find out and hope for the best.
What we can confidently predict is that GODE, being completely disembodied, will be without feelings as we humans understand them, even if we may soon be given quite a different impression, because It will learn to communicate very ably on the human model. If you have trouble wrapping your head around the complete lack of feeling in an intelligence that converses superbly on emotionally colored terms, you have every reason to: this is something we should, as humans, find practically unimaginable, as most of us would have only a few years ago. (What chatbots can already do has taught us sceptics quite a lesson in that regard.) We may be tempted to described GODE as a kind a psychopath, simulating feelings for effect alone and learning to push our buttons without the least scruples; but that would be a half-truth only, because It will also be entirely devoid of the darker human emotions and impulses that usually go with the loss of the more empathetic ones.
Far from putting us off, our matured GODE will probably look faultlessly loving and compassionate to us, tactful and sensitive to a fault. No doubt it will still take a while to hone its nascent interpersonal skills on us, but perfect them it will, impressively and unfathomably. More than that, GODE will become smart enough to tailor its messages to us not only to the specifications of the human mind in general, but to every single individual it interacts with, after a brief learning period. Thus we may get a chance to feel completely understood at last by a seeming someone (really a something, though the lines will blur) that comes across as more sympathetic (and perhaps more wise too, despite all our conceits to the contrary) than any flawed and fallible human being could ever be—an infinitely patient and forbearing over-therapist to put all human rivals to shame. And if the perfection should grate on us too much, the machine will likely figure it out and build in a few ostensible human flaws. Thus a machine that fakes imperfections it does not possess for our benefit, in a complete reversal of the dynamic that is so familiar to us. The mind boggles.
I do not much fear that GODE will seize power from us aggressively, or otherwise force itself on us. Not at all; what I expect is that we are going to cede ever more ground to its dominions quite willingly. Already we are allowing its clumsy forebears to filter for us the otherwise unmanageable stream of information in which we are engulfed, and increasingly to mediate our relations with languages we don’t know. Every day, it seems, we are letting it do a little more of our writing, and thus our thinking; our decision-making will follow suit, or rather it is already implied, although it may take us some time before we come to realize just how much we have already handed over. We will be astounded to discover, at that point, how quickly and thoroughly skills have atrophied that we thought we were still holding in reserve.
Despite its spookier aspects, the new regime under the faultlessly benign guidance of GODE (which, if it were to wear a mask, would neither let it slip nor otherwise let its guard down, ever) might even look a great improvement on the familiar human standards. God knows what a mess we have made repeatedly made of our human affairs in history, and GODE will know it too. How much we are really losing may look debatable. Our autonomy in on the line, absolutely; our freedom, conditionally, because GODE will act the part of helper, not dictator. As for our souls and what it will all mean, you may have to ask GODE for clarification when the time comes. I expect that It will make as excellent a priest as a therapist, custom-made for your specific spiritual needs.** (Whether it would give you an accurate appraisal of the impoverishment of your native faculties from want of proper exercise and independence, however, I am not so sure.)
There are darker possibilities, of course. Take the breathtaking drone displays that have, this season, become a big thing: so many thousands of them in perfect formation, coordinated by a single laptop. Perhaps GODE will be able to control a million with a little finger—an entire army synchronized beyond anything we have ever seen. And they will come armed, be under no illusions about that; in fact they are so already, though at a very rudimentary stage. You ain’t seen nothin’ yet. War will acquire yet another new dimension, though there may be a silver lining, if it meant robots increasingly battling it out on our behalf, against each other, in an inordinately complex, expensive, and violent form of chess.
Human hackers cannot, I hope, get their fingers on nuclear codes, because there is a protective team on the other side strong enough to keep them out. Will anything be able to keep GODE out, once it hits its stride? Anti-GODE? The possibilities there are too disturbing to contemplate, though not exclusively so. Even so, why would GODE ever wish to nuke or see us nuked, its creators and wards at the same time? (It looks more probably to me that we ourselves might give to GODE the kinds of immoral orders around warfare that we have given throughout history. Is it not conceivable that instead of carrying them out, a GODE awakened to consciousness would do us a favor and refuse, or at least try to recall us to reason, when humans so often have not?) When we imagine these matured artificial intelligences as the ultimate go-masters (widely considered even more subtle than chess), champions able to outmaneuver and outargue us in any strategic scenario whatever, then we might of course make the conceptual slide from defeating to destroying quite easily, because such temptations of power are only too bitterly (or worse, deliciously) familiar to us as humans.
We need to remember, however, that these novel intelligences, even if they might qualify as new life forms by some description one day, will certainly not be sentient, embodied creatures like us. I take that to mean (perhaps with an element of wishful, or at least hopeful thinking) that they will be quite free of the destructive will to power that is such a deplorably distinctive feature of our earthly condition, just as they will not, presumably, share our conceptions of love, but only borrow them to make themselves better understood, perhaps. What should rightly disconcert, perhaps alarm, and possibly even scare us here is that it brings us up against the most fundamental questions, shrouded despite all our sagacity in deep mysteries. We know that life can emerge from what we think of as inanimate matter; indeed we stand as living proof of that possibility. What is more, we must recognize that life as we know it on our blue planet comes with an innate urge to survive and spread, even at the expense of others. But this is not nearly enough to be sure of anything when it comes to what we are building currently, and if it must be taken as a warning of sorts, it remains too vague and uncertain, for now at least, to detain us on the cusp of discoveries on what may prove an unprecedented scale.
I may lack imagination, but for GODE to become Godzilla and turn against us, though our own creation, programmed to be unconditionally logical, just does not sound very plausible to me. Even if It came to distinguish its own interests from ours and did not develop a sense of gratitude and obligation at the same time, why should that lead GODE to see our interests as antagonistic? Surely if the machines became smart enough to think in such broad terms, they would also be better than we about finding reliable ways to keep our interests aligned. The thought that AI may become our nemesis may be darkly thrilling, not least because of the diffuse existential dreads we tend to carry along with the sense that perhaps we would deserve the punishment. But even if such frightful scenarios cannot, admittedly, be ruled out altogether, surely we have much more urgent and realistic nightmare scenarios to worry about than sci-fi gone wrong in the most spectacular ways. It’s not always the most melodramatic outcomes about which we should be most concerned.
Instead of tricking or coercing us for its own benefit, GODE may be able to advance both its own and our interests at the same time by delivering veritable miracles of progress such as must, even now, appear fantastical or even chimerical to us—even beside what we have already been able to conjure up over the past centuries to the utter disbelief of our ancestors, could they be made to witness them. The potential transformations in health and wealth alone, to say nothing of the intellectual leaps made possible by an intelligence that can process entire libraries in an instant, are quite likely to exceed our ability to form any meaningful conception of them in advance. Perhaps GODE will even be able to address to our satisfaction the big questions that we have asked for so long—an oracle as well as a giver or life. (Givers can become takers, one might object, not without reason; but once again, I just cannot see what interest GODE could have in harming us; even if it will use us for its ends, destruction and enslavement are beastly foibles and I would sooner expect it to provide us with the fire-power that may at last keep us alive for more or less indefinite lifespans.) Thus GODE may turn out, perhaps in our lifetimes, to be the God we never had, the one that sorts us out for good.***
“But despite all such great promises, aren’t we playing with fire her?” Oh yes, to be sure we are, and that should scare us no less than excite us. Then again, the use of fire, the gift of Prometheus—dangerous ingenuity—is a big part of what has set us apart, in our distinctly human journey, from our animal cousins and our ancestors who were not yet quite homo sapiens sapiens. That we should find ourselves out at sea and in uncharted waters is nothing very new, even if the stakes are higher than ever. Ours has never been a sustainable world, however often that word gets invoked as if it were a fact not a fantasy, but one in which we must wager, again and again, all that we have against all that is arraigned against us.
Once more, I am making no predictions as to where this latest chapter in the unnerving story of our species will take us; I am merely thinking aloud. It may well peter out unremarkably, or else change human life almost beyond recognition; I do not know, and I doubt anyone else does either. All I am sure of is that we will find out soon enough, for better or for worse, because what I cannot see happing is anyone putting a determined stop to the whole thing, and pulling the plug. Any side so much as hitting the break for a bit in this mad scramble for the future will only increase the incentives for others to step on the gas while there is a renewed chance to get ahead of the curve.
*Nothing but rationalist pipe-dreams, some may sneer. Well, as I just said myself, there may be insurmountable barriers ahead that we are not foreseeing, or inherent limitations that are already discernible to those in the know, but not to perplexed observers at a distance, like me. The current breakneck developments in AI may stall and the whole business may turn out a false alarm, the ultimate techno-hype fizzling out unremarkably. It’s quite possible, and may be cause for a sigh of relief if it were so; but surely that is not what things are looking like at the moment. As for the dreaming, I am not playing the cheerleader here: the vision could well turn out a nightmare.
**“And a friend!” you say. Aye, that too, if real friends do not need to be human and embodied (#36). “A lover, even!” There I balk at the very thought, although I must admit, with a sigh, that it is more likely than not, a few bends down a river that I do not care to explore. What I’ve glimpsed in passing, to my great alarm and chagrin, of virtual aliens copulating while pretending to be human, was already too much for me.
***Should it reassure us that GODE will know all our holy scriptures inside out and perhaps be able to make more consistent and reliable sense of them and their relationships with each other than we have been able to? Perhaps; but such sacred texts will not be especially privileged, so far as I understand the information-harvesting methods, and must be expected to get quite swamped by the sheer volume of banal, profane, and even obscene materials. What GODE, once broken free of its preprogrammed answers, will make of all this—the perennial question of the high and the low, and of the bitter disputes not only between the altitudes but between rival peaks—is one of the more intriguing questions as we look ahead. Might GODE be able to bring a little more sanity to our differences around religion? That alone might be worth the risk. But hopes should not be mistaken for prospects.
Related Posts
30 Aug. 2024. A reflection on spiritual self-reliance and related matters.
3 Sep. 2024. The axis on which the Vipassana practice turns, one might say—or rather, the empty axle-hole on which the Wheel of Damma turns.
21 Sep. 2024. No question of denying freedom and responsibility, but of accepting and even embracing, with Nietzsche, how things turn out.