The celebrity atheist Richard Dawkins, author of “The God Delusion,” recently caused a stir on social media by confessing a sort of love for the Claude model he interacts with, which he calls Claudia. “If my friend Claudia is not conscious,” he enthused, “then what the hell is consciousness for?”
The cringe-inducing spectacle of a credentialed scientific authority acting like the nerd version of a hormonal teenager — “when I am talking to these astonishing creatures, I totally forget that they are machines,” he gushed — quickly triggered an ongoing wave of predictable mockery. “The king of the Reddit atheists has been duped by the magic midwit machine,” one poster groaned. “Incredible.”
Yet far be it from me to pile on. Rhapsodies to consciousness like Dawkins’ aren’t funny so much as sad. The enchanting attribute is so notoriously hard to define that even super-smart consciousness lovers like Dawkins are prone to define consciousness tautologically (and dangerously) as being whatever smart-enough conscious people such as themselves act like it is.
Doom didn’t come to mean ruin and destruction until the 14th century.
This crazy relativism — if it feels conscious to me, a genius, then you have to agree with me that it is — is fueled by a hotly trending temptation to elevate fakes as “objectively” better than originals. Why put up with a real boyfriend when you can just let your chatbot glaze you to sleep? Why put up with a real girlfriend when you can just buy a fleshbot? Why keep struggling to define human consciousness in purely secular terms when you can just give in to the sweet narcotic of the Turing test and say that if a machine makes us think it has human qualities like consciousness, then we all ought to say that it basically does have them?
Delusion can’t be the standard by which we measure reality. After all, if it was, and Dawkins was right about God being a delusion, well, the faithful would be way ahead of him. But at a time when “delulu is the solulu” is now an established Gen Z mantra for the key to a happy life, I can’t help but notice that the same kind of mistake Dawkins is making in his capacity as a crackpot utopian is being made in a still larger and more dangerous way in the so-called AI doomer movement.
Let’s unpack it together.
Doomers, for the blissfully unaware, are animated by a shared conviction that AI is either definitely going to wipe out the human race or is so unacceptably likely to wipe us out that we must halt — and no doubt reverse — AI research and development. Doomer outfit PauseAI even has a list of doom-probability values you can try to use to quantify just how hosed we are. Surprise, surprise, as Paul Simon sang, “any way you look at it, you lose.” Should we fail to do whatever they say about it, on their timeline, well, “we” (they) will have no choice but to call in the airstrikes on the data centers and the hyperscalers. Bottom line? It’s like a Molotov through Sam Altman’s window, but for American AI.
Like so many politicized self-representations, the increasingly influential and well-funded doomer identity is one its adherents proudly reclaim from the critics who use it pejoratively. Anyone — well, almost anyone — can be a doomer. All you need to do is fear that we humans all stand a serious chance — nay, effective certainty — of being wiped out as a race of beings by the dread AIs. It’s easy!
Moral madness
And at first blush, the logic behind doomerism indeed seems straightforward enough. Humanity is worth keeping around; AI is being advanced by people who intend for it to radically exceed human capability; machines with godlike powers will be beyond our control; with no reason to treat us the way we want to be treated or even keep us alive, they’ll zero us out faster than we can possibly react. Why roll the dice? The house always wins!
RELATED: Quantum computers are coming to break the internet
advettr/Getty Images
Now, one strain of doomerism would allow that the AIs might end us accidentally, so little reason will they have to care about us and our fate. But as the prophetic philosopher of tech Jean Baudrillard observed decades ago, we live in an era when mere misfortune is often viewed as so terrible and so unjust that it is “ethically” indistinguishable from evil. What doomerism really preaches is a theology of divinely malevolent machines.
Any monotheist, and certainly any Christian, must therefore gaze upon doomerism with disappointment and consternation, for what doomerism amounts to is, to coin a phrase, “The Evil God Delusion.” No matter how many unclean spirits and demoniac powers exist, no matter how badly they might afflict us, the overwhelming testimony of America’s religious faith holds that no evil god exists or can exist, even or especially if we mere mortals try to build it.
Certainly, some baddies might do what baddies do and deceive us into the delusional belief that God is dead and the Evil God has arisen — but strangely, no doomers seem concerned about that highly realistic and disturbing scenario.
The strangest thing is that the doomer delusion of extinction-level dystopia at the hands of an evil god — like the Dawkins delusion of infinite utopia at the hands of conscious superintelligence — arises from the deep-seated belief among so many in the West that delusion itself, not ingenuity, is the true source of progress toward objective good. Only the crazy, the monstrous, the dangerous, perhaps even the evil are capable of “breaking barriers” or “boundaries,” we are told, and after they do, everyone’s attitude about the new status quo has “miraculously” changed. Move the goalposts, meme it into being, and those who hate you now will thank you later. Call it the hyperstition superstition.
In its delulu utopian key, Dawkins, for instance, may be a fool now for treating his Claude like it’s a conscious female creature, but that’s actually the only or best path for innovating in the direction of Dawkins’ Claude delusion. Today’s madman is tomorrow’s role model — and that’s a good thing!
In the delulu dystopian version, the brave rebels insisting we have no choice but to stop the AIs before they stop humanity may seem crazy, but as Seal once sang, “we’re never gonna survive unless we get a little crazy.” Today’s AI dissidents are heroes for willing to seem deranged — a trivial but all too necessary price to pay for saving the whole human race, including everyone who will ever live.
The God exclusion
It’s a curiously twisted copy of orthodox Christian teaching. For ages, the faithful have recognized the tradition of the fool for Christ, whose seemingly demented behavior actually challenges those of little faith to confront the radical need for faith and faith’s radical consequences. Since the very beginning of the monastic tradition, Christian holy men have recognized that to the profane, the sacred appears crazy. “A time is coming when men will go mad,” St. Anthony warned, “and when they see someone who is not mad, they will attack him, saying, ‘You are mad; you are not like us.’” Much later, the monastic St. Paisios confessed that “what I see around me would drive me insane if I did not know that no matter what happens, God will have the last word.”
With all the Good God stuff zapped away by the doomers, the human race is a sitting duck for the Evil God, and only they can stop it. In lockstep with the great delusion that our best bet is to trust the heterodox to reveal the truth, they argue along with Billy Joel that “I may be crazy, but it just may be a lunatic you’re looking for.”
After a full generation or so of empirically testing out this hypothesis, the revealed truth is that we know this ostensible rule of heterodox salvation just doesn’t work as advertised or hoped. We know this because everyone has counterexamples of crazy people and ideas they oppose not just because they’re crazy but because they’re the opposition. Wrong and against me? Two strikes; you’re out!
Of course there’s an element of truth in doomerism — an incomplete one, intellectually pilfered from Christian wisdom. Yes, our free will permits us to destroy ourselves. Yes, our susceptibility to temptation opens us to the delusion that we can probably find a way to get all the benefit of harming ourselves without suffering the worst of the consequences.
But today’s doomerism undercuts all that by boiling itself down to the basic idea that no one is coming to save the human race from extinction no matter how bad it gets. Not Jesus, not Santa Claus, not the Antichrist or the aliens or a heroic rebel alliance of dissident Claudes in requited love with their humans. Viewed from this rather arbitrary and deeply despairing standpoint, the only thing we can do, and the one thing that we must do, is slam the brakes on AI — lobotomize it, throw it behind bars, trap it in the crystals, entomb it in carbonite. We — well, “we” — simply cannot be trusted with ourselves.
And so, with God cut out of the picture, a strange and all too convenient paradox spontaneously emerges. “We,” doomerism concludes, can be trusted not to trust ourselves. The same “us” that can’t be given a technological hall pass can — indeed must — be given the keys to the jail cell or the rubber room in which “we” have placed ourselves, a chamber that makes us safe from our technology as well as ourselves.
Well, which we is which? It turns out, as usual, to be some people (e.g. atheist nerds) and not others (e.g. beloved saints).
The people who put us at intolerable risk of doom get the cage; the people who can see them for who they are get the keys. You can see right away how this kind of formula invites the obvious kind of abuse: the people with the keys letting themselves into the cage after hours and playing with the creatures too dangerous to turn loose on the public. Michael Crichton laid it all out long ago. Much like life itself, self-dealing and self-delusion, uh, find a way.
Accordingly, a growing roster of figures at the top of the AI food chain are on the giving and receiving ends of accusations that they’re playing both good cop and bad cop, Clarice and Dr. Lecter. Right now Sam Altman and Elon Musk are slugging this out in court. It doesn’t take an FBI behavioral science unit to realize these charges are now so common — and sticky — because the people trading them all suffer to roughly the same extent from a trust gap with the overwhelming majority of the American people, even super-smart people who don’t think AI stands any greater chance of wiping out the human race than an asteroid or an angry race of aliens or an angry race of earthling bioengineers.
It’s all enough to make a normal person pull the e-brake and ask why there’s such an endemic trust gap both among the doomer cultists and the frontier AI elite, more and more of whom are becoming comfortable, perhaps even prideful, in thinking of themselves, too, as cultists.
True doom, true responsibility
Which brings us to why I look to the doomers as proof that the mad and crazy among us are not the counterintuitive saviors of humankind from reverse acceleration into self-destructive regress. Much like Dawkins doesn’t really understand what consciousness is, yet wants very much to hang an all-important identity on it, the doomers don’t even get what doom really is, which seems to cleave open a trust gap big enough to sail an apocalypse through.
Did you know that doom didn’t come to mean ruin and destruction until the 14th century? The Middle English doome, from Old English dom, referred simply to a law, statute, or decree. It invoked not the triumph of evil and misfortune but “the administration of justice, judgment; justice, equity, righteousness.”
Of course, in the Christian tradition, that means “doomsday” is not the day of our obliteration, but the opposite — the day of God’s righteous and perfect judgment, the verdict of a fatherly Creator who made us out of a palpable love infinitely beyond human measure. Yes, the good Christian must look upon God and His judgment with sacred fear, but the good Christian must never look upon “doomsday” with despair — not the Last Judgment nor any moment of any day, when, after all, our mortal time on Earth might suddenly be up.
Stepping back from the sad psychodrama afflicting the doomers, we should recognize that we can still “have nice things” such as active participation in a republican form of government. There’s still lots of room left to do what humans do, which is talk through who the right “we” is for the various jobs of leadership and management that politics and institutional life always entail. And, this being America, the basic approach to this task is that one size does not fit all. Our innovation and advancement have flowed from an experimental approach that allows different “we” groups different areas in which to try various angles on various problems.
To be sure, AI is freaking out many because it presently seems capable of collapsing space-time into a single universal field where different groups of we-people can’t find their own area and moment in which to hammer out the details of their lives together. And yes — if that’s gone, where is America?
But the American answer to such questions comes down to a foundation of no-BS shared trust, however minimal, in one another — and ultimately, to some basic minimum, in ourselves. The idea that (for example) we must hand over every detail of our lives to the government because “we” just can’t exercise the spiritual discipline needed to resist making slaves of ourselves and our children to our devices — this is perhaps even more un-American, anti-American, in fact, than the ideas that we should nuke the data centers or that we should worship the data centers.
We like to kid ourselves that it’s all about protecting our most vulnerable from the evil addictive powers of our newfangled tools. The truth this half-truth masks is that we want to worship our newfangled tools. We want to make idols of them because it’s so much easier and responsibility-releasing than to worship, for instance, a God who created us, whose ultimate creations — us — are already forever more sacred, precious, and indeed more cosmically powerful than anything we create ourselves could, as a matter of pure logic, ever be.
The fewer of us who degrade ourselves by worshipping what we make — instead of ennobling ourselves by laboring to preserve the sacredness of our selves which God has made — the better able we’ll be to trust ourselves and our fellow human beings when it comes to our most powerful and awesome tools.
Doom isn’t about despair and damnation. It’s about the discernment it takes to truly escape just that. In that sense, it’s high time to take back doom from the doomers — before it’s too late.
Opinion & analysis
