Ai Ai Ai, its armageddon

It isn’t long ago that a man wearing a cardboard sign that read “The end of the world is nigh” would be regarded as a crackpot. We might have chuckled at the prophet of doom.

These days his sign is a headline and the computer boffins who are creating a genuine alien intelligence use a concept Pdoom. The P stands for probability (not prophecy) that a super intelligent computer system will extinguish all human life.

We’ve been on the brink of calamity for some time now.

Here’s a story from 1962 about how apocalypse was narrowly avoided.

That’s 64 years ago.

The United States and Russia were arguing about each other’s proliferation of nuclear missiles. Russia didn’t like it that America had their doomsday devices set up in Turkey and pointed at Moscow. The Americans didn’t like it that Russia had set up their Armageddon machines in Cuba and had them pointed at Washington and New York and Los Angeles.

Tensions were running high. America was causing mayhem in Cuba. The Russians were making threats and had sent 4 nuclear armed submarines which were submerged in the Caribbean off the coast of that island state.

The American navy was patrolling the surface, hunting the submarines. They were dropping depth charges into the ocean creating explosions near the Russian vessels as a warning to surface.

To launch their weapons of mass incineration, these submarines needed the order of both the captain and the political officer. One of the submarines, designated B-59, had three leaders on board, the Captain, the political officer and a man called Vasily Aleksandrovich Arkhipov who was deputy commander of the fleet.

B-59 had been submerged for longer than its battery specification allowed. They had had no communication with Moscow or any command structure for more than 2 weeks.  Its air cooling system had failed and so the temperature in this submerged tin can was over 40 degrees centigrade. The hull of the ship was reverberating with the thunder of the signalling depth charges the Americans were dropping.

The captain and the pollical officer were convinced that the war had started and were prepared to launch their payload of nuclear missiles. Ordinarily this would have been enough but Vasily’s presence as senior officer meant they needed his consent.

In a dimly lit room, in hellish heat, cut off from the rest of the world, hundreds of meters underwater, with shock waves from the depth charges shaking the boat, the three men had an intense argument. The Captain lost his temper.

Can you imagine the smell of the men?

 Vasily held his ground and ordered that they surface and meet the Americans face to face.

---

Reflecting on this story, known in the west as “The Cuban Missile crisis”, it occurs to me that it might have some allegorical relevance to our current doomsday scenario.

First, this is probably the most dangerous moment in human history and nobody knew anything about it until 30 years after the event. The cold war shrouded Vasily’s moment in secrecy. No headline, no fanfare, no viral insta post.

Second, although this event and the geopolitical machinations surrounding it led to the establishment of a hotline between the two superpowers of the cold war and the mutual agreement to restrict nuclear armament, it was the moral choice of a single human being that averted catastrophe. The systems of the day, which were meant to prevent erroneous decision making, had given the green light to unprecedented destruction. One human being made a good decision based on restraint and caution. He was also willing to risk censure from state authority.

This is moral courage.

In the third instance, the story set me wondering about whether our discovery of things as powerful as nuclear fission and understanding the basic building blocks of all matter and the energy processes  that create the world, brings with it a shadow element, destruction. Does our progressive understanding of and capacity to work with the creative forces of life, bring with it the possibility of life’s annihilation?

The first thing we did with our discovery of the power of the atom was to turn it into something to hurt each other with. No, that’s incorrect, it was the need to hurt each other that fuelled the race to harness the atom’s power.

Certainly, we participated in the escalating cycle of nuclear armament that reached denoument in the sweaty twilight of B-59 in full knowledge of the apocalyptic possibilities. America had already conducted that experiment on mannequins in the Nevada desert and on people in Japan.

We knew that we risked complete annihilation, but we went ahead anyway.

---

Here’s another, much older story about our dabbling with the creative forces of the universe and the existential dread that this precipitates in us. This time a myth rather than history. It’s the story of the golem.

A golem is a humanoid creature made from clay  and animated by the performance of sacred ritual. It occurs first in ancient Judaism. A golem cannot speak because, “In the beginning was the Word, and the Word was with God, and the Word was God”.

So the word is reserved for the divine and as we know the LLMs are all about words so this lends our creation of our new techno golems a more transgressive edge and perhaps elevates our anticipation of punishment by extinction.

Anyway in one version of this myth from the 16^th century, Rabbi Judah Loew ben Bezalel, the Maharal of Prague, makes a golem from river clay to protect the ghetto against anti-semitic violence. In the most frequently occurring version of the story, Rabbi Judah forgets to deactivate the golem on the sabbath and because of this lack of observance of pious law, the golem turns on the people and destroys the ghetto.

In the end, the rabbi manages to turn the golem off by erasing one of the glyphs on his forehead. In one version he is suffocated to death by the clay that de-animates on top of him, in another he stores the clay in the attic of the  Old New Synagogue in Prague, and forbids entrance to all and sundry. Of course we’ve never been good at being forbidden. The clay isn’t actually there.

We checked.

Islam also refers to golem like creatures called jinn and here’s another cautionary tale from the ancient stories. Sulaymān (that’s Solomon in the Judeo Christian context) is granted use of a workforce of superhuman labourers to build a temple. Their control is bound to a seal ring that Solomon wears. Solomon then dies but is propped up by his staff. The jinn continue to labour despite the absence of oversight and unaware that their divine mandate to their master has ended until a termite gnaws through Solomon’s staff and he falls to the ground. Only then do the jinn realise they are free.

Until then they labour away unsupervised.

Some of the jinn are quite malicious and resentful of having been compelled into servitude and they set about slandering Solomon with falsehoods that his power was not divinely granted but was derived from wicked sorcery. Lots of people believe this and start dabbling in dark magic with a variety of mythically bad consequences.

---

It seems tampering with the forces of creation, brings with it existential dread of destruction.

Could it be that the fevered surge of anxiety laden headlines about the AI apocalypse we are unleashing is simply the manifestation of embedded Abrahamic guilt?

Or perhaps these myths and many others like them from various cultures draw on a referent baked into reality, something like “Watch out folks, don’t get ahead of yourselves”.

More likely the referent is deep knowledge of the nature of homo sapiens. It is Freud’s drives, the complementary forces of Eros and Thanatos, life and death, creation and destruction, sex and aggression. It is baked into every single one of us.

Perhaps another part of the referent is the self-knowledge stored in our mythologies, we will eat the fruit of the forbidden tree, we will open Pandoras box, we will use our sorcery to create a literal minded apprentice despite all injunctions against doing so.

We can’t help ourselves.

---

I notice these myths imply a fourth allegorical point. While it may be the golem that smashes the ghetto, or the jinn that carries on labouring without supervision and then undermines Solomon’s legacy, it is the humans in the story who are the real problem. Rabbi Judah evinces a lack of spiritual clarity and this is what unleashes the might of the golem against the people it was created to protect. Solomon dies but appears to live the jinn and it is this that lends narrative impetus to the problems that follow.

I laugh when I hear the techno billionaires talking about alignment, and the critical importance of ensuring that AI is aligned with our human goals and strivings. It seems to me there scant evidence that en masse, we are in any way capable of acting in our own best interests.

I wonder if it is less artificial intelligence that poses the greatest risk to humanity, and more the collective genuine stupidity of 7 billion crazy apes.

I have come to see the fevered twittering about the threat that AI poses as the mass projection of all of the worst of humankind onto a machine of our own making. We fear that we will lose control of it and that it will kill us all, dismantle our infrastructure and use the elements of the earth for its own ends. We fear that it will do to us what we have done to each other over and over again since before recorded history and what we have done, and continue to do to all the other life forms on the planet.

For those who don’t know what projection is, it is a psychological defense mechanism whereby  a person unconsciously takes aspects and characteristics of their own experience and behaviour that they find unacceptable and situates them in another person and then relates to those qualities as if they were attributes actually possessed by the other. We can project positive, likeable things on to other people which leads us to approach them and possibly fall in love with them, or less positive, destructive things which then leads us into conflict with or avoidance of the other person.

And this handful of techno overlords, whose companies are absorbing a significant portion of global capital (for that read our pension funds invested on our behalf by multinational finance houses), keep thrusting the misbehaviour of their latest models into the news cycle telling us how dangerous they are.

It is as if they are saying, “if you think what humanity has done with hydrogen bombs and napalm and the internal combustion engine is bad, take a look at what we’ve built. It’s going to be the nail in the coffin for us.”

Of course, these LLMs are very appropriate targets for this projection. After all, they trained on the sum total of human output and so it is quite reasonable that we will see in them, our own capacity for wanton, selfish, destructive behaviour.

The big upside of projection in psychological terms is that the ego, having been divested of the undesirable qualities that are now so convincingly embodied in the projective target, becomes convinced of its own righteousness and can set about living under the delightful illusion that it is an agent for good in the world, now engaged in a morally legitimate fight against the bad, conveniently situated in this other entity. It’s usually a person, but in this case it is Claude or one of his more advanced iterations.

This doesn’t mean that AI is not dangerous to us. It does mean though, that we are the generative partner of that danger. The point here is less “Is AI going to kill us” and more, “How much collective psychic energy are we giving to the small handful of people who own the machines and how dangerous is it for us to do that”.

Because that’s one of the basic principles involved in projection. What psychology shows us is that what is denied in us controls us. If we do not confront our own destructiveness, we will live it out in amplified form.

And it is not only our instinctual drive towards destruction we are projecting on the machines, we are also delegating all manner of ordinary human processes to them. Under the guise of helping and empowering, we have outsourced the social processes that regulate intimate partner selection (Tinder, Bumble, Hinge and so on) and perhaps even more perniciously, the ordinary process of validating ourselves, forming and norming our opinions, discussing, comparing and evaluating our opinions. All of these are no longer situated in everyday conversation but in processes and on media owned by this same minority.

A significant and well documented effect of this is that the voices that emerge from this new technologically oriented social process are explicitly guided by the attention economy. That is, a single imperative, the most advanced computers ever built at a scale never before manufactured, processing hundreds of billions of datapoints and selecting for one thing only, maximising how long people spend looking at the screen.

This has largely usurped other culturally embedded social processes that, while imperfect, messy and convoluted, tend to employ a more varied set of imperatives not least of which is the deemphasis of extreme viewpoints. It is tempting to lay responsibility for all of the current ills of the world at the door of the techno billionaires and their machines but this would be another projection. The horrors of the 20^th century significantly predate the computer and produced human fatalities that make the genocidal wars currently being waged look like small potatoes.

For the purposes of this essay though, the point is that globally, the technologies antecedent to AI already commandeer a massive amount of our personal, interpersonal and cultural energy. And they want more.

And that’s why they keep telling us how dangerous AI is. These people, who also own google and X and Facebook and Instagram are not happy with how much of our life force, our collective “chi”, or as Freud would have it, our libidinal energy, they have already co-opted, they want it all.

And they want the money and power that accompanies it. And tacit in what we are being told is, “Hey it’s not us, we’re trying to do good here. It’s this machine”. It’s not the flesh and bone and greed and our self-destructive tendencies that’s causing the problem, it’s this infernal silicone that we can’t control.

It is as if Rabbi Judah tried to convince the smouldering ruins of the ghetto that it wasn’t him who did this, it was the clay.