The meaning of life as the cherry on top of neural network glitches

In the comments of my previous post, I was asked to tell more about the speech given by the esteemed Atomic Cherry during a Sunday conversation with Kirill Grebnev at the Educational Resort.

Kirill brought up the question of the meaning of life. For me, the concept of meaning is by definition retrospective: the past is always what is given meaning, through reflection on what was planned, what was achieved, by what means, and, accordingly, whether there was meaning and what it was. Therefore, talking about meaning in the future can only be done hypothetically. For instance, practice shows that doing certain things is meaningful, therefore in the current situation it makes sense to apply a certain tactic. Whether there will actually be meaning in its application will be known in fact. The same applies to the meaning of life. Only the already lived segment of it is given meaning. Boring, instrumental, nothing to argue about.

As for Cherry, he didn’t interfere in the conversation for a long time; then, after a powerful puff of his pipe, he broke into a half-hour lecture, delivered at a very brisk pace so as not to be interrupted. I reproduce this to the extent that I managed to grasp.

Many animals are capable of advanced group interaction. Monkeys possess a highly developed intelligence that allows them to use tools, diversionary maneuvers, attack from ambushes, and enter into alliances. The genus Homo is capable of the same. However, all these wonders of social behavior hit the limit of Dunbar’s number. For larger collectives, the brain’s neural network resources are no longer sufficient, and emotional attachment weakens.

Cro-Magnon solved this problem through emotional attachment to fetishes and the construction of conceptual subjects, including collective ones (see my piece on spirits in the Libertarian Theory of War). Now, for him, “his own” are not only those with whom he has spent his whole life hanging out and hitting strangers over the head, but also anyone who is a descendant of the same totem animal, who believes in the same gods, and cracks an egg from the same end. It was precisely this ability to abstract attachments that allowed humans to assemble alliances of tribes, thanks to which they wiped out the Neanderthals completely. And after that came states, world religions, conspiracy theories, and other neural trash.

The thing is, a human cannot arbitrarily switch off this ability to abstract. He looks for patterns and finds them, always. Every neural network hallucinates. We see this most clearly in the example of large language models, but we ourselves are no better. Our consciousness, with all its searches for a higher meaning, is exactly those neural network hallucinations. If you want abstract thinking, accept the abstract glitches.

Abstract thinking has a practical value — it serves the biological need of humans to assert themselves by dominating large collectives. If we remove the ability to abstract, we simply return humans to the level of interaction of primitive tribes. And if we remove the drive to assert oneself over others, we get a pure hallucinating neural network, devoid of practical application, and therefore bogged down in fruitless spiritual quests.

Yes, the need to assert oneself over other people can be sublimated through creativity, asserting oneself over the forces of nature, and similar challenges. But without respect and recognition from other people, these activities lose their meaning for the human.

Leave a Reply