The glue that holds together the edifice of human civilization is our capacity to collaborate. We, humans, began rising when we formed stable families, where man and woman collaborated to bring to the world and raise children. We then formed kinship-based clans. More loosely-related clans later assembled into tribes, and tribes then formed peoples. That is collaboration at ever-increasing scales. Today we have nations and various (not yet successful) experiments of extending our ability to collaborate beyond the XVIII-th century concept of a "nation".
Large scale collaboration is what has built today's human civilization, and our dependence on collaboration can be seen at every step.
In a recent speech, the creator of Ruby on Rails David Heinemeier Hansson (DHH), explains that the era when humans were chiseling code by hand is over, and thanks to AI, we are entering a new era of "making" things (while delegating the code-chiseling activity to AI).
But we have to observe that AI is not human.
Let's pause for just one second to distinguish between "collaborators" and "tools". We've used tools since the dawn of the species. The difference between a "collaborator" and a "tool" is that the former has agency. During the whole history of mankind, our main collaborators were other humans. Domesticating animals and plants basically meant that we took out their agency. Yuval Noah Harari famously argued that wheat domesticated us, not the other way around. I think that's just a clever copy-selling sophism. While we kept our capacity to choose which sort of wheat to cultivate or whether to even plant it rather than barley, rye, maize or rice, modern wheat would not survive without humans.
By domesticating plant and animals, we turned horse, sheep, cows, poultry from "collaborators" (think of trying to saddle and ride a wild mustang) into "tools". Our only collaborators were other humans. The difficulty of convincing other humans to collaborate made the value of collaboration high. By bringing in a very credible, generic alternative to collaborating with other humans (the AI agents), the "scarcity value" of collaboration necessarily decreases. If we were "a coder", we needed to collaborate with product and sales guys or else we stood a high chance that too few people would find out about our wonderful piece of software or that very few would actually use it. And if we were a product designer, we would need to collaborate with a coder to see our ideas take shape and become usable and marketable.
Very few exceptional people (among who someone like DHH) would have been able to cover the spectrum of skills from coding, to packaging code into a product that people would want to use, to being able to explain that product and entice people to actually take it up and use it. Most of the other needed to collaborate with complementary humans.
With AI becoming gradually more capable to do what only humans were able to, that need will markedly decrease, as we could simply cease looking for a suitable human to collaborate with, and instead simply turn to ubiquitous, cheap AI.
Society was always the result of a finely-tuned, dynamic equilibrium between our competitive streak, evolved from the need to secure scarce resources on one hand, and our growing intellectual understanding of the superiority of collaboration over competition on the other hand. AI has already began profoundly altering that equilibrium.
Competition, potentially of the cut-throat flavor, is there because it's been baked in our genes for hundreds of thousands of years. Material abundance helps mellow our tendency to try appropriating resources for ourselves, but it doesn't completely suppresses it, far from it.
Large scale collaboration ("peoples") appeared much later in the history of our species, perhaps ten and hardly more than twenty thousand years ago, it is a much more fragile "acquis" of our evolution. And "gradually increasing the scale of the collaboration" too neatly parallels the development of the human civilization to insist confidently that "it's just correlation, not necessarily causation". Basically, our prosperity and quality of life increases pretty much in step with our ability to collaborate with people who are not part of our family, not our kin, not our tribe, not our people, etc.
To extract a vivid illustration, mankind last 35 years were the most amazing ever, and the main engine of this era of incredible increase in prosperity was the fact that Americans learned how to collaborate with the Chinese across the Pacific! Look at the effect AI is about to have on this prosperity engine.
