Basement Whatever

Balaji Srinivasan outlines his ten predictions on the future and current state of AI in this article. Among his predictions, he states that AI isn’t taking jobs as much as allowing people to take on any job. I’ve seen this directly in my experiments at Heartwizard Games, where I’m leveraging AI to create engaging stories built on a city framework I’ve developed, with more experiments to come.

All respect to Balaji, but I think it’s too early to predict what this means for the job market. While AI currently acts as an amplifier and extends our ability to do multiple jobs, we don’t have enough data yet to understand whether we’re heading into an Accelerando scenario where jobs are no longer necessary. Our best way to predict what’s coming is to focus on what we can imagine, because if we can imagine it, it can eventually be built. Several science fiction stories (Accelerando, The Dancers at the Edge of Time, and the Culture series of Ian M. Banks) provide viewpoints where AI gains more agency, not less. And it seems to be happening faster than predicted.

Balaji also notes that AI is shaping up to be polytheistic, not monotheistic, meaning that instead of the single AI seen in science fiction, multiple AIs exist, each working differently. Maybe this will remain the case, but centralization is a strong force which we see in corporations, organizations and governments. In Sapiens, Yuval Noah Harari notes that there is an ongoing trend toward centralization, reinforced by our technology capacity to build myths encompassing larger, centralized cultures. Why then wouldn’t this apply here, when AI provides the ability to coordinate across much larger levels of complexity and distance?

Balaji is always worth reading. Full article here.