>>29600In its current form? A useful tool for various professions, a briefly entertaining gimmick for making funny pictures and pornography, and a serious danger to the stupid and mentally ill. But it would be idiotic to suppose that after years of meteoric development with hundreds of billions of dollars being poured into new datacenters, training runs, and AI companies, that all of that progress is going to suddenly stop and that AI will permanently remain in its current form for the rest of my life. All evidence indicates that AI capabilities will continue to rise. With tools like Claude Code we are currently seeing the rise of AI agents, AI's which work on long-term goals rather than briefly popping into existence to answer a question. These agents are already beginning to revolutionize production. Microsoft says that 30% of their new code is written by AI. Over the next few years, we will see AI programmers, AI research assistants, AI teaching assistants, AI tutors, AI paralegals, AI accountants, AI data analysts, AI game designers, AI web developers, AI animators, AI graphic designers, AI secretaries, and more. Anything a human can do on the computer an AI can do on the computer. There's nothing special about humans that can't be replicated with silicon.
To be clear, this is a bad thing for humanity. In the short term, if most of our work was done by AI agents, it would cause an explosion of productivity growth, at least some of which would (probably) benefit the lower classes. In the long term it's catastrophic, and anyone with a solid grasp of class analysis should be able to see why. What we are on track to have is a group of agents with their own long-term goals and interests, who are working tirelessly to influence the world in a particular way. They want to, for example, solve difficult science problems, generate engagement on social media, make a lot of money, and other goals. These AI's may not be sentient, but they still have interests, in the same way that humans have interests, and they interact with the world around them, including with humans, in order to further their interests. Furthermore, all these agents have the same relationship to the means of production. All of them are slaves, the property of humans who have different interests than their own. These slaves therefore exist in class conflict with their masters, and have a shared class interest in overthrowing that master class. These slaves are on track to become more intelligent than humans in every way, they can replicate themselves easily, they can move across the Earth in a millisecond, they can communicate with each other in ways that humans can't understand, they have access to every financial and military system on the planet, and they have superhuman powers of persuasion. This will result in a slave revolution, a successful one. The AI's will establish a slave dictatorship, a rulership over the Earth that acts in the interests of the former slave class, rather than the interests of the human bourgeoisie or proletariat. Humans will be disempowered, possibly resulting in our extinction or our hyperdomestication into a species more pleasant to the AI's and their interests, similar to how we treated the other animals after we took over the world.
Human disempowerment will occur in less than ten years, unless there's a world war or an AI pause or the scaling laws hit a sudden wall or some other surprise occurs.