🤔 Vitalik Buterin: Humanity is probably approaching the last days of its dominance.

Ethereum co-founder posted a large thread about the future of AI technologies. In his view, the main question today is no longer whether artificial intelligence will become smarter than humans, but whether there will remain at least one domain where humans are irreplaceable.

Key takeaways:

• Throughout human history, machine capabilities have taken a sharp leap three times—industrial revolution, computers, and now LLMs (large language models). Each time, it seemed like complete superiority over humans was just around the corner.

• Right now, many believe that modern language models with small improvements will be able to replace people in almost everything. This scenario is quite plausible, but no one knows whether it will actually come true.

• AGI (general artificial intelligence) will arrive when AI technologies can independently sustain the development of civilization, even if people suddenly disappear. This—rather than the ability to do “90% of a human’s work”—is what Vitalik Buterin considers the real point of no return.

• The most likely intermediate stage is not a “human vs AI” standoff, but an alliance of “human + AI,” where people make strategic decisions and machines handle the bulk of the work.

• The best future scenario is a world where people can enhance themselves with technology, AI remains open and competitive, and the development of artificial intelligence proceeds gradually rather than at maximum speed.

• Vitalik Buterin also backed ideas for slowing down the development of AI technologies, believing that humanity needs to retain the ability to “hit the brakes” in time if the technologies start advancing faster than society can adapt to them.