@aryehazan I don't think human are this narrowly self–interested. Soviet Union, despite its many flaws in other aspects, do have biomedical research that is not driven by profit.
I think people are drawing the wrong conclusion. Because the core capability people are likely difficult to find and perhaps straight-up hidden, people may be missing what's actually going on. Obviously, they need ~0 beginners to join the core capability team at the current stage (unless you're Terence?).
@tyler_m_john I really, like really, think people are underindexing the human aspect of the socio-economic change that will be brought about by AI. For the first time, it will be true in literal and mundane sense that capitalists don't need workers while workers also don't need capitalist
The most obvious problem is that you simply assumed unequal ownership of natural resources from the very beginning. Obviously, for people who think that human equality demands equality, whenever inequality arises at any intermediate step, whether through unequal acquisition of natural resources or through other factors, they will not agree that it is good
@JonahDispatch You know it's comically evil that someone proudly say that even if he could save a lot of lives he wouldn't do it because it wouldn't make him richer than others?
@Geiger_Capital The same sort of bizarre reasoning could be used to justify any evil.
An Aztec moral reformer says, "We should abolish cannibalism and human sacrifice."
"But do you know that human flesh makes up only 5% of our total nutritional intake?"
I quite literally don't understand this line of reasoning. GPT-5.4/5.5 Pro had been out there for several months. You could simply throw arbitrarily hard graduate-level math/science problems at it, so what bubble could there possibly be? Are these all overseas contractors typing really fast?
My point is You could basically substitute all the keywords that people use to defend wealth inequality with arguments defending slavery, hereditary aristocracy etc. "Without Aztec nobles maintaining social order, everyone—including the victims of human sacrifice—would be worse off." And with the advent of extremely powerful AI systems, what else could we look forward to besides spectacular cruelty imposed on the poor? I truly hope we do not live in such a horrible world.
@teortaxesTex I think people are definitely underestimating the current state of affairs. There is no need to imagine anything speculative, just imagine the combination of a Mythos-level base model with the reasoning post-training used in 5.2–5.5 Pro.