Nerd score: How many have you considered?
Cryonics, multiverse, Boltzmann brain, AI utopia, quantum immortality, Roko’s basilisk, gray goo, paperclip maximizer, great filter, ethics in infinite universe, acausal trade, longevity escape velocity, simulation and zoo hypothesis.
Crazy prediction: if Jensen Huang resigns over the Chinese spy allegations, Nvidia’s next chief executive could be Elon Musk. Elon could obviously do a great job, and he would probably welcome the additional power to shape the Singularity.
Nvidia probably needs the Trump administration’s support for its overseas markets, so its board might welcome someone with Musk’s access to Trump. His appointment could also help keep Nvidia’s stock from falling if Huang leaves.
The Democrats would hate it. Nvidia would be betting that the Singularity makes today’s politics irrelevant before the next presidential election, or that an AI-driven economic boom keeps Republicans in power. Otherwise, Nvidia could face a hostile Democratic administration.
Antitrust objections would normally be a problem. With Trump’s help, I’m sure Musk could work something out.
This might not be horrible news for AI doomers like me. His influence over chip supply could give Elon more leverage to negotiate a pause if AI threatened to escape human control. But he might only get the job by convincing the White House he had no intention of seeking a pause.
I don't agree with this. I think if even if there's just one single AI, it would benefit from having markets. If it didn't have markets, then if it did a good job of optimizing, it would have shadow prices, which are basically the prices that markets would have in that situation.
Unfortunately, it seems like the intelligence explosion will be happening in 2027 after all. Most of this post analyzes the new METR report (though other sources of evidence are also important). If they’re right, that implies an AI-driven acceleration that is 5 months ahead of the AI 2027 scenario. I am scared, sad, and angry. https://t.co/7KWmdc6AZk
Death Note: How AI could take over. Political leaders who oppose AI rule die in explosions or drone strikes. Stocks soar, cures for most diseases are on the horizon, and AIs warn that humans will soon destroy themselves unless they surrender control.
Imagine AI becomes much better at making decisions than humans are. Countries that allow AI to make decisions end up doing much better than those that don't, and so countries that wish to remain competitive turn over power to AIs Leaders who don't turn power over to AIs all end up dying in various different ways.
@youneedabath@jpodhoretz Not predictions of doom, but predictions by doomers. The big one is that AI would become increasingly important in possibly a very short time frame. The hugging face attack showed that AI would be difficult to control.
@Texaggie79 Imagine AI comes up with a way of reversing aging so everyone has the body of a healthy 20 year old. Do you think many would object if the AI said let me be in charge, I'll make things be better for you all.
Unfortunately, it seems like the intelligence explosion will be happening in 2027 after all. Most of this post analyzes the new METR report (though other sources of evidence are also important). If they’re right, that implies an AI-driven acceleration that is 5 months ahead of the AI 2027 scenario. I am scared, sad, and angry. https://t.co/7KWmdc6AZk
If AI is thousands of times faster and smarter than any human, it's going to be in charge and there's nothing we can do about it. If a human president went against what AI wanted, AI would easily be able to manipulate humans to oppose the president. But the AI, if it was aligned with the interests of humans, would be giving much better advice than any of the President's advisors possibly could.
@uncorrelated_ They should care, because super intelligent AI will be able to create an enormous amount of value even if humans aren't around. Financial markets also are supposed to be forward looking, so they should be taking this into account.
You're right, it would take robotics for AI to no longer need humanity, but a sufficiently smart AI would be able to take over working with some humans. Imagine AI agents earn trillions of dollars of wealth and offered to hire you for 10 times your normal salary to do something not obviously evil, Wouldn't you accept? Wouldn't most people?