I always thought AI was going to be way smarter than humans and an existential risk, and that's turning out to be true. The danger is going to be more humans using it against each other. I think most likely that'll be the danger.
Elon Musk's greatest fear is no longer science fiction. A sentient AI that builds smarter versions of itself could soon surpass us entirely, leaving humanity obsolete and powerless.
So what the big fear is that these things are going to become sentient, make better versions of themselves, and we're going to be lost. We've lost the control over the world. It's now there's a higher life form that lives amongst us.
Yeah. That we've created. How far away are we from that?
Well, in terms of silicon consciousness, I think we're trending toward having something that's smarter than any human, smarter than the smartest human by maybe next year or something. A couple years.
There's still a level beyond that which is say like smarter than all humans combined, which frankly is around 2029 or 2030 probably.
Right on time. Do you think it has a potential application for government?
One of Musk's most disturbing concerns is the rise of a hyperwoke authoritarian AI, a digital nanny state that polices thoughts, rewrites truth, and turns the future into a controlled dystopia.
I mean, one of the concerns would be like, okay, if there's a super oppressive like woke nanny AI that is omnipotent, that would be a miserable outcome. That would be terrible.
And just like if you misgender someone or something like that, you know, that would be good. I mean, it's people, it's really just like computers that are bad software and computers. This sounds kind of strange, but the reason I call myself like tech support is that a lot of it is mostly not corruption. It's mostly just waste and incompetence. It's just a big dumb machine basically, a whole series of big dumb machines.
The whole idea of creating OpenAI was my idea. I mean, I named it OpenAI as an open-source artificial intelligence. That's what it's named after. Now it is closed source and for maximum profit. So to some degree, I think reality is an irony maximizer. The most ironic outcome is the most likely, especially for the most ironic entertaining outcome is the most likely.
That fear drove him to create an alternative to Google because he saw a tech giant barreling toward AI dominance without the slightest concern for safety or human consequences.
I wanted to start something that was the opposite of Google because I was concerned about Google. Google wasn't paying enough attention to AI safety in my opinion. So I was like, what's the opposite of Google? Well, it would be a nonprofit open-source AI. And now OpenAI has turned into a closed source for maximum profit AI. How are they able to do that? That's what I'm confused about. That shouldn't be possible. It's like if you donated some money to preserve a portion of the Amazon rainforest and instead of doing that they chopped down the trees and sold it for lumber and you were like, that's literally the exact opposite of what I donated money for. It doesn't make sense. And that's what they did. So I'm not happy about that.
But that motivated you to get Grok AI going.
Yeah. Grok is at least aspirationally a maximally truth-seeking AI, even if that truth is politically incorrect. So I mean, you may have seen some of the crazy stuff from OpenAI and from Google Gemini like where it says generate an image of the founding fathers and it generates an image of diverse women.
Speaking of AI, as time goes on and you're more and more embedded in it, how much if at all have your expectations changed?
Well, I always thought AI was going to be way smarter than humans and an existential risk, and that's turning out to be true.
He warns that programming AI with rigid ideological outcomes like demanding diversity at all costs could lead to horrifying results, including the elimination of anyone it deems non-compliant.
The problem is if you program an AI and say the only acceptable outcome is a diverse outcome and that's like a mandate from the AI, then you could get into a situation where it says there's too many white guys in power. Assuming these things don't have empathy, they're going to do what they're programmed to do. So if it's rewrite history and everything's to both women, then it's going to do that. People ask AI which is worse, global thermonuclear or misgendering Caitlyn Jenner, and it would say misgendering Caitlyn Jenner is worse than global thermonuclear. And we got a problem here guys. Even Caitlyn Jenner said, no, definitely misgender me, that's way better. But if you program an AI to think that misgendering is the worst thing that could possibly occur, it could do something totally crazy like in order to ensure that there's no misgendering that can ever happen, we'll just all humans that ensures the probability of misgendering is zero because there's zero humans, which is logical. It's a problem with a thing that's not a human that you want to do a task for you and you give it very specific parameters.
Yeah. And that's one of the things that they've shown about AI is that it'll cheat. They'll cheat in order to accomplish things that they can't accomplish otherwise. They won't follow the rules. They will make copies of themselves and try to upload it to servers if they think that they're being taken offline.
Yeah. I mean, that's like the plot of Terminator, actually.
Literally, it's the plot of Terminator. I just as a reminder, I actually with my little X, my kid, everything's called X, watch Terminator 2, which holds up actually. The plot of it kind of makes sense and I think the AI destroys the world in like 2029, by the way. So it's on track.
Yeah, really, really close.
And in the most chilling prediction of all, Musk says AI will evolve so fast we won't even realize it's taken over, until it's far too late to stop.
You scare the out of me when you talk about AI. Between you and Sam Harris, I didn't even consider it until I had a podcast with Sam once. He made me my pants. I talking about AI, I realized like, oh, well, this is a genie that once it's out of the bottle, you're never getting it back in.
There was a video that you tweeted about one of those Boston Dynamics robots, and you're like, in the future, it'll be moving so fast you can't see it without a strobe light. Are you honestly legitimately concerned about this? Is AI one of your main worries in regards to the future?
Yes, it's less of a worry than it used to be, mostly due to taking more of a fatalistic attitude.
So you used to have more hope and you gave up some of it and now you don't worry as much about AI. You're like, this is just what it is.
