# Elon Musk on AI: humans will no longer be in control in ten years  | The Economist

*The Economist* · https://www.youtube.com/watch?v=1X-rr1DKSbY

## Summary

In this short interview with The Economist, Elon Musk predicts AI will exceed the sum of human intelligence in roughly five years and expects an age of "amazing abundance" within a decade, while conceding humans are unlikely to remain in control. He says he has accepted AI's momentum as unstoppable, even though he once tried to slow it by creating OpenAI as a counterweight to Google. He proposes that leading AI companies hold regular calls, give each other early access to review new models, and only involve government when a company fails to address a flagged risk.

## Key points

- Musk guesses AI will exceed the sum of human intelligence in roughly 5 years.
- He predicts an age of "amazing abundance" by 2036, barring a global thermonuclear war.
- He thinks it is unlikely humans will be in control in 10 years, likening it to chimpanzees being in charge.
- He still sees non-zero risk from AI and robots but says the momentum is unstoppable and even a stop button probably shouldn't be pressed.
- He created OpenAI as a counterweight to Google, but says this and Anthropic's spin-out accidentally accelerated AI.
- He proposes leading AI companies hold weekly or bi-weekly safety calls and give competitors one to two weeks of early access to review new models.
- Government should step in only when a company fails to address a risk flagged by competitors.
- He criticises Sam Altman over OpenAI's shift from nonprofit to an $800 billion for-profit, but praises Dario Amodei as principled.

## Notes

### Predictions
- Musk guesses AI will exceed the sum of human intelligence in roughly 5 years.
- After that, there won't be much AI can't do better than humans, apart from being human.
- Assuming no World War III, he expects an age of "amazing abundance" by 2036, where anyone can have anything they can think of.
- He thinks it is unlikely humans will be in control in 10 years, comparing the situation to chimpanzees being in charge if the intelligence gap were reversed.

### On risk and inevitability
- He still sees non-zero risk from AI and robots, including his earlier 10-20% estimate of killer robots wiping out humanity.
- He has concluded he cannot stop AI's momentum, and even if there were a stop button, it probably shouldn't be pressed because abundance is the most likely outcome.
- His philosophy: "let's enjoy the ride" and minimise the probability of a bad outcome.
- He notes the universe ends in heat death, so "it really is all about the journey."

### His own role in acceleration
- Musk long declined to participate in AI, then created OpenAI as a counterweight to Google's near-monopoly.
- Anthropic spun out of OpenAI and is, in his words, currently the leader in AI.
- These actions accelerated AI, which wasn't his intention; he concludes "all roads lead to acceleration of AI" and decided to join the club.

### Proposed safety mechanism
- Leading AI companies should meet or have a call once every few weeks (weekly or bi-weekly) to discuss safety and security issues.
- Competitors should get one to two weeks of early access to review a new model before release.
- Competitors have the technical depth and the incentive to keep each other honest, unlike government officials without frontier expertise.
- Government should step in only if a company fails to address a risk flagged by competitors — which he says is literally what occurred with his model.
- This should be set up immediately; six months is a long time given multiple AI breakthrough announcements per day.

### Personal dynamics
- Musk criticises Sam Altman for turning a nonprofit, open-source, world-owned company into an $800 billion closed-source for-profit.
- He says Anthropic exists because its founders didn't trust Sam Altman, and calls Dario Amodei a principled person who cares about the future of the world.
- No one he has met at Anthropic "set off my evil detector," though he warns against complacency since the road to hell is mostly paved with bad intentions.
- He is willing to set aside personal differences and talk "for the good of the world."

## Transcript

[00:00] I think AI may exceed the sum of human intelligence in about in around 5 years. >> In 5 years? >> Roughly 5 years is my guess. Um So there really won't be anything that AI can can't do better than humans apart from being human perhaps. >> In a more prosaic level what will life be like? >> The most likely outcome is an age of amazing abundance uh where anyone can have anything they can think of. This is this is this may sound preposterous but well here we are in 2026. Let's see where we stand in 2036. Now of course there are things that could uh derail this like a a massive global thermonuclear war or something like that. Um but assuming that there's not World War III uh I think I think we're headed for an age of

[00:59] amazing abundance. So this is I guess a message of of optimism and uh excitement about the future. >> You're sure that humans will no longer be in control in 10 years time? >> I think it is unlikely um in the same way that if you had a vastly more like a a you know the like like if if if the difference in the intelligence between AI and humans is vastly greater than the difference in intelligence between AI and chimpanzees it's hard to imagine that the chimpanzees would be in charge. >> I'm really stuck by this. >> um an evolved form of chimpanzee. So it wasn't that long ago that we were swinging through the trees eating bananas um in the jungle. And we still eat bananas. >> You said there was a 10 to 20% chance of

[01:55] killer robots wiping out humanity. So, >> you sounded worried then. Now you seem very relaxed about it. Have you changed your view? >> I still think there's risk associated with AI and robots. It's not zero. Um I I guess I've come to my my I sort of philosophical conclusion is to look on the bright side. Um I like I I can't see any way to really stop this incredible momentum of AI and robots. Um and and at times I sort of think, well, perhaps, you know, we're even if there was a stop button, we probably shouldn't press it uh because I cuz the most likely outcome is incredible abundance for for all. >> That's what I thought, actually. It struck me that now you seem to have decided it's inevitable.

[02:51] Let's hope for the best and go for the ride. >> Yeah. I mean yes, pretty much. Let's uh let's enjoy the ride as as I think my philosophy at this point. Cuz I mean, honestly, this this it's it's there seems to be just an inexorable uh progress here in in AI and and in robots that um >> But you >> you know, even if I wanted to stop it, I I don't I couldn't. >> Do you think there's still a 10 to 20% chance that killer robots mean the end of humanity? That's a high bar to as they say. >> Well, in long run we're all dead, of course. Um you know, um So, I mean, I I I so >> I mean >> the let let's according to the standard model of physics, the end of the universe is um a long, boring sort of thin, gray soup of

[03:51] uh gradual cooling. Um I mean, it is sort of the heat death of the universe is called. So, now if if the heat death of the universe is in fact the outcome for our reality, then it really is all about the journey. Because the destination is terrible. >> years is not [laughter] very Yeah, yeah, I get that, but that could be a long way ahead. >> long very very very very long way. >> If it's 10 to 20% in 10 years, you would you go into one of your rockets if you thought there was a 10 to 20% chance of it blowing up and killing you? >> Yes, but I I would >> [laughter] >> But let's say there's you can't do anything about it. >> Okay. >> But I guess the only thing we can try to do is minimize the probability of the

[04:31] bad outcome. Um but trying to stop the intelligence in robots um I mean, for the longest time I declined to participate in AI or was, you know, you know, created OpenAI as essentially a counterweight to to Google cuz at the time they were had more or less a monopoly on AI. In creating OpenAI and then having Anthropic spin out of that and now currently Anthropic's leader in AI. Um I mean, these these actions have actually resulted in knock-on effects that accelerated AI which wasn't really my intention. So, it just seems like all roads lead to acceleration of AI. So, then I'm like, okay, well, you can just sort of be sad about it or join the club, I suppose. >> Well, you've certainly joined the club.

[05:30] >> Yeah. >> I mean, you've >> [laughter] >> you've definitely joined the club. >> If you can't beat them, join them, you know. >> A number of the other leaders of of model makers have been pushing for different ways to try and minimize the risks. Um you know, just last week Demis Hassabis came out with a proposal for a private public regulator. Do you think you guys should be working more actively to get something in place to control the risk, or do you think that's impossible, too? >> Yeah. My recommendation to Demis cuz we I talked to him for a few hours before he published his his piece was that um the most immediate thing that we could do is to have the leading AI companies um at least just meet uh or have a some

[06:13] sort of call once every few weeks um and just uh discuss say any safety and security issues. >> But why why haven't you done that yet? I mean, you don't all like each other very much. Yeah, but you don't you're not all I mean, you're also all competing, right? You all want to be first. >> but that's that's where I think um good the competitors can keep each keep each other honest. So, meaning meaning like if um like I think it's quite difficult for someone in the government who doesn't have a deep technical understanding and isn't driving the frontier of of AI to know whether something should be released or not. Um however, the competitors I think what can, you know, if given sort of a week or two to review a new

[07:00] model um can highlight issues and uh and have an incentive to keep the all all the competitors have an incentive to keep the others honest. >> So, you'd have a system where the model makers would check each other's work. If there was something that worried them, you'd tell the government. >> Yeah. >> The government would then take >> If if there's something else was worrying them and that company was not doing anything to address that risk, then that would be the the moment for government to step in and take action, which is literally what just what occurred with my model. >> this all have to happen? >> Well, I think it'd be good to do that. I think immediately, really. >> Yeah, that's what people keep telling

[07:38] me. We we have to get this some kind of system in the next 6 months because of the pace at which the models are improving. >> Yeah, 6 months is a long time. Um >> So, you think shorter than that? >> Well, I mean there's there's so many AI breakthrough announcements. Sometimes multiple per day. I mean, how many were there last week? >> A lot. >> I lost track, you know. >> Yeah, a lot. >> Um >> So, so you you basically, you know, whether it's Demis's proposal or others, you need to get together and get this set up kind of now. It sounds like you're telling me. >> Yeah, I I I would recommend that we at least have some sort of informal um you know, weekly or bi-weekly um call and and and that like so just that that

[08:27] there's that that um there's maybe a week or 2 weeks of of early access by competitors. And and this is why I think the incentives work is because if if, you know, competing companies I think are likely to understand if something is a security risk. And they're they're not going to be shy about highlighting that that their their competitors' models should be delayed in release. >> You guys don't much like each other and you don't much trust each other. You're insulting each other on on social media constantly. So, do you do you think you can do that with people that you dislike that that where there seems to be such kind of personal vitriol between some of these leaders? >> Um Yes, well, I'm not I'm not a fan of Sam

[09:11] Altman because if you started a nonprofit that was meant to be an open-source AI company and um you know, owned by the world and it somehow got turned into an $800 billion for-profit company with closed source, I think you'd be like, "Well, wait a second. That's the exact opposite of what I donated the money for." That that's my issue. I think it's a legitimate one. Um And then Dario I think also is not a fan of Sam Altman because he felt in fact the reason the Anthropic team left OpenAI is because they didn't trust Sam Altman. Otherwise Anthropic wouldn't exist. They would still be at OpenAI. Um I I think Dario is a very principled person and he and he cares about things a lot. Uh he cares about the future of of the

[09:59] world and I I think everyone I've met at Anthropic so far has been I'm on Like no one set off my evil detector. You know, they always seem seem like they're well-intentioned. Um and uh and the road to hell is I think mostly paved with bad intentions. There are a few well-intentioned paving stones in there, so we we don't want to be complacent. Um but uh you know, at the end of the day if we have to talk, we'll talk. I mean set aside our personal differences for the good of the world type of thing.