Elon Musk's Latest Interview: "Transformers 3" is Coming, AI "Self-Improvement" Has Already Occurred, After the AI Singularity "Money Will No Longer Matter"

On March 11th, local time, during a public dialogue at the “Abundance Summit” technology conference, Tesla and xAI founder Elon Musk discussed AI progress, the mass production pace of Optimus 3, and the post-“singularity” economic landscape.

In the interview, he provided several clear judgments: AI has entered a self-improvement stage, humanoid robots are about to enter large-scale production, and the economy after the AI “singularity” is hard to predict, but he bets on deflation and that “money will no longer be important.”

When asked about the timeline for data center construction related to SpaceX, Musk did not elaborate, citing “SpaceX is in a silent period.”

“Optimus 3” is about to enter production: large-scale production possible next year

Regarding humanoid robots, Musk revealed that Tesla’s “Optimus 3” (Optimus 3) is nearing completion.

He said: “We are finishing the final stage of Optimus 3, which is likely the most advanced robot in the world right now; other products haven’t come close to its level.”

According to him, Tesla plans to:

• Start production this summer

• Initial output will be low

• Aim for high-volume production next year

Musk emphasized that robot manufacturing also follows the common manufacturing S-curve: slow at first, then rapid expansion.

He said: “Manufacturing output usually follows an S-curve: starts slow, then climbs rapidly.”

Meanwhile, Tesla is designing a new robot manufacturing plant. He disclosed that the design of this factory is significantly different from traditional factories, with the goal of continuously updating robot versions, “possibly launching new robot designs every year.”

AI has entered “recursive self-improvement”: training AI with AI, reducing human involvement

On the pace of AI development, Musk’s judgment is equally aggressive.

When asked whether AI has entered the “recursive self-improvement” stage, he replied: “Actually, this has been happening for some time.”

He explained that current large models have formed a cycle:

• New models are trained with the participation of previous generation models

• Humans still supervise

• But their involvement is decreasing

He said: “Humans’ role in the loop is diminishing. Each generation of models helps build the next.”

He predicts that this process could soon reach higher levels of automation: “Fully automated self-improvement might happen by the end of this year, or at the latest, next year.”

In his view, breakthroughs in AI are accelerating: “Right now, I see an AI breakthrough before bed, and when I wake up, there’s another.”

Post-“singularity” AI: Unpredictable, but he bets on deflation, universal income, and that “money will no longer matter”

Regarding whether institutions can keep pace with AI and robots’ impact, Musk uses the “singularity” as a metaphor: “The singularity is called a singularity because it’s very hard to predict what will happen.”

He mentioned Grok’s hallmark as “the halo around a black hole,” and said, “What happens inside the singularity is hard to know, but it will be very interesting.”

On macro judgments, he provides a clear optimistic premise: he believes the future “will have a series of possible outcomes, not all good,” but “most likely very good,” with a probability “possibly 80% or higher.”

Without extreme exogenous shocks, he said he feels quite comfortable about economic growth: “Without a third world war… I think a tenfold increase in economic scale within 10 years is a fairly conservative prediction.”

He directly attributes inflation/deflation logic to supply explosion: “We will have universal income—basically, giving money to people.” The reason is that “the output of goods and services will far exceed the money supply,” leading to deflation: “Deflation is the ratio of output to money supply… if goods and services grow faster than the money supply, you get deflation.”

Looking further ahead, he believes the importance of money will decline: “Money will become irrelevant at some point in the future.”

He even proposes a valuation method for a “non-human economy”: “I think future AI won’t use human currency; it will only care about energy and quality—wattage and tonnage.”

Employment and “robots making robots”: no layoffs, but expansion, with individual productivity “absurdly high”

Regarding the question of when robots will participate massively in manufacturing and replace humans, Musk emphasizes that the current reality is “there are still many people”: Tesla’s total staff is “about 150,000,” with “roughly two-thirds working in factories in some form”; its supply chain workforce “may be between 1 million and 2 million.”

But his judgment is that efficiency will leap significantly: “We do not plan to lay off or reduce staff. On the contrary, we will increase the number of employees.” The real change is in output per person: “Tesla’s per-person productivity will become insanely high.”

This aligns with his description of technological progress paths: whether AI or manufacturing, “it’s often an S-curve or a series of overlapping S-curves: slow at first, exponential growth, reaching a plateau, then another breakthrough initiating the next curve.”

Full interview translation:

Host Peter H. Diamandis: Audience, as you can see, I am still working hard to turn “hope” into reality.

Musk: You look great.

Diamandis: I feel very good.

Musk: Did you use some anti-aging serum or something?

Diamandis: That’s our “longevity express train,” we’re moving toward that goal. You’re on that path too. I think during our last conversation, you already started embracing the idea of extending lifespan.

Musk: To some extent, yes. I don’t know if we want everyone to live forever, but I think extending “healthy lifespan”—not experiencing a long period of decline and drooling all over oneself—is a good idea. We want to avoid that.

Diamandis: First, congratulations on the collaboration between SpaceX and xAI. It’s a brilliant move that will power humanity’s first “Dyson cloud.” I’m very curious about your timeline for building these data centers. How much bandwidth do you expect to gain in the first year? Please tell us about the speed of achieving this goal.

Musk: SpaceX is currently in a silent period; I cannot disclose information that might cause trouble.

Diamandis: Okay, we won’t discuss that. I understand, but I look forward to the speed.

Earlier this week, we spoke here with Erik Schmidt and another executive from a large-scale cloud service company. I won’t name them, but I’m curious—what do you think about our current stage of “recursive self-improvement” in AI? Have we reached it? Do you think Grok is currently undergoing recursive self-improvement? How is it happening? What’s the timeline for Artificial General Intelligence (AGI) and Super Artificial Intelligence (ASI)? Please give us a rough overview.

Musk: I think we’ve been in the recursive self-improvement stage for some time. Are you referring to fully automated, human-involved self-improvement?

Diamandis: Yes, I mean in terms of AI software.

Musk: In the recursive self-improvement process, human involvement is indeed decreasing. Each new model is built with the participation of the previous one. This has largely happened, but full automation has not yet been achieved. Maybe by the end of this year, or at the latest, next year.

Diamandis: Do you think there will be a “hard takeoff” (sudden explosion) of intelligence by then?

Musk: We are already in a hard takeoff.

Diamandis: Okay.

Musk: Yes, right now.

Diamandis: Yes.

Musk: At this stage, I go to sleep and see a major breakthrough in AI; when I wake up, there’s another.

Diamandis: Truly.

Musk: Honestly, it’s hard to keep up. It’s quite dizzying.

Diamandis: I think many breakthroughs that make you dizzy are actually brought by you.

Musk: Grok is doing very well. In some metrics, it’s the best. For example, it’s the best at predicting things, which can be considered one of the best indicators of intelligence. Grok’s new version is outstanding.

We are still behind in programming capabilities. I was late because I just attended a company-wide meeting on programming, sorting out all the work needed to catch up and surpass our competitors. I believe we can do it. I think we’ll achieve this goal around mid-year.

Also, I think people don’t fully realize how vast intelligence will become or how it will surpass human intelligence to the point of incomprehensibility.

Imagine this: Suppose we use energy that’s a million times more than the current total electricity consumption on Earth—that’s still only about one-millionth of the Sun’s energy output. Essentially, if you expand the US economy by a million times, it would still only consume a tiny fraction of solar energy. If we scale to the size of solar energy, even a million-fold expansion of current economy and electricity use would only utilize about one-millionth of the Sun’s energy.

But what would an economy or intelligence using a hundred million times more electricity than the entire current civilization look like? What would it think? What would it do? That would be an incredible scene. The challenge is that even faintly understanding such vast intelligence is very difficult. But one thing is certain: it could solve all problems you can think of.

Diamandis: Yes, it’s like a long journey. It might sound a bit ridiculous, but I truly appreciate this relentless optimism.

Musk: I see you turning “hope” into reality. Remember that phrase—“turning hope into reality”—it’s quite interesting.

Diamandis: That’s Grok’s achievement. That’s the marketing advice Grok gave me when you were harshly criticizing me.

Musk: Right? So you’re turning hope into reality. But you’ve also been turning pain into reality before.

Diamandis: That’s for sure.

Musk: When AI and robots increase economic output by several orders of magnitude, it will be beyond our imagination.

Diamandis: We might become the planet’s minority of intelligent beings in a very short time, then an even smaller minority, and finally a tiny fraction.

Musk: Yes. Not only on Earth, but across the entire solar system. Because if you develop intelligence only on Earth, the best outcome—meaning the energy it can utilize—is about one-billionth of the Sun’s energy. Limiting ourselves to Earth, that’s the best we can do.

Diamandis: So that’s the part of energy we can intercept, right?

Musk: Yes. Earth only receives a small part of the Sun’s energy, which is the majority of what exists and can be accessed in the universe. So, the intelligence across the entire solar system will be many orders of magnitude higher than on Earth.

Diamandis: Elon, can I ask how far into the future you can see? How many years ahead can you reasonably predict?

Musk: It’s very hard to predict the future precisely. Many things tend to follow an S-curve or a series of S-curves. Development starts slow, then grows exponentially, reaches a plateau, then another breakthrough triggers the next curve.

This is roughly what I see in AI breakthroughs. For example, after a breakthrough, it follows an S-curve that seems to grow infinitely, but then the gains show logarithmic diminishing returns until the next breakthrough. So, AI progress is basically a series of overlapping, interconnected S-curves.

Diamandis: At one point, you might have been able to predict the next ten or twenty years. What’s your current thinking?

Musk: What I’m about to say might sound crazy.

Diamandis: That’s okay. Because you’ve always been excellent at bold predictions.

Musk: Yes, I think within 10 years, the economic scale will be ten times or even more than now.

Diamandis: Yes. You did mention that in over five years, GDP would grow three digits, reaching ten times the current size. But given your predictive ability…

Musk: I think about a tenfold growth in about ten years, which is actually quite conservative. Unless something like a third world war disrupts these plans. But without a third world war, if current trends continue, I’d say the economy will grow ten times in ten years.

Diamandis: I like that. Can you give us an example?

Musk: Humans will establish bases on the Moon.

Diamandis: Yes. And we will have humans…

Musk: Land on Mars.

Diamandis: We will also build mass projection devices on the Moon.

Musk: I think so. I believe within ten years, we can have mass projection devices on the Moon.

Diamandis: Amazing. Gerard K. O’Neill’s space vision is coming true.

At this year’s “Abundance Summit,” four robots appeared on stage together. I’m very excited about “Optimus.” I’m especially curious about the timeline for Optimus 3—when I can buy one or two? When do you expect Optimus to be commercially available? Or will you adopt a leasing model?

Musk: We are currently in the final stages of Optimus 3. It will be the most advanced robot in the world, far surpassing any other. Honestly, I haven’t seen any robot demonstration as impressive as Optimus 3. Frankly, there might be such robots or they are kept secret, but I haven’t seen them. Of course, I need to ensure what I say can be appropriately disclosed.

Diamandis: We are live streaming this conversation on the X platform.

Musk: Okay, then it’s already very public.

Diamandis: Yes.

Musk: I think we will start production of Optimus 3 this summer, but initial capacity will be very slow. The production volume will follow a classic S-curve, gradually increasing, and probably reach mass production sometime next year. After Optimus 4, we will accelerate design iterations. I might try to release a new robot design every year, with yearly improvements.

Diamandis: When I visited the Gigafactory with Dave Bondy, it was an extraordinary experience. Tesla’s factory covers 11.5 million square feet. You also mentioned building a 9.5 million square foot factory there for Optimus—that’s incredible.

Musk: That’s roughly the size.

Diamandis: It must be.

Musk: That would be quite impressive. It will be a completely new factory design, very different from others.

Diamandis: How far are we from “robots manufacturing robots”? You’ve automated most parts of the Gigafactory, with humans playing a very small role. Will robots replace humans in manufacturing?

Musk: We still have many human workers involved. Those directly employed by Tesla, working on the front lines or managing manufacturing, number around 100,000. So, we have many employees. Tesla’s total staff is about 150,000, with roughly two-thirds working in factories in some capacity. Our suppliers might have another 1 to 2 million people.

So, the number involved is very large. We expect Tesla’s per-person productivity to become extremely high. Therefore, we have no plans to lay off staff; in fact, we will increase the workforce. The real change is in output per person: “Tesla’s output per individual will become insanely high.”

Diamandis: That’s incredible.

When you came to join the podcast, we discussed the concept of “sustainable abundance.” You emphasized that we will enter an era of “Universal High Income (UHI),” which goes beyond “Universal Basic Income (UBI).” I wonder if you have further thoughts on how to achieve this goal? Do you have additional insights?

Also, we previously discussed a potential social upheaval period of about two to five years. During this time, until we achieve “de-monetization” and deflation, and reach the stage of UHI, we might need to distribute many relief checks similar to those during the pandemic. Do you have more thoughts on this? Do people really need that hope and vision?

Musk: To be clear, I think we shouldn’t become complacent. We need to be cautious because the future outcomes are diverse, and not all will be good. But for now, I agree with your view that the future is likely to be very good. There’s about an 80% or higher chance it will be very positive.

I do believe we will achieve universal income. Essentially, it’s direct cash transfers. Because the output of goods and services will far exceed the money supply, leading to deflation. Deflation is the ratio of output to money supply… If the growth rate of goods and services exceeds that of the money supply (which I predict will happen), then deflation will occur.

Diamandis: Yes. Many will start new companies and compete, which will accelerate variability and deepen deflation.

Musk: Basically, AI and robots will produce massive quantities of products and provide abundant services, to the point where humans have nothing to do manually. There will always be moments when human desires are satisfied.

So, returning to my earlier example: if the economy is a million times larger than the US economy, all human desires would already be satisfied. Even if it’s a thousand times larger, you might have already met all material needs you can think of.

Diamandis: So, do you think the value of money will significantly decline? Will we enter a post-capitalist era?

Musk: Yes, I believe that at some point in the future, money will no longer have practical significance.

Diamandis: So, you’re heading toward…

Musk: The future societal model might resemble what’s described in Iain Banks’ sci-fi series “The Culture.” I think future AI won’t use human currency; it will only care about energy and quality—power and tonnage.

Diamandis: That sounds a bit ironic, right? Just as you’re about to become a super billionaire worth trillions, money begins to lose its value.

Musk: More or less. All these wealth figures essentially represent my ownership stake in the companies I founded. The money isn’t in bank accounts. Strictly speaking, I only own part of these companies. These companies are doing many useful things, and as they create value, their valuation increases, which translates into my wealth, making the number look high.

Diamandis: Someone once asked me what motivates you. I said Elon’s motivation is problem-solving. He repeatedly tackles the biggest challenges to make the world better. If others can solve these problems, he doesn’t need to do it himself, but the problem is, no one else is doing it. So, I just want to thank you.

Musk: You’re welcome.

Diamandis: I’m curious—do you think democratic institutions and modern organizations can keep up with this “supersonic tsunami” coming? Will they be overwhelmed by the tide of the times? Will they collapse entirely? How should we respond?

Musk: That’s called the “singularity” for a reason. It’s very hard to predict what will happen inside the singularity. Grok’s hallmark is the singularity.

Diamandis: I like that. By the way, your background logo is very beautiful and magnificent.

Musk: Thank you. The halo symbolizes the mass and light around a black hole falling into it. It’s hard to know what will happen inside the singularity, but it will be very interesting. I am very confident that future life will be fantastic. Frankly, I believe AI and robotics are the only ways to solve our budget deficits and avoid national bankruptcy. Your influence has made me more optimistic now. We should indeed be more optimistic.

Diamandis: Thank you, my friend.

Musk: I used to be not very optimistic, perhaps because I was too immersed in negativity.

Diamandis: Combining optimism with realism is ultimately beneficial.

Musk: Absolutely. You must not be complacent or blindly assume everything will go smoothly. Instead, you should strive to steer things toward the good. I mean, amazing things will happen in the future. If we have highly dexterous and extremely intelligent bionic robots, every person on Earth could receive better healthcare than the richest today. By the way, people say I am the wealthiest, but I think those monarchs are actually much richer.

For example, I had to undergo three neck surgeries because the first two failed. My back still hurts sometimes. I wonder if AI can help solve back pain? If it can, that would be a huge victory. I believe it can. Back pain is really uncomfortable. It worsens sleep quality and makes people irritable.

Diamandis: This morning, David Sinclair came to our stage. He’s conducting clinical trials on epigenetic reprogramming. A recent paper shows this therapy can repair joints. So, back pain might be one of the conditions it can eliminate.

Musk: That would be incredible. Of course, just solving back pain would greatly improve human happiness because it’s not a matter of whether you get it, but that you will eventually. The human body has some design flaws.

Diamandis: I’ve always wanted to invite you to Fountain Life Medical Center in Dallas. We’ll help you. When you’re free, definitely come.

Musk: What equipment do you have there? I know you have MRI and CT scans, but what do you do after getting the results?

Diamandis: I’d be happy to send you a list of services via private message. You’re very generous. Next, another great “moon landing” entrepreneur, Ben Lamm, will join me. He runs Colossal, a company resurrecting extinct species, working on mammoths and 15 other species. I heard you want a mini mammoth—true?

Musk: Yes, I think having a mini mammoth as a pet would be very cool. That would be epic. Those furry, adorable little creatures running around and roaring—just thinking about it excites me.

Diamandis: Great, I’ll mention it to Ben. That’s amazing.

Musk: Can someone build a real-life “Jurassic Park”? If it’s built, even with the risk of death, I’d definitely go see it. That would be awesome.

Diamandis: I think if anyone can do that, it’s Ben Lamm and Colossal. He’s engineering life itself. Someone recently asked if he could make a Pikachu, and he said maybe.

Musk: Yes, “Jurassic World” and such would be fantastic.

Diamandis: Okay, I’ll ask him. Elon, thank you very much for joining us and sharing. Thank you, my friend. Let’s give a round of applause to Elon Musk! (Background music: Nothing can stop us now)