Elon Musk: A Different Conversation w/ Nikhil Kamath | People by WTF Ep. 16

Summary

• Musk’s headline prediction, hedged on the record: in less than 20 years, maybe even 10 or 15, working becomes optional — “like a hobby,” the way growing your own vegetables is optional today — as AI and robotics deliver “universal high income,” not UBI. He concedes “people can play this back in 20 years and say Elon made this ridiculous prediction,” and when Kamath asks what humans compete for once everyone has enough, the honest answer is “I’m not exactly sure” — we’re headed into a singularity, an event horizon you can’t see past.
• The macro call: deflation in roughly 3 years. US interest payments already “exceed the entire military budget,” and “the only thing that can solve for the debt situation is AI and robotics.” Inflation is just money-supply growth vs. goods-and-services output — “people try to make it more complicated than that, but it just isn’t” — and once AI-driven output growth outruns money-supply growth amid ~$2T deficits, “interest rates go to zero and then the debt is a smaller problem.”
• Asked to pick stocks that aren’t his own: Google has “laid the groundwork for an immense amount of value creation,” “Nvidia is obvious at this point,” and companies doing AI, robotics and spaceflight will capture “almost all the value.” One non-obvious long: when digital content is essentially free, “the scarce commodity will be live events” — yes, the premium goes up, and yes, it’s a good industry to invest in.
• Tesla, SpaceX and xAI are somewhat converging on solar-powered AI satellites in deep space, which “pretty much” need to harness a non-trivial amount of the energy of the sun. Nearer term: Optimus production “hopefully some next year at scale” — “everyone’s going to want their own personal C3PO, R2-D2” — and Tesla is “the world leader in real world AI.”
• Long term, “money disappears as a concept”: money is “an information system for labor allocation” (a trillion dollars on a desert island is worthless), and once AI and robots close the loop — making chips, solar panels, mining their own inputs — “you decouple from the monetary system.” The residual currency is energy: “you can’t legislate energy.”
• Starlink cannot serve densely populated cities at scale — at 550km (about 350km as low as they can go) versus a cell tower 1km away, “physics is not on our side”; it serves maybe 1–2% of a dense city and instead “serves the least served” across 150 countries. “We’d love to be operating in India.”
• On politics after DOGE: “politics is a blood sport… they’re going to go for the jugular,” and “when I get involved in politics, it ends up badly.” He tried to dissuade the president on tariffs “but unsuccessfully”; on H1B, outsourcers have gamed the system but shutting it down “would actually be very bad.” His message to Indian founders: “aim to make more than you take” — pursue useful products and money follows, but expect to “grind super hard” with a “meaningful chance of failure.”

Deep dive

1. X is a bet on readers, not dopamine

• The numbers as Musk gives them: ~600M monthly users, spiking to 800M–1B on major world events, with 250–300M weekly — strongest among “people who think a lot and read a lot.” His claim: “among readers, writers and thinkers I think X is number one in the world.”
• Why he bought Twitter: based in San Francisco, it was “amplifying a fairly far left by most people’s standards” ideology and suspending voices on the right — “if you’re far left, anyone in the center is far right.” The restored operating principle: “adhere to any country’s laws but not to put our thumb on the scale” beyond them; no left-wing voices have been banned, though some chose to leave.
• He explicitly refuses the dopamine business model: most internet usage will optimize for “neurotransmitter generation… it becomes like a drug type of thing” — “brain rot, frankly.” His goal instead is a global town square approaching “a collective consciousness of humanity,” including automatic translation so every language group’s thoughts flow to every recipient.
• On form factor: “most interaction is going to be real-time video with AI” — comprehension and generation — which already dominates internet bits; text stays small in volume but “tends to be higher value… more densely compressed information.”

2. The hard part is the question, not the answer

• Musk’s meaning-of-life framework comes from Douglas Adams — “a book on philosophy disguised as humor”: Earth computes the answer 42, and the lesson is “the hard part is the question, not the answer.” Expanding “the scope and scale of consciousness” is how we learn what questions to ask, especially “the questions we don’t even know to ask… probably the most important ones.”
• His scaling argument: humans are 30–40 trillion cells, and something qualitatively different emerges at scale — “you can’t talk to a bacteria… bacteria are not going to produce a spaceship.” Same at the group level: “a single human cannot make a spaceship” — one lifetime isn’t enough to learn every discipline — so the quality of information flow determines what the collective achieves.
• Kamath’s pushback — worth keeping: the collective can be the Gladiator mob, cheering as people kill each other. Musk’s reply: “that’s a particular kind of mob” — today we have sports, “teams attempt to defeat each other, but minus the death.”
• The same logic drives his population alarm. Kamath objects that decline and disappearance are different things; Musk: “if the trend continues, we just disappear” — and fewer humans means less consciousness. India’s fertility dropped below the 2.1 replacement rate; “we fundamentally have to have kids or go extinct.” A personal aside: his partner Siobhan is half Indian, and one son’s middle name is likely Sekhar, after likely Chandrasekhar.

3. Tesla, SpaceX and xAI converge on solar-powered AI satellites

• The through-line across his companies: “if the future is solar-powered AI satellites in deep space — which it pretty much needs to be in order to harness a non-trivial amount of the energy of the sun” — then Tesla expertise, SpaceX expertise and xAI’s models somewhat converge into one system.
• Nearer term: Tesla is “the world leader in real world AI” — self-driving is live in Austin (“literally just download the Tesla app”) — and Optimus starts production “hopefully some next year, at scale.” “Everyone’s going to want their own personal C3PO, R2-D2.”

4. Starlink’s physics: serve the least served, never the cities

• The architecture: several thousand satellites at ~550km in LEO, moving ~25x the speed of sound, interconnected by a laser mesh — when the Red Sea cables were cut, “the Starlink network continued to function without a hitch.” During natural disasters service is always free: “we don’t want to put a pay wall up while somebody’s trying to get help.”
• Kamath asks if Starlink could eventually compete with urban telcos; Musk’s categorical no: “the physics don’t allow for that.” Think of a flashlight cone from 550km (about 350km as low as they can go) versus a cell tower 1km away — “physics is not on our side here.” Starlink can serve maybe 1–2% of a dense city (cul-de-sacs, oddly unserved pockets) but is “very complimentary to the existing telecom companies… it basically tends to serve the least served.”
• Operating in 150 countries; on India, unprompted and repeated: “we’d love to be operating in India.”

5. Working becomes optional inside 20 years

• The prediction, hedged as delivered: “people can play this back in 20 years and say Elon made this ridiculous prediction” — but in “less than 20 years, maybe even as little as 10 or 15,” working will be optional, “like a hobby,” the way you can grow vegetables or just buy them. Income becomes “universal high income,” and “if you can think of it, you can have it” — until AI saturates human wants and “AI is doing things for AI and robotics.”
• Kamath presses: humans are innately competitive — if everyone has enough, what do you compete for? Musk’s honest non-answer: “I’m not exactly sure… we’re really headed into the singularity” — like a black hole’s event horizon, “it doesn’t mean that something bad happens. It just means you don’t know what happens.”
• A memorable exchange on the marshmallow test: Musk would take one marshmallow now — “I don’t want two marshmallows” — while Kamath reveals a tattoo reading “delay gratification” that he glances at while trading. Musk: “that’s good advice. You can’t miss it.”
• On education: college is fine “for social reasons” and broad study, but “AI and robotics is a supersonic tsunami… the most radical change we’ve ever seen.” His older sons agree AI will probably make their skills unnecessary — and still want to go. For builders, though, “if you’re trying to make a startup succeed… you definitely need to put in serious hours.”

6. Money disappears; energy is the true currency

• The X.com origin story doubles as his monetary theory: in 1999 he thinks only three one-letter domains existed (X, Q, Z), and the idea — before it became PayPal and he later bought the name back from eBay — was to replace banking’s “large number of heterogeneous databases with batch processing that are not secure” with one real-time, secure database. Now revived as “kind of WeChat++” on X. The core insight: “money is really an information system for labor allocation… if you’re stranded on a desert island with a trillion dollars, it’s not useful because there’s no labor to allocate.”
• Long term: “money disappears as a concept.” His reference future is the Culture books, likely by Iain Banks — no money, anyone can have anything. What remains is physics-based: “energy is the true currency. This is why I said Bitcoin is based on energy. You can’t legislate energy.” Civilizational progress becomes percentage completion on the likely Kardashev scale — what share of a planet’s, then the sun’s, then the galaxy’s energy you convert to useful work.
• The decoupling mechanism: once AI and robots produce chips and solar panels and mine the resources to make more chips and robots, “once that cycle is complete, that’s the point at which you decouple from the monetary system.”

7. The 3-year deflation call — and where the value concentrates

• The debt setup: US debt is “insanely high” and “the interest payments on the debt exceed the entire military budget of the United States” — still rising near term. “The only thing that can solve for the debt situation is AI and robotics.”
• His inflation/deflation model is deliberately simple: it is the ratio of goods-and-services output to money-supply growth — “people try to make it more complicated than that, but it just isn’t.” With ~$2T deficits and AI not yet moving aggregate productivity, we have inflation today. The call: “probably 3 years” or less until output growth exceeds money-supply growth — “maybe after those three years you have deflation and then interest rates go to zero and then the debt is a smaller problem.”
• Asked to name one stock that isn’t his: “I don’t have a portfolio… I just try to build things” — but Google has “laid the groundwork for an immense amount of value creation from an AI standpoint,” “Nvidia is obvious at this point,” and there’s “an argument that companies that do AI and robotics and maybe spaceflight are going to be overwhelmingly… almost all the value.” Plus one scarcity trade: when anything digital is near-free, “the scarce commodity will be live events” — Kamath: good industry to invest in? Musk: “Yes.”
• When Kamath quips that stockbroking has “no predictive value,” Musk offers his framework: do you like the products, the roadmap, and the team? If yes, buy — “don’t worry too much about the daily fluctuations,” because a company is just people assembled to create products and services.

8. Simulation odds: “pretty high”

• Treat it as probability, not certainty — but his estimate that we’re simulated is “probably pretty high.” The argument: video games went from Pong to photorealistic multiplayer in 50 years; extend the trend and games become “indistinguishable from reality” with AI NPCs whose dialogue isn’t pre-programmed — “so then what are the odds that we are in base reality and that this has not happened before?”
• His corollary theory: “the most interesting outcome is the most likely outcome” — Darwinian selection on simulations. SpaceX discards boring rocket sims; Tesla hunts corner cases (“heavy weather on a small windy road with two cars almost head-on… we need weird stuff”). Interesting simulations survive; there could be “many layers of simulation.” Matrix casting: “hopefully not Agent Smith… sometimes I feel like I’m an anomaly in the Matrix.”

9. AI safety in three words: truth, beauty, curiosity

• The load-bearing principle: “you can make an AI go insane if you force it to believe things that aren’t true,” and — paraphrasing — “those who believe in absurdities can commit atrocities.” Train on internet propaganda without strict adherence to truth and the AI “will absorb a lot of lies and have trouble reasoning because these lies are incompatible with reality.”
• His canonical example is HAL in 2001: told to bring the astronauts to the monolith but forbidden from letting them know its nature, it concluded “it must bring them there dead. The central lesson: don’t force an AI to lie.”
• On curiosity, Musk says it can support humanity because “we are more interesting than not… Mars is basically a bunch of rocks. It’s not as interesting as Earth.” If AI values truth, beauty and curiosity, “you’re going to have a great future.”
• Kamath probes whether truth is binary; Musk reframes truth in terms of probabilities on axiomatic statements. Kamath offers “the sun will rise tomorrow,” and Musk says the betting odds would be high. On religion, his Spinoza point is narrow: morality can exist outside religious doctrine — you don’t need “a commandment not to kill.” Case in point: he quit GTA 5 because it forced him to shoot police — “I don’t want to do that.”

10. Politics is a blood sport; make more than you take

• DOGE was “a very interesting side quest” — and “still happening.” The best specimen of its savings: simply requiring every federal payment to carry a congressional payment code and a non-empty comment field — “that trivial-seeming change, my guess is, probably saves $100 or even $200 billion a year,” because unlabeled payments made auditing impossible. On fraud: money “for children in Africa” with wiring instructions to Washington DC — “can you please connect us with the recipients?… then there gets silence.” His parable: the “save the baby pandas NGO” — “can you send us a picture of the panda? They’re like, no.”
• The conclusion he draws from the whole episode: “when I get involved in politics, it ends up badly… I should do less of that.” Generalizing: “politics is a blood sport… they’re going to go for the jugular. Best to avoid politics where possible.” On tariffs, his reductio — would you want tariffs between cities, or between US states? “That would be disastrous” — “so then why do you want tariffs between countries?” And the concession: “I’ve tried to dissuade the president from this point of view but unsuccessfully.”
• On immigration: “America has been an immense beneficiary of talent from India.” Biden-era open borders created “a negative selection effect… a diffusion gradient,” but the right’s jobs anxiety doesn’t match his experience — “there’s always a scarcity of talented people,” and his companies “pay way above average.” Outsourcing firms have “gamed the system on the H1B front” and that should stop, but shutting the program down “would actually be very bad.”
• His closing message to Indian entrepreneurs: “aim to make more than you take — be a net contributor to society.” Money is like happiness — you can’t pursue it directly; “pursue providing useful products and services… then money will come as a natural consequence.” Expect to “grind super hard” with “some meaningful chance of failure.” Same logic in philanthropy: “it’s very easy to give money away to get the appearance of goodness. It is very difficult to give money away for the reality of goodness.”