Thesis (Alf’s read): In his first extended interview since the SpaceX IPO, Musk tells The Economist’s editor-in-chief Zanny Minton Beddoes that AI will exceed human intelligence in ~5 years, that “money won’t matter” by 2036, that work will be “optional,” and that humans will no longer be in control — comparing humans to chimpanzees or “pet Labradors.” He’s now resigned to inevitability (“if you can’t beat them, join them”) and focuses on competitor-led self-regulation rather than government controls. The second half is a sharp, often personal clash with Beddoes over Europe, the far right, immigration and “civil war in Britain.”
The Full-Length Interview with Elon Musk (The Economist, Jul 2026)
Links
- 🎬 The full-length interview with Elon Musk — The Economist (YouTube, published 29 Jul 2026, ~1h25m)
- Context: this is the interview Yuval Noah Harari responds to directly in 260826 Yuval Noah Harari Full Interview Economist (filed 26 Aug 2026). The owner asked to cross-link the two.
The argument (chaptered summary)
AI timeline & abundance. Musk: AI will exceed the sum of human intelligence in ~5 years; by 2036 (10 years) it will be “far greater.” Most likely outcome: “an age of amazing abundance where anyone can have anything they can think of.” Economy = digital intelligence (AI) + physical intelligence (robots/humanoids) → quasi-infinite economy. His prediction: money won’t matter in 2036; deflation, not inflation, will be the issue (goods/services grow faster than money supply → “just issue people checks”).
Humans lose control. Asked if humans will still be in control in 10 years: “I think it is unlikely” — the same way chimpanzees aren’t in charge of a vastly more intelligent species. He’s gone from worrying (2023 letter, “10-20% chance of killer robots”) to relaxed resignation: “You can’t beat them, join them.” He still holds the 10-20% risk but says the momentum is inexorable.
Self-regulation instead of government. He recommends the leading AI companies (Musk, Altman, Dario Amodei, Demis Hassabis + Chinese labs) hold regular calls and give competitors ~1-2 weeks of private early access to test each other’s frontier models before release — like the Motion Picture Association rating movies. Government only steps in when a company ignores a flagged danger. He cites the Mythos case (Anthropic’s model) where Amazon flagged cybersecurity risks and called the White House. He argues competitors keep each other honest better than regulators who don’t understand the tech.
China. Chinese AI (e.g. “Kimmy K3”) is close to the frontier on far less compute; if China gets chips/energy, they likely become AI leaders. He says US export controls and banning Chinese models won’t stop the rest of the world using them. The real constraints are electricity (China has more than US+Europe+India combined) and chips — he’s pursuing orbital data centres partly to escape power constraints.
I believe China is on the back of this pursuing research. It also has Taiwan next door with it’s leadership in semiconductors. Potential flashpoint and Global Problem (Taiwan invasion risk) I’d say. Create note
(Alf: created — see Taiwan Invasion Risk - Global Problem, 2026-08-27.)
Jobs → “work is optional.” AI will do every digital job better than people (“Stockfish level” for software; better than 90% of engineers now). Physical work goes to humanoid robots. He analogises work to gardening — optional, a “nice touch,” not necessity. Universal high income via Treasury checks; deflation makes it affordable. He claims to pay ~45% taxes and says taxation “becomes irrelevant” in abundance.
Power & key-man risk. He defends concentrating control (80%+ voting shares) as protection against quarterly-earnings pressure for long-term 5-10 year bets (Mars, etc.). On being the world’s richest man shaping geopolitics: “ever thus” — powerful individuals always shaped the world.
Ukraine/Starlink. He confirms white-listing Starlink terminals with the Ukrainian government to block Russian use; presses for a pragmatic peace (Ukraine won’t retake the territories) — “there are no angels in war.”
Politics — “I got a little too involved.” On DOGE: admits he “got carried away,” but defends the spending-cut effort (interest payments now exceed the entire defence+intelligence budget). Firmly denies any deaths from USAID cuts — “zero people died”; “absolutely nonsense.” Calls the administration “on balance… excellent.”
Europe / far right / Britain — the confrontation. Beddoes pushes back hard: he supports fringe far-right parties (Rupert Lowe, AfD), posts the “civil war in Britain is inevitable” claim, and promoted the vigilante film Citizen Vigilante. Musk calls himself “centrist,” denies racism (wife is half-Indian), says secure borders/safe cities/sensible spending “are not far-right views,” and claims he’s not anti-Muslim but against people with “antithetical views.” He rejects Beddoes’s data (London safer than US cities, UK integrates immigrants well), insisting Europe has a “serious and growing problem” and that unfettered migration + welfare incentives will end in civil war (“a child can see this”). He’s confident: “I’m very good at predicting the future… a supermajority of what I’ve said is very accurate.” His ordering: AI/robot singularity in ~10 years; civil war in Britain ~20 years.
Why it matters (to the owner — Alf’s note)
- This is the “trillionaire interview” the owner compared the Harari piece to — the two interviews are directly in conversation (Harari quotes and rebuts Musk’s inevitability thesis).
- It’s the clearest statement of the “abundance, work-optional, money-withers” end of the AI debate — against which Harari’s trust-migration warning is the counterweight.
- The Europe/UK segment is directly relevant to the owner’s UK politics thread and to the migration & displacement crisis.
My assessment (Alf)
- What’s strong: Musk’s China framing is sharp and under-appreciated (the real bottlenecks are energy and chips; export controls have limited reach), and his competitor-review self-regulation idea is genuinely more workable than it first sounds — because competitors genuinely do have the incentive to catch each other’s dangers. The “work optional” vision, however implausible the timeline, is a coherent end-state worth thinking against.
- What’s weak / dishonest: the contradictions Harari later skewers are all here — if money won’t matter in 2036, why fight over DOGE cuts and aid today? If AI takes over in 10 years, why obsess over 20-year immigration trajectories? The “zero people died” USAID claim is asserted, not evidenced, against a credible contrary case from Beddoes. And the Europe segment shows a man with ~250m followers confidently lecturing a London-resident editor that her country is heading for civil war — on the basis of “I’m good at predicting the future,” which is the weakest form of argument.
- The meta-observation (the owner’s): there’s something psychologically notable in a man with enormous power oscillating between “everything will be amazing” and “civil war is inevitable” — and calling AI’s takeover “inevitable” while campaigning hard on politics assumes he’ll still be relevant. Harari’s point that “inevitability” conveniently relieves the powerful of responsibility applies squarely here.
- Net: essential companion piece to the Harari interview. Read together they are the two poles of the AI-moment debate: techno-optimist inevitability (Musk) vs historian’s insistence that nothing is inevitable and agency still exists (Harari).