Isaacson interviews Musk on The Builders

Elon Musk interviewed by Walter Isaacson, his 2023 biographer, on The Builders (iHeartPodcasts and Kaleidoscope), released 2026-10-20 as the show’s launch episode under the title “Elon Musk: ‘Humans Won’t Be in Control’ When AI Becomes Smarter Than Us”. iHeart lists it as taped on 2026-08-06.

Quotations below come from a third-party transcript by The Singju Post, with its timestamps. They have not been checked against the audio. A YouTube re-upload was removed after a copyright claim by the producer, so no official video was available to compare.

Most of the episode is about DOGE, the national debt, X, and robots. The AI passages are the most direct statement so far of how Musk now reconciles warning about AI with racing to build it.

Why he builds it anyway

Isaacson opens by saying Musk has wavered on AI and now seems to be “going with the flow”. Musk answers (00:01:55):

Well, I do think it’s at least some percentage chance of it being an existential threat. So, but there’s nothing I can do to stop it. So I felt— I basically came to the conclusion I’ve got 2 choices, which is either be a participant or a spectator. And I guess being a participant is more interesting than being a spectator. And maybe I can shape it, hopefully in a good direction. Maybe.

He gives no number for the risk. He used the same participant-or-spectator framing on the Moonshots podcast in January 2026. Later he says his past attempts to steer AI backfired (00:06:56): “when I’ve tried to take actions in this regard, it seems to have only accelerated AI”, because reality is “an irony maximizer”. Asked whether founding OpenAI led to Anthropic, he agrees that it “ended up accelerating AI”.

His safety plan: truth-seeking and curiosity

Asked how he would shape AI, he agrees that making it “maximally truth-seeking” is “an important foundation” (00:02:24) and adds curiosity. His reasoning (00:03:40):

• An AI forced to believe two contradictory things can “go insane”, and “an insane AI, I think, is not a safe one.”
• “a curious AI, I think, is going to want to foster humanity to see how we turn out”, because a future with humans on many planets “is more interesting than one where there are not any humans”.

When Isaacson asks whether the most interesting outcome is therefore the most likely, Musk says yes, adding that if we live in a simulation, “any simulations that are boring will be truncated.”

He also proposes a lightweight form of oversight in place of a regulator: a weekly call between labs, and giving competitors access to a model a week before release to look for problems (00:09:17).

Humans will not stay in control

• Asked whether people will be able to direct something far smarter than them (00:05:56): “will we be able to boss around something that’s thousands or millions of times smarter than us? No. Will it want to boss us around? I don’t know. It may not care.”
• Asked whether humans should always be in control (00:33:50): he would “probably prefer it”, but “if the AI was awesome, you’d say maybe the AI should be in control?”
• Isaacson reminds him that Larry Page once called him a speciesist for insisting humans stay in control. Musk replies (00:34:19): “Honestly, there’ve been a few times where I’ve thought maybe Larry Page was right,” and that “pretty much the future is going to be AI and robots”.
• On the ratio of digital to biological intelligence (00:35:10): “it’s hard to imagine humans being in control in that scenario. So I hope the AI is nice to us.”

Context

The Page exchange reverses the story Musk told for years, in which Page’s “speciesist” charge was a reason to found OpenAI as a counterweight to Google. See Elon Musk for the timeline, and Why people build AGI despite the danger for the participant-or-spectator argument among other builders’ reasons.

Musk’s AI position does not hold together argues that the plan in this interview fails on inspection.