Walter Isaacson on Elon Musk, Tesla/SpaceX merger thesis, Elizabeth Holmes, and who should be the Steve Jobs of AI

Walter Isaacson on Musk, AI alignment, and the Steve Jobs question

Walter Isaacson, whose two-year embedded access to Elon Musk produced one of the most detailed portraits of the Tesla and SpaceX CEO, offers a grounded read on a subject the tech world tends to mythologize or vilify.

Tesla/SpaceX convergence

The new Roadster, Isaacson argues, is effectively a joint product of Tesla and SpaceX — the propulsion architecture draws directly on rocket technology — and he expects a formal merger of the two companies eventually. The Roadster reveal, coming years before production, fits a familiar Musk pattern. The Cybertruck chapters in Isaacson's book show Musk overriding nearly everyone around him, insisting on stainless steel and designs pulled from video games and comic books, driven in part by his son Saxon's question: "Dad, why doesn't the future look like the future?"

Isaacson pushes back on the idea that the Cursor acquisition signals a broader shift in Musk toward deal-making. His read is that Twitter was money and impulse, not strategy, and that Musk remains a build-it-end-to-end operator — citing TeraFab, Musk's chip manufacturing effort, as the clearest evidence.

“This is almost a joint product of Tesla and SpaceX... this notion that you can have propulsion done this way for a car, I think it shows why eventually Tesla and SpaceX are gonna merge. Steve Jobs would be appalled by agentic AI — it doesn't connect human creativity to the product and these people aren't stepping back and saying, why are we making this?”

Demon mode

Musk's volatility isn't a new development or a recent deterioration. Isaacson watched it up close for two years, living in an Airstream trailer next to Musk in Boca Chica, sitting through meetings where Gwynne Shotwell was texted to intervene when Musk was about to kill a program. People like Shotwell and designer Franz von Holzhausen know how to pull him out of what the inner circle calls "demon mode." Isaacson's conclusion: don't read any recent appearance as evidence that Musk has become less mercurial.

Elizabeth Holmes

Isaacson met Holmes early, before she became famous or infamous. She had read his Steve Jobs biography more times than he had and dressed to match. He declined to write about her or moderate panels with her because the biochemistry didn't add up — a tiny drop of blood, and she wouldn't explain the mechanism. His ongoing question about Holmes isn't whether she defrauded investors but whether she had conned herself. He still doesn't have a clear answer.

The Steve Jobs of AI

Isaacson frames the central tension in AI around two schools of thought that run back to the 1840s. Ada Lovelace argued machines would never generate their own goals — creativity would always require humans. Alan Turing called this "Lady Lovelace's objection" and said she was wrong. Steve Jobs belonged firmly to the Lovelace school: the computer as a bicycle for the mind, augmenting rather than replacing human creativity, always connected to the human at the other end.

Isaacson's concern isn't with tools like Claude used directly. It's with agentic AI — systems designed to act autonomously, outside human control. Jobs, he argues, would have been appalled: not because agents start without human input, but because the design philosophy treats autonomy as the goal rather than the connection to human creativity.

He's spending time with Anthropic and others in San Francisco thinking through who might write the next chapter on AI alignment — specifically, the philosophers working on how to teach models not just the rules but the reasoning behind them. He has a Marie Curie book coming first, but the alignment question is clearly where his attention is drifting.

On competitive intensity

The current AI rivalries — Musk versus Altman, the shifting Google-Anthropic dynamic, Nvidia's simultaneous role as indispensable supplier and target — read to Isaacson as normal. The Gates-Jobs rivalry was sharper. Carnegie and Rockefeller were rougher still. The frenemy dynamic, he argues, is healthy precisely because it keeps any single player from locking up the field. His one warning: if the competitors start cooperating too much, they'll crowd out newcomers, and that's when antitrust becomes necessary.