Zara Reyes: Megan, quick question — if you lost a lawsuit on a technicality, would your next move be to call the person you're suing a potential collaborator in The Economist?
Megan Skiendel: Honestly? Only if I was still appealing the lawsuit at the same time.
Zara Reyes: Which is — that's precisely what happened. July 23rd, 2026, Musk sits down with The Economist and says, quote, 'for the good of the world, personal disagreements can wait.' He is talking about Sam Altman. A man he is actively suing in federal court.
Megan Skiendel: And appealing. The Oakland jury ruled against him — unanimously — but on statute-of-limitations grounds. Not on whether OpenAI actually broke faith with its nonprofit mission. That question is still open.
Zara Reyes: Wait — I want to make sure people actually feel how weird that is. He didn't lose the argument. Judge Yvonne Gonzalez Rogers accepted a verdict that basically says: you waited too long to sue. The underlying claim was never touched.
Megan Skiendel: So the 'OpenAI became an $800 billion Microsoft enterprise and abandoned its mission' framing — still lives. Unruled on.
Zara Reyes: And Musk also runs xAI — a direct competitor. So the question we're actually trying to answer today is whether any version of this cooperation is structurally possible, or whether we're watching very expensive theater.
Megan Skiendel: Frankly — I think the answer is somewhere between those two, and it depends entirely on whose incentives you follow.
Zara Reyes: But follow the incentives for a second — because the part that breaks the clean version is Greg Brockman. OpenAI's president publicly endorsing Musk's call for safety meetings within days of that Economist interview. That's not Musk performing cooperation. That's OpenAI accepting the performance.
Megan Skiendel: Right — but that's actually the new thing. Not Musk's outreach. Brockman's yes.
Zara Reyes: Exactly what I'm saying. Brockman signing on is the signal. Everything else is Musk managing his narrative post-verdict.
Megan Skiendel: Think of it like — two business partners split bitterly, one sues the other, loses on a paperwork deadline, then calls and says let's grab coffee. The fight about who cheated whom? Never settled. Procedurally dismissed. That's the whole situation. The 2017 meeting where Musk demanded full control of OpenAI — Ilya Sutskever witnessed that, it's in the court record — that never got adjudicated. The technicality let Musk reframe without conceding a single inch.
Zara Reyes: So the founding mission dispute — the thing Musk and Altman co-founded OpenAI over in December 2015 — still lives, untouched.
Megan Skiendel: Completely untouched. And xAI is still racing. Honestly, Musk proposing industry peer review while running a frontier lab is — I mean, it's not even subtle. He wants the legitimacy of the safety frame and the competitive speed. Both.
Zara Reyes: No but — the headline says feud ends. What's actually new is one endorsement from Brockman. That's it. The ideology? Intact.
Megan Skiendel: One endorsement and zero resolution on the merits. That's the whole signal-to-noise problem with this story.
Zara Reyes: And the clean version — the one circulating — is that this isn't just Musk. Demis Hassabis at Google DeepMind independently proposed an industry-led safety evaluation body. Musk cited it approvingly. So the read is: cross-industry convergence, not a solo performance. That's the take. And I think it's wrong.
Megan Skiendel: The Hassabis endorsement is real, though. You can't just wave it off.
Zara Reyes: No, it's real — but the symmetry question is the tell. Does xAI submit its models to the same inspection regime? Because if peer review applies to OpenAI's models before release and not to Grok, that's not governance. That's a speed bump for competitors.
Megan Skiendel: I mean — I've watched self-regulatory frameworks get drafted. The entities proposing the rules almost never end up subject to the same timeline as everyone else. That's not cynicism, that's just how it goes.
Zara Reyes: Okay but here's the part that's actually kind of genius — and I mean that in the most unsettling way. The Chinese lab inclusion. Musk says invite Chinese labs to inspect each other's models before release. That sounds multilateral, cooperative, good-faith. It is also completely unexecutable.
Megan Skiendel: No framework exists for that. IP, national security — there's no mechanism.
Zara Reyes: Which means it fails at implementation. Which means Musk never has to enforce it against xAI. The proposal is designed to sound serious and collapse quietly. Honestly — that could be intentional strategy, or it could be someone who genuinely didn't think through the geopolitics. Both are possible. But either way the asymmetry holds.
Megan Skiendel: And while all this plays out — he was posting 'don't let your loved ones use ChatGPT' on X. That's not a philosophical position you just pivot off of into cooperative peer review.
Zara Reyes: The courts aren't waiting for the governance theater to resolve — and that's actually the part we need to get to, because what Scott Winters filed against OpenAI over ChatGPT medical advice makes this whole peer-review conversation feel very abstract very fast.
Megan Skiendel: Scott Winters is the whole argument in one person. Florida pastor, no insurance probably, it's two in the morning, he types his symptoms into ChatGPT, follows what it tells him — and it nearly kills him. He's now suing OpenAI and Sam Altman directly. That case is already in a courtroom. Not a framework. Not a proposal. An actual courtroom.
Zara Reyes: Courtroom liability moves faster than any peer-review regime Musk can sketch out in The Economist.
Megan Skiendel: Exactly — and that's the enforcement mechanism the industry doesn't control. They can draft all the safety proposals they want. A jury deciding ChatGPT gave lethal medical advice? That's binding. That changes product behavior overnight.
Zara Reyes: No but — and this is the part that should make everyone uncomfortable — the 2015 OpenAI founding was itself the industry-led safety initiative. Same principals. Musk, Altman, the whole founding cohort said we're doing this to keep AI safe. That was the pitch. And now they're in adversarial litigation over whether it stayed true to that mission.
Megan Skiendel: Oh, honestly — that's the one that should land harder than it does.
Zara Reyes: The self-regulation was the founding document. And it produced — I mean, it produced a ninety-seven-point-four-billion-dollar takeover bid. February 2025, Musk tries to buy OpenAI's nonprofit controlling entity for $97.4 billion. While talking safety. The safety frame and the aggressive competitive move coexisted the entire time. That's not a contradiction he resolved — that's the actual pattern.
Megan Skiendel: So what do you actually watch for? Specifically — because 'does the proposal get enforcement mechanisms' sounds right but it's still abstract.
Zara Reyes: Whether xAI submits to the same inspection timeline as OpenAI. That's the number. Not the press release — the actual submission date. If Grok gets a carve-out or a delay, the whole thing collapses into exactly what Judge Yvonne Gonzalez Rogers's accepted verdict already signals: the legal record moves, the governance theater doesn't.
Megan Skiendel: And Scott Winters doesn't wait for that answer. His lawyers are already in discovery.
Zara Reyes: And that's — I mean, that's actually the thing I keep sitting with. Scott Winters is in discovery, xAI is still racing, Musk is still appealing. All three of those are happening at the same time right now. So the question I don't have an answer to is: if his motives are completely cynical — pure narrative rehab after the verdict — but the peer-review framework somehow gets real enforcement teeth anyway, does the cynicism matter at that point?
Megan Skiendel: Honestly — I'd flip it. If the motives are genuine, but the moment OpenAI or xAI faces an actual competitive disadvantage from the inspection timeline, the framework quietly dissolves. Same way the original OpenAI nonprofit mission dissolved under competitive pressure. Does the sincerity matter then?
Zara Reyes: No. I don't think it does.
Megan Skiendel: Neither do I. Which is — I'm not sure if that's a bleak answer or a useful one.
Zara Reyes: Probably both. Thanks for walking through all of it.
