I worked at CourtTV for half a decade. And I say that not as a flex (it’s still a flex, right?) but to give everyone some insight into how trials happen — because they rarely do actually happen.

My job at CourtTV was technically “Associate Producer” but I was really tasked with scouring the country to find interesting trials. Each week I’d write up and present a packet of anywhere from 4-6 upcoming cases to the executive producers. They would then ask a bunch of questions and ultimately decide which ones were worth covering. We’d go through an application process with the courts and judges in the states that allowed cameras, book hotels for hosts and field producers, coordinate with other networks who applied to cover things and then, the Friday before the trial was about to begin… it would be delayed by six months. Or plea out. Almost every time.

The ones that didn’t reach agreements were the high-profile ones (Anna Nicole Smith, O.J. Simpson, etc) or some random heinous acts that defense attorneys hoped would end with better outcomes than the deals offered (if any were at all). Still, by and large, cases resolved themselves early — sometimes the morning of jury selection.

But the ones that didn’t? Big or small, they usually became compelling TV because there were no plea deals to be made — each side had a solid argument, and the outcomes were usually 50/50. Even when they announced a verdict was reached, we usually had no idea what it would be outside of some lucky “tell”s (a quick deliberation, some revealing questions the foreman asked, etc).

I bring this all up because Elon Musk has gone to the trial phase against OpenAI and Sam Altman. Of the initial 26 claims, he eventually settled on two, and is seeking $180 billion dollars — which he wants directed to OpenAI’s non-profit arm. The New York Times has a great story and live blog on the whole thing.

This trial is important to us for two reasons. Actually probably far more than that, but for the purposes of this newsletter, we’ll leave it at two. First, the outcome of this trial (and we won’t have a “true” outcome for years, as the loser here will almost definitely appeal) could help influence the future of AI and, of course, humanity as a whole.

Secondly, we care because Kalshi allows you to try your hand at predicting the outcome.

It’s a fantastic case, full of star power (an unpredictable Musk will be on his third day of testimony on Thursday), regrettably-sent emails, legitimate geniuses, and the still-evolving world of AI. It’ll all be decided by a jury that the judge tried to eliminate news junkies from. And it should be a legit 50/50 shot in either direction, despite the Wall Street Journal painting Musk as an underdog before the trial kicked off.

And it has delivered. And the markets have become a roller coaster as each day passes.

At 4:17 a.m. ET on Wednesday, the odds shot up to 67% on YES that Musk would win. By 5:45 p.m. Wednesday, YES had plummeted to 48%. Musk’s day on the stand featured a relatively positive morning followed by combative questioning from OpenAI’s lawyers in the afternoon — his somewhat flummoxed responses likely pushed his numbers down. By midnight they had gotten back up to 50-52%.

And that’s when I jumped in. We’re officially on YES!

And here’s why I’m not sitting this one out: people are generally scared out of their minds by what AI can become in the wrong hands. And Elon Musk, whether you believe him or not, has positioned himself as the last line of defense against all the evils that could come from it, driven by greed and bad actors.

So my hunch is that, despite jurors trying their best to be impartial, you’re going to get some people who see Musk as genuinely trying to do what is best for humanity. There are recurring themes like this one throughout the trial and pre-trial coverage:

It’s not just about Musk not wanting the company to be for profit — he has repeatedly said that his role is to very literally protect humankind:

And so if the jury gets sympathetic (even subconsciously) to Musk’s “savior of humanity” role? He gets the win. I think he gets the win, at least. There’s a twist — the jurors in this trial are acting in an advisory role to the judge. She can do her own thing if she chooses, but if the overwhelming feeling is that Musk has won, and she goes against it? That could get ugly.

If a couple jurors stubbornly hold out? We normally would have a hung jury, which would lead to, as the rules state, “no superior outcome for either party.” But in this case… I’m not sure. This is an odd little setup that’s harder to predict because of the power put in the judge that is somewhat limited by the jury.

Keep in mind, though, if something happens to sway public support — maybe analysis from a CNN or Fox News panel that Musk is “definitely going to win after what we saw today in court,” or a tweet from a giant legal/political/news account that says he has this won — and the YES goes up to 75 or 80 cents? We can sell. That is the beauty of the markets.

Sam Altman is also not a super magnetic character. He was fired once from OpenAI in a decision reached by several people. He’s done some questionable things. And advising jury could be turned off by Musk’s combative second day, but they could be even more turned off by Altman. We just need Musk to be, in the most cynical way of putting it, the lesser of two evils. And he may actually be the good guy here. We’ll never know.

So our play of the day is Buy YES on Elon Musk as close to 50% as you can, and let the media paint him as the “good” guy, so there’s at least a public opinion-fueled cash out opportunity if the judge sides with OpenAI. Despite OpenAI’s attorneys trying to strategically remind jurors that Musk is a Trump ally, the idea of him standing in the way of out-of-control-AI is a much more important factor a juror/judge combo would consider. This is a game of biases. And I am guessing public sentiment will be on his side for a win.

One important “small print” note, too — If the two sides decide to not put their faith in the hands of an advisory jury and the judge and settle it at any point before a verdict, a publicly disclosed settlement (again, according to the fine print, it has to be publicly disclosed) would end up as as “win” for YES. So the OpenAI charity getting $50 million as part of an agreement to end the lawsuit before a verdict means YES wins.

You’ll often see conclusions in this newsletter along the lines of “this is why markets are great! They make trials fun!” but in this case the trial doesn’t need any help. It’s an instant classic already.

Still, it does add a decent layer to it.

Good luck everyone! And may all of your AI ventures have safety valves!