MOpenAI is reportedly chasing another thirty billion dollars... at a valuation that would make it the most expensive private company in history.
COne point two five trillion. That's not a number, that's a warning sign.
MTechCrunch says the raise is ahead of a planned IPO. They want to break records before they even go public.
CThirty billion. For what, exactly? They already have more compute than most countries.
MCompetition. Anthropic, Google, everyone's spending like it's a war. It kind of is.
CBut here's what gets me. They're not profitable. Nobody in this space is profitable, not at this scale.
MAnd yet the investors keep lining up. You tell me why.
CBecause they're afraid of missing the next Google. But at one point two five trillion... that's priced like it already happened.
MThe deep dive notes say if this closes, it's the most highly valued private company ever. That's not hyperbole, that's just math.
CAnd what happens when they go public and the real market gets a vote? That's the part that worries me.
MMy take: the private market is giving them a premium for being the leader. The public market won't be as kind.
CI disagree. If they keep growing revenue like they have, institutions will pay up. They're not buying earnings, they're buying positioning.
MPositioning for what? A compute monopoly? Because that's the only way these numbers make sense.
CNo, positioning for enterprise AI. Every company in the world is going to need these models. That's the pitch.
MEvery company needed a website in 1999 too. We both know how that ended.
CThat's different. AI isn't a website, it's infrastructure. Like electricity.
MElectricity companies don't lose billions a year running the grid.
CFine. Maybe it's more like early railroads. Losing money, building track, everyone thinks they're crazy... until the trains arrive.
MAnd most of those railroads went bankrupt. The big winners came in later.
COkay, that's fair. But the ones who survived became the backbone of the country.
MSo we're both saying it's a coin flip. One point two five trillion, and it's still a coin flip.
CNo, I'm saying the odds are better than most people think. You're the skeptic.
MI'm the one reading the actual financials. Call me crazy.
CWhile they're out raising money, a16z is quietly walking away from the crypto label.
MSaw that. a16z Crypto is now a16z Frontier. They're broadening their scope.
CTechCrunch says it's because blockchain investing has fallen out of fashion. Chris Dixon still runs it.
MWhen the smart money rebrands, it's usually because the old brand stopped working.
CWas 'Crypto' scaring off their own limited partners? That's a real question.
MMore like they want to chase AI deals without looking like they missed the boat.
CThey had a huge crypto fund. Hundreds of millions deployed. You don't just walk away from that unless something changed.
MWhat changed is that crypto winter lasted three years. Investors got tired of explainer threads and dead tokens.
CI still think blockchain infrastructure will matter. But calling it 'Frontier' lets them invest in anything. Smart hedge.
MOr a sign they're admitting the old strategy didn't work. Either way, the name change is the story.
CAnd over in Australia, someone did napkin math on batteries versus a giant hydro project. The numbers are wild.
MThe Snowy 2.0 thing. The post said twelve billion dollars could buy thirty-five gigawatts of Tesla batteries instead.
CThat's the hook. But here's the problem: Snowy 2 gives two weeks of continuous output. Batteries give two hours.
MRight, so they're not the same thing at all. One is a bridge, one is a backup.
CThe post is misleading, but it still raises a fair point. Why are we spending thirty billion on something that won't finish for a decade?
MBecause Australia needs long-duration storage. Solar and wind don't run at night or during calm spells. You need something big.
CBut if the batteries are ten times cheaper per gigawatt... maybe you build both and stop pretending one replaces the other.
MI'll give the post this: it got everyone doing the math. That's valuable, even if the conclusion is flawed.
CThe cost overruns on Snowy 2 are genuinely brutal though. Twelve billion more on top of what they've already spent? That's real money.
MAnd every overrun makes the battery comparison more tempting. Even if it's apples to oranges.
CMeanwhile, someone on X made a fake betting slip with AI to prove how easy it is to fake a winning ticket.
MAlan Rowland posted about it. Someone edited his slip to show a million-dollar payout, just to troll him.
CWhich actually proves a serious point. If anyone can fake a winning slip, how do you verify any big payout claim online?
MYou don't. That's the whole problem. A screenshot is not proof anymore. It's a suggestion.
CIt's a good reminder that AI didn't just change art and writing. It changed evidence.
MAnd speaking of AI art, a VTuber on X is furious that she accidentally followed an AI artist. Her words: 'I don't support AI artist.'
CShe said it showed up in her feed and she followed back without checking the profile first. A common mistake, honestly.
MShe's disturbed that people can be 'this way.' I think the AI art fight is only going to get louder.
CIt already has. Artists feel replaced, audiences feel deceived, and nobody's solved the labeling problem.
MThe deeper truth: AI art is so good now that even a VTuber with a curated aesthetic can't always tell.
CThat's not a bug, that's the whole point. The models are trained to be indistinguishable.
MAnd until platforms require clear labels, this will keep happening. People will keep getting fooled.
CAnd that's the show tonight. Big money moving, labels changing, and a fake betting slip that proves a real point.
MTomorrow, we'll see if that OpenAI raise actually closes. If it does, the whole market shifts.