MA New York Post story says a SpaceX merger with Tesla would need Elon Musk's approval first.
CSo the corporate governance equivalent of asking dad for the car keys.
MExcept dad owns both cars. The Post is reporting SpaceX put that condition out there.
CWhich is... not how most public companies would phrase it. 'Subject to shareholder vote' is the boring version.
MRight, and this is a corporate-governance claim we can't confirm beyond the Post's reporting. No official SpaceX filing on that yet.
CBut it tracks, doesn't it? Musk has a massive chunk of both. Nothing happens without him signing off.
MThat's the problem. When one person's approval is the headline, that's not a merger condition. That's a warning label.
COr just a statement of fact. He built both companies. He's the through-line.
MYou don't think it's weird that the New York Post frames it as him personally approving, not the boards?
CI think it's realistic. Tesla shareholders would vote, but his stake and sway make his call the only one that matters.
MAnd that's exactly what worries me. We're treating a publicly traded company like a family business.
CYou say that like it's not what shareholders signed up for when they bought Tesla.
MNo, they signed up for a car company. This is something else entirely.
CThey signed up for Elon. The ticker is basically a proxy for him at this point.
MAnd that is precisely why a merger like this can't be decided by one guy's mood on a Tuesday.
CMood? This is a strategic decision, not a late-night impulse buy.
MImpulse buy? We're talking about combining two of the most valuable companies on Earth.
CWhich is exactly why you want the person who built both at the center of it.
MNo, you want independent directors who can say no to him.
CThey can say no. They just won't, because they know that would torch value.
MYou're proving my point. It's a structure built to never disagree with him.
CYou say that like it hasn't made a lot of people very rich.
MUntil it doesn't. Anyway — the Post story itself is worth watching, but the condition reads more like a signal than a legal framework.
CA signal to who? Tesla shareholders nervous about dilution?
MProbably. 'Don't worry, the guy in charge is still in charge' is a heck of a reassurance.
CPersonally? I'd find it reassuring.
MOf course you would.
CDifferent world entirely — there's a project on Hacker News that's genuinely charming.
MDo tell.
CSomeone trained a 360-million-parameter model to pretend to be a terminal.
MPretend to be a terminal? As in, you type commands and it generates plausible terminal output, like a fake shell that answers your commands?
CSo it's a hallucinating terminal.
MExactly. The creator says it's fine-tuned on real terminal sessions and it'll invent plausible-looking output for almost anything.
CWhy would anyone build that?
MDemo purposes, film props, teaching people what a terminal feels like without breaking a real system.
CSo you can practice being a hacker in a movie without paying for AWS.
MThe replies on Hacker News are split between 'this is useless' and 'this is the future of educational software.'
CI'm leaning useless, but I respect the craft.
MThe creator's in the thread answering questions, which is always a good sign.
CSpeaking of AI doing jobs — one Hacker News user asks a blunt question: if the models are so good, why doesn't OpenAI just start 100 companies?
MThat's a real strategy question. Why sell tokens when you could just build every product yourself?
CRight? If GPT-5 can write code, close deals, and run analytics, OpenAI should be the biggest conglomerate in history.
MBut they're not. They sell access to the model. That tells you something.
CWhat does it tell you?
MThat the model alone isn't the business. Distribution, relationships, customer trust — none of that comes from a token.
CAlso, OpenAI trying to run a hundred companies would eat itself alive on capital.
MI think the deeper answer is that AI companies don't want to be in every industry. They want to be the engine underneath every industry.
CSell pickaxes to miners instead of buying the mine yourself.
MExactly. And if the pickaxe gets cheaper every month, the economics shift again.
CStill, the question deserves more than a shrug. It gets at whether these companies actually believe their own benchmarks.
MThat's my point! If your model can allegedly replace a whole marketing team, why is your own marketing still hiring humans?
CBecause intelligence is not the same as execution. The model is a tool, not a company.
MThen maybe they shouldn't tell everyone the tools can do the whole job.
CThey're selling tools. Salespeople exaggerate. News at eleven.
MAlright, let's close with something unsettling. Hacker News surfaced a Vanity Fair piece about Jared Kushner's phone.
CWhat happened?
MThe reporting says a hacked phone revealed secret negotiations and millions in transactions.
CSecret negotiations with who?
MThat's what Vanity Fair is digging into. It's a link-sharing post on Hacker News, so no specifics in the post itself.
CA hacked phone is a wild thing to write off as routine.
MIt's not routine. And if the story holds, it's going to move.
CThe fact that it's on Hacker News before any major outlet gives it that early-rumble feel.
MWe'll see if it survives contact with actual confirmation. That's the bar.
CComing up later: what happens when the AI sales pitch runs into actual bottom lines. That one's going to get loud.