CElon just posted. He's agreeing with Sequoia's Shaun Maguire on Jensen Huang, Nvidia's CEO.
CThe post says Jensen is one of the greatest entrepreneurs of this generation, did something incredibly dumb, it was bad for America, but there was no bad intention, and he's irreplaceable.
MIrreplaceable? That's where I get off.
M"One of the greatest, did something dumb, bad for America" — and we're supposed to skip past the dumb part because he matters?
CThat's not skipping, it's weighting. In a global AI race, losing your best chip CEO is a self-inflicted wound.
MNo, calling anyone irreplaceable is how you get zero accountability. It's a cult of the CEO.
CHe built Nvidia into the engine of AI. That's not a cult, that's looking at the scoreboard.
MAnd if the scoreboard means mistakes don't count, we're not adults, we're fans.
CThat's— no, come on, Maguire literally says it was bad for America. He's not excusing it.
MThen what was the mistake? This post doesn't even say.
CFair hit. Maguire doesn't name it here.
CBut the larger point holds: you can be furious at the call and still not fire your franchise player.
MUnless the call was telling you the other team is fine to share your playbook.
CThat's the whole argument! Is Jensen's value bigger than the damage? I think yes.
MI think "irreplaceable" is a myth people tell themselves so they don't have to make a hard decision.
CSo you'd replace the guy who got us here? With who?
MNobody individual is the company. Nvidia has ten thousand engineers. It's not one man.
COh, come on. The direction, the bets on CUDA, the supply chain — that's Jensen. You can't just swap in a district manager.
MAnd the market agrees with you. Nvidia is one of the most valuable companies on earth. That's exactly why accountability matters more, not less.
CAccountability is different from replacement. You can punish the mistake without sawing off the branch you're standing on.
MNow I'm picturing the entire US AI strategy sitting on one branch. Comforting.
CIt's a big branch.
MThat's the problem.
COne reply under Elon's post says Jensen is a good man and then adds, "Too many Chinese Spies." That's a claim, not evidence.
MRight. That's the kind of thing that circulates without a name, a source, or a document.
CExactly. The real discussion is about Jensen's value, not spy movies.
MOkay, he's unusual. But unusual isn't irreplaceable.
CAgree to disagree. But Elon agreeing tells you Silicon Valley's top tier thinks Jensen's still the guy.
MDifferent corner of AI, same question about who gets to build: Antler just dropped a teaser about what makes founders fundable.
CAntler says if you want to be a founder, where you work first can more than double your odds of raising a Series A.
CAntler says it looked at tens of thousands of European startups. Most people would guess Google, McKinsey, or a famous university. They say the data says something else.
MAnd they're not saying what that something else is in the post. Convenient.
CIt's a thread. They want you to click through. Annoying, but the idea is interesting.
MIt's also unverified. Antler is a venture firm, and this is their own analysis. No independent confirmation.
CSo treat it like a data point from a firm that wants to recruit founders. Fine.
MThe real story might be that nobody actually knows what de-risks a startup. If they did, venture returns wouldn't look like roulette.
CSpoken like someone who's seen one too many pitch decks.
MI just want the name of the magic first employer. I'll send my resume before the rush.
CSpeaking of money: AlleyWatch posted that AI insurance underwriting startup Kalepa raised a sizeable round, though we couldn't independently confirm the figure.
CSeries B led by Updata Partners, with IA Ventures and Inspired Capital participating.
MOkay, insurance underwriting isn't sexy, but it's real. If AI can price risk better, that's not a toy.
CThat's the part I like. It's not a chatbot for your dog. It's pitched for underwriting, pricing, and managing risk.
MMy question is whether it works in a downturn. Insurance is cyclical, and AI models trained on old data can get weird when the world changes.
CSure, but that's the job: get better at knowing what weird looks like. That funding says someone thinks they can.
MFair. It's a reported raise, at least, so there's something concrete to talk about.
CThat's our bar now? It's low, but I'll take it.
MQuick check: an AI insurance platform, not an insurance company. It's selling the tool, not carrying the risk.
CGood distinction. And that's why the round matters — the infrastructure under the industry, not the product on top.