On Oct. 13, 2024, during Starship Flight 5, SpaceX launched its roughly 400-foot Starship/Super Heavy stack from Starbase, Texas. After separation, the Super Heavy booster returned to the launch tower and was caught by the tower's giant mechanical “chopstick” arms — the first mid-air catch of a rocket stage. The booster was about 226 feet tall and weighed roughly 11 million pounds, per Contrary Research.
Musk had previewed the chopstick-catch idea in August 2021, calling it an attempt to catch the largest flying object ever with robot chopsticks and comparing it to the Karate Kid fly scene. The goal is full reusability: removing landing legs reduces weight and complexity, and catching the booster at the tower is meant to speed turnaround for future Moon and Mars missions.
In August 2021, Musk tweeted that SpaceX would “try to catch largest ever flying object with robot chopsticks,” sharing a Karate Kid fly-with-chopsticks clip. SpaceX then caught the Super Heavy booster with the launch tower's “chopsticks” on Oct. 13, 2024, during Starship IFT-5.
Infisical, a San Francisco Y Combinator-backed secrets management company, announced a $40 million Series B at a $600 million valuation led by Elad Gil, per Forbes. The “lying to AI agents” line is a colorful description of its agent-security work, not a fraudulent fundraise.
Infisical's announced $40 million Series B at a $600 million valuation is for secrets management and agent security. The post's explanation of Series B and secrets management is accurate.
AlleyWatch reports Kalepa, a New York City-based AI insurance platform, raised a $75 million Series B led by Updata Partners, with IA Ventures and Inspired Capital participating.
Wired reports the Cybertruck is being axed/discontinued, the second large EV to get the ax in the last year, and argues that smaller EV proliferation could be better for the U.S. battery supply chain.
Wired reports Stephen Miller will serve as White House chief of staff; the article's
CElon just posted on X a five-year-old clip of himself saying SpaceX would catch a rocket with giant chopstick arms.
MHe's reposting his own 'I told you so'? Of course he is.
CBut the clip is wild, because SpaceX actually did it.
MCaught mid-air. No landing legs. That part is real.
COctober thirteenth, twenty twenty-four, Starship Flight Five. The Super Heavy booster came back to the tower and got caught.
MThey call it Mechazilla. The booster was about two hundred twenty-six feet tall and eleven million pounds.
CThe BBC's science correspondent, Pallab Ghosh, called it 'absolutely astonishing.'
MSpace.com reported the real landing matched SpaceX's pre-flight animation almost shot-for-shot.
CAnd the booster was about one second away from aborting the tower approach.
MSo even the engineers knew the catch was on a knife's edge.
CWhich makes Elon's five-year-old Karate Kid fly comparison less like mad hype and more like a target.
MOh, come on. The target was bananas. The engineers are the ones who turned it into hardware.
CThat's exactly my point! The vision was bananas, and the team still pulled it off.
MNo, no, wait—he's spent years making wild promises. This one landed because the engineers were excellent.
CBut the leadership said, build a giant tower with chopsticks, and they built it. That's not luck.
MIt's survivorship. For every chopstick catch, there are a hundred abandoned promises.
CThis is the one we're talking about, and it worked. Reusability got real.
MFair. It's a genuine engineering milestone. I'll give him the clip.
MSpeaking of wild claims, here's one from the AI world that needs a closer look.
CLet me guess, another startup that raised millions to solve a problem we didn't know we had.
MSaif Ullah posted on X, quote: 'A three-year-old startup just raised forty million dollars by lying to AI agents, on purpose.'
CThat sounds like fraud.
MForbes says the round is real. The fraud framing is clickbait.
COkay, so what actually happened?
MY Combinator-backed Infisical raised a forty million dollar Series B at a six hundred million dollar valuation, led by Elad Gil.
CThe company does secrets management. It stores app keys and passwords so apps, and now AI agents, can use them safely.
MSo the 'lying to AI agents' part is just a colorful way of saying they sell agent security.
CWhich is a real problem. If you let an AI agent loose with a password, it can leak it fast.
MSo you vault credentials and enforce least privilege. That's standard security.
CBut agents don't use dashboards. They call services at machine speed. If one gets phished, it's not one login—it's a hundred.
MThat's still secrets management. Adding 'agent' to the name doesn't make a new category.
CIt does if companies are scared their AI assistant will wire money to the wrong account.
MTerrified, yes. Real market, maybe. But the post said they raised by lying to AI agents. That's not what happened.
CFine. It was a cheap hook. But the round is real, and the fear is real.
MFear sells. Just don't call it lying.
CYou'd think a security company would avoid the word 'lying' in its own narrative.
CMeanwhile, another raise, but this one wants to put AI in the insurance office.
MAlleyWatch posted on X that Kalepa, a New York company, raised seventy-five million dollars in a Series B led by Updata Partners.
CKalepa builds AI for insurers to underwrite, price, and manage risk.
MSo the algorithm decides what your home insurance costs. What could go wrong?
CInsurers already use models. This is just faster and more granular.
MMy worry is black-box pricing. If an algorithm denies you coverage, you can't even see the reason.
COld actuarial tables were black boxes too, just slower.
MThere's a difference between a table and a neural net. One you can audit.
CFair. But if it cuts premiums for safer homes, I'm not going to hate it.
MUntil it's wrong. Let's see the data.