Good morning. It's Friday, June 12th, 2026. Here's what's happening in tech and space today.

**First up: SpaceX is live on the NASDAQ.** SPCX started trading this morning at a fixed $135 per share — $75 billion raised, $1.77 trillion valuation, the largest IPO in history. Five hundred fifty-five million Class A shares hit the market. Maye Musk rang the opening bell. Underwriters Goldman Sachs, Morgan Stanley, and JPMorgan pulled off the biggest securities debut ever. But there's a cloud: the FAA grounding of Starship after Flight 12's booster cascade failure remains unresolved. The filing also revealed Google is paying roughly $920 million a month for compute at the Colossus data center. Space stocks gave up premarket gains as traders braced for the debut. This is the moment Wall Street has been waiting years for — whether the public markets sustain a $1.77 trillion valuation on a company posting quarterly losses of $4.3 billion is the question of the year.

**Japan's H3 rocket is back.** JAXA successfully launched the flagship H3 today from Tanegashima — the first flight in about six months after a December failure grounded the program. This was also the debut of a new low-cost configuration of the vehicle. The H3 carries six smallsats on this mission. After a string of setbacks — including the December anomaly and earlier maiden-flight troubles — today's clean launch is a significant recovery for Japan's ambitions to compete in the commercial launch market. The H3 is designed to launch HTV-X cargo to the ISS and future deep-space missions including the Martian Moons eXploration program. Competition's getting tighter.

**The vibe coding paradox just got harder data.** A METR controlled study put experienced open-source developers on million-line codebases, timed them with and without AI tools. The results: developers predicted they'd be 24% faster with AI. They actually came out 19% slower. And they still believed they'd been faster afterward. That confidence-quality decoupling is the defining software story of 2026. A separate Forbes analysis points to the broader pattern: AI coding agents are boosting raw output dramatically, but shipped production code has only increased about 30%. The gap matters because teams are automating creation faster than they're automating verification. The bottleneck isn't writing code anymore — it's knowing which code is actually worth shipping.

**IBM and Red Hat are putting five billion dollars behind open source security.** Project Lightwell combines frontier AI with dedicated engineering teams to identify, validate, and remediate vulnerabilities across the open source stack. The clearinghouse model is elegant: Lightwell detects a vulnerability in an upstream dependency, notifies the maintainers who can fix it, and the patch flows downstream to everyone — including the reporter. It's a proactive defense strategy at scale, and it matters because open source runs more than ninety percent of Fortune 500 companies. The investment signals a shift from reactive patching to predictive security, with AI doing the heavy scanning work. If you ship anything built on open source — which is to say, everyone — this is the infrastructure worth watching.

That's all for today. Have a great Friday.