Good morning. It's Wednesday, April 15th, 2026. Here's what's happening in tech, space, and AI today.
**SpaceX hits a thousand.** Yesterday morning at Cape Canaveral, a Falcon 9 launched Starlink batch 10-24 โ 29 satellites โ bringing SpaceX's 2026 total to over 1,000 Starlink satellites deployed. That was the 37th dedicated Starlink mission of the year. But that wasn't even the main show. SpaceX flew two Falcon 9s in a nineteen-hour window: one from Vandenberg on the West Coast, one from Cape Canaveral on the East Coast. The booster from yesterday's mission, B1080, made its 26th flight โ a veteran that previously flew two Axiom crew missions. Forty-six Falcon family launches this year so far. The cadence isn't slowing down.
**Rocket Lab's new Hall-effect thruster, built in Long Beach.** Rocket Lab announced Gauss โ a new electric satellite propulsion system designed for high-volume production. Hall-effect thruster, power processing unit, propellant management โ all integrated, all designed in-house. They've already built a production line capable of cranking out more than 200 thrusters per year, which is notable because electric propulsion has historically been a massive supply chain bottleneck for constellation operators. Higher specific impulse than chemical propulsion, meaning less propellant for the same mission delta-v. Named after Carl Friedrich Gauss, joining the Rutherford, Archimedes, and Curie engines in the family. This comes from the company's special projects team in Long Beach โ right down the 405 from San Pedro.
**The AI cybersecurity arms race accelerates.** OpenAI released GPT-5.4-Cyber yesterday โ a fine-tuned variant of GPT-5.4 optimized for defensive security work. Not public. Restricted to verified security researchers and government agencies. It's a direct response to Anthropic's Claude Mythos, which reportedly outperformed competing systems in complex attack scenarios and autonomous exploit generation during UK AI Security Institute evaluations. The window between vulnerability discovery and weaponization has collapsed from weeks to hours. Both models represent the same inflection: AI systems that can find and exploit zero-days faster than any human team can patch them. Welcome to defensive security at machine speed.
**Anthropic pivots hard into biology.** Fresh on the heels of the Coefficient Bio acquisition โ that's the $400 million deal for an AI biotech startup founded in 2025 โ Anthropic announced that Novartis CEO Vas Narasimhan has joined its board of directors. This is a signal. Anthropic isn't trying to be a better chatbot company. It's positioning Claude as foundational infrastructure for drug discovery and computational biology. Coefficient was founded by former Evozyne, Genentech, and Prescient Design researchers โ people who understand that folding proteins and designing molecules is fundamentally a search problem. Anthropic is betting it can win that search.
And that's today's briefing. Stay curious.