Good morning, it's Friday, April 17th, 2026. Here's your morning voicecast.

**Story one: SpaceX completes full-duration static fire of Starship Version 3 ahead of a crucial May test flight.** This week, SpaceX stacked the Super Heavy booster at Starbase and fired up all 33 Raptor engines in a full-duration static fire — first time on a Block 3 vehicle. The upper stage completed its own test the following day. Block 3 brings an integrated heat shield, removes the center engine shielding, and is designed for rapid reusability. The stacked rocket stands 124 meters tall, capable of delivering over 100 tons to orbit. The test flight is set for May, and NASA is watching closely — the next several flights will likely determine whether Starship can support a crewed lunar landing before the end of the decade.

**Story two: Blue Origin prepares for its first re-flight of a New Glenn booster.** On Thursday morning, Blue Origin completed a 20-second static fire of a previously-flown New Glenn booster at Cape Canaveral. The booster, named "Never Tell Me the Odds," flew and landed successfully on the NG-2 mission last year. If the data checks out, NG-3 launching AST SpaceMobile's BlueBird-7 satellite could go as early as Sunday. Blue Origin replaced all seven engines on this flight and tested a new thermal protection system on one nozzle, planning to reuse the originals on a future launch. This makes them just the second company to re-fly an orbital-class booster — after SpaceX, which has reflown Falcon boosters over 550 times.

**Story three: Stanford's AI Index says China has nearly erased the US lead in model performance.** Stanford's HAI institute released its 2026 AI Index this week, and the headline is striking. The Arena score gap between top US and Chinese models has collapsed from over 1,300 points in 2023 to just 39 points — a 2.7 percent margin. Claude Opus 4.6 leads for the US; China's Dola-Seed 2.0 is right behind it. China now leads in AI patent output, publication citations at 20.6 percent versus America's 12.6 percent, and has installed nearly 295,000 industrial robots compared to 34,000 in the US. All of this despite less investment capital and tighter export controls.

**Story four: Robotics startup Nomagic hires a DeepMind researcher to lead Physical AI for warehouse robots.** Nomagic has appointed Dr. Markus Wulfmeier as Chief Scientist, moving from Google DeepMind where he worked on robotics and reinforcement learning. He'll develop Vision-Language-Action models trained on what the company calls its "Library of Chaos" — millions of real-world edge cases collected from live warehouse operations. The goal: robots that can handle the messy, unpredictable reality of actual fulfillment centers. It's the same paradigm transformers brought to language — applied to physical manipulation.

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