The AI Grand Prix takes flight with $500k and a job at Anduril on the line. This week, we hosted the Physical Qualifier at Anduril HQ with competitors from 8 countries testing their code live under pressure. USA 🇺🇸, Japan 🇯🇵, South Korea 🇰🇷, Taiwan 🇹🇼, the Netherlands 🇳🇱, Estonia 🇪🇪, Kazakhstan 🇰🇿, and China 🇨🇳. The event doubled as a recruiting event for eligible global talent. Over 3,000 teams registered. 10 teams remain. Find out who wins on Dec 11th. The first-ever AI–GP takes place at Arsenal-1 in Ohio.

Fight Unfair

Most founder and VC types love to publicly celebrate the wins, the product launches, the major milestones... And fair enough! Building new hard tech is indeed incredibly rewarding when things work. Anduril does plenty of celebrating! But what doesn't get talked about nearly enough is how often things *don't* work. Every company building rockets, aircraft, engines, missiles, drones, or any other difficult tech has LOTS of days where something doesn’t work, when a test goes badly, when you thought you got it juuuuust right but womp, not so much. Those days suck, but they are also inevitable on the path to success. Well, Friday was one of those days for Anduril. A solid rocket motor exploded during a test fire at our factory in Mississippi. Most importantly, no one was hurt. The safety systems worked exactly as designed. The team responded exactly the way they've trained to, and damages to our test stand were minimal. By the end of the day everyone was already focused on understanding what went wrong and getting ready for the next test. There's a reason for the cliché "not rocket science" -- because rocket science is actually quite difficult! Friday's test failure is just one example of why. Development testing exists to answer difficult questions before the design goes into production. Obviously we aim and hope for successful tests, but just as obvious, every test failure gives us data. Every anomaly improves the next design. Every test, in the long run, makes the program stronger. Even if Friday sucked (and it did), we take a deep breath and move onto the next one. We'll be back to test firing rockets within weeks. Anduril is continuing to build and test rocket motors weekly, and the production facility remains on schedule. Disciplined iteration begets steady progress, and we're already putting the pieces of our test stand back together for the next test. Onwards.

New test footage: first missile shot from YFQ-44A. YFQ-44A executed an end-to-end, beyond-line-of-sight strike against a simulated target. The test, executed out of @EdwardsAFB, represents an important step in turning CCA into an operational capability.