I'm starting something new.
During my time at xAI, I got to see a clear path towards hill climbing any problem that can be defined in a measurable way. At the same time, I've seen how raw intelligence can get lobotomized by the finest human errors introduced in the agent harness.
The gradients must flow. Learning shouldn't stop at the model weights, but continue to improve every part of an AI system.
Today, I’m launching Nuraline: the missing infrastructure that enables AI-native software to continuously self-improve.