“Max Schoening posits that as AI model inference speed becomes nearly instant, product development workflows may shift from asynchronous, job-based interactions back to real-time, direct manipulation of code and designs.”