The Nov/Dec 2025 models + harnesses showed the first complete loop to train and deploy AI reasoning to achieve PMF in a domain (coding). Nothing stops simply repeating this domain by domain other than focus.