No, that’s reinforcement learning and small incremental model updates. The real initial training & model deployment is more akin to DNA. Models cannot “learn” the same way humans do.