You have it, profit must go.
I need to stop using the old term AI, although that is now generally recognized, and replace it with the correct term ML (Machine Learning). That may make things clearer. The problem is ML algorithms being made by people with inherent and unrecognized bigotry. There are several major projects now to get ML systems to both monitor themselves and to maintain key examples of the model content they used to achieve a decision. This is encouraging.
And you are correct as to the approach. I do not think we have time to do the slow bottom up approach even if we can learn what we need to solve our problems hence the need for rapid inclusion of ML in the administrative process. But I’m using that also to get around the human stalemate of emotional thinking. We need to get a range of decisions removed from human emotion. There are places where human emotion is good, e.g. compassion and true mindfulness that are far too complex for ML but the systemic problems we are dealing with are administrative and engineering and should be objective. Examples of this that are currently totally bogged down in stupid emotion are: 1) human migration to maximize intellectual capital and efficiently use resources, 2) Financial resource distribution design for UBI and resource allocation, 3) Continuous micro poling to present the best solutions for confirmation by residents to local problems. etc.
Profit as a motivator must be replaced quickly. Sustainability requires for our survival. It is already clear that concept of material capital as the basis of value cannot be supported. The new basis for value is intellectual capital or creativity in all of its forms. This cannot be horded or seen as a zero sum item and that allows the elimination of profit. The issue resource allocation to produce the maximum value in intellectual capital.