The goal of putting AI to work on mathematical or scientific research, whether by letting AI generate knowledge on its own or by having AI assist humans in solving problems, is to solve problems and surface new knowledge. Competing for priority of discovery is not the objective. Yet ever since the scientific paradigm became the dominant human cognitive framework, many mathematicians and scientists have treated the rights of first discovery as a central concern. Official recognition of priority delivers reputation and tangible rewards. In the coming AI paradigm era, the way value is captured from new knowledge will change. Recognition will no longer start with validation by a small scientific community and only later expand to wider reputation and benefits. The incentive mechanisms for knowledge discovery and invention under the AI paradigm will therefore differ from those of the scientific paradigm. They will be algorithmic and mathematical rather than dependent on narrow human scientific communities.

C@C

(view original post →)