Autocatalytic knowledge dynamics in the AI era
In the AI era, knowledge is growing at an accelerated rate. The goal of this study is to construct the macrodynamics of knowledge and elucidate the role of AI in this process. An autonomous knowledge production system is examined. The driving force behind the system's development is the disturbance of its homeostasis under the impact of external or internal factors. To solve emerging problems, the system uses available or generates new knowledge. The process is autocatalytic in the sense that knowledge accumulated over the past is used to create new knowledge. Knowledge production is hampered by resource scarcity and environmental disturbances that worsen the system's state. The loss of knowledge, in particular due to its obsolescence, contributes to the process' slowdown. Information noise also inhibits the advancement of knowledge, leading to energy dissipation. The combination of autocatalysis, aimed at generating new knowledge, with counteracting forces caused by resource scarcity, environmental problems, knowledge loss, and information noise provides the basis for building knowledge dynamics. The resulting dynamic equation shows that, depending on the balance of the forces acting, the rate of knowledge production can change as follows: 1) increase, ensuring system's development; 2) decrease to a finite (non-zero) level: the system degrades but survives; and 3) decline to zero: the system collapses. The model was applied to WIPO patent filing data, yielding parameter values. The influence of AI enhances autocatalysis, but counteracting factors also increase. A shift in the balance of forces leads to a phase transition with a change in knowledge dynamics. The system's evolution can proceed through a series of phase transitions. Key milestones in AI development are compared with the forces represented in knowledge dynamics. Barriers to knowledge growth are discussed.