Artificial Intelligence-Augmented Machine Learning for Autonomous Scientific Discovery in Interdisciplinary Research

Abstract

When Artificial Intelligence (AI) and Machine Learning (ML) are applied together, they vastly accelerated the process of improvement of numerous fields of science. However, despite the existence of existing systems, even now people have to be involved in the process of hypothesis formulation and adjustment of the models to a great extent. In this paper, a vision of fully autonomous transformation of the AI-augmented ML systems is extrapolated, with the agent being free to hypothesize, simulate, interpret the outcome, and optimize the models independently of the existential guiding presence of human. The trends in genomics, environmental sciences, and quantum chemistry have prompted the creation of the described framework, which is the accomplishment of a new breed of intelligent digital scientists who could increase the speed and reach of the multi-disciplinary research many times over.