On scientific understanding with artificial intelligence

Mario Krenn,Robert Pollice,Si Yue Guo,Matteo Aldeghi,Alba Cervera-Lierta,Pascal Friederich,Gabriel dos Passos Gomes,Florian Häse,Adrian Jinich,AkshatKumar Nigam,Zhenpeng Yao,Alán Aspuru-Guzik

DOI: https://doi.org/10.1038/s42254-022-00518-3

2022-04-04

Abstract:Imagine an oracle that correctly predicts the outcome of every particle physics experiment, the products of every chemical reaction, or the function of every protein. Such an oracle would revolutionize science and technology as we know them. However, as scientists, we would not be satisfied with the oracle itself. We want more. We want to comprehend how the oracle conceived these predictions. This feat, denoted as scientific understanding, has frequently been recognized as the essential aim of science. Now, the ever-growing power of computers and artificial intelligence poses one ultimate question: How can advanced artificial systems contribute to scientific understanding or achieve it autonomously? We are convinced that this is not a mere technical question but lies at the core of science. Therefore, here we set out to answer where we are and where we can go from here. We first seek advice from the philosophy of science to understand scientific understanding. Then we review the current state of the art, both from literature and by collecting dozens of anecdotes from scientists about how they acquired new conceptual understanding with the help of computers. Those combined insights help us to define three dimensions of android-assisted scientific understanding: The android as a I) computational microscope, II) resource of inspiration and the ultimate, not yet existent III) agent of understanding. For each dimension, we explain new avenues to push beyond the status quo and unleash the full power of artificial intelligence's contribution to the central aim of science. We hope our perspective inspires and focuses research towards androids that get new scientific understanding and ultimately bring us closer to true artificial scientists.

Computers and Society,Machine Learning,Chemical Physics

What problem does this paper attempt to address?

The paper aims to explore how artificial intelligence systems can facilitate scientific understanding and defines three fundamental roles that AI might play in this process.

1. **Computational Microscope**: AI as a computational tool reveals phenomena that are difficult or impossible to observe directly in experiments. For example, molecular dynamics simulations can uncover new biological functions of coronaviruses, altering our understanding of carbohydrates in biological systems and inspiring new analytical methods that can apply these insights without the need for complete computations.

2. **Source of Inspiration**: AI as a source of inspiration expands human imagination and creativity by generating novel concepts or ideas. For instance, automatically identifying anomalies in scientific research literature or discovering surprising concepts through the interpretation of existing models. This type of work helps scientists form new hypotheses or understandings that can be applied in different contexts.

3. **Understanding Subject**: This is the ultimate goal of future development, where computers themselves become the subjects of understanding, capable of inducting concepts from observations and applying them to new phenomena without human intervention. Although there is currently no evidence that machines can independently achieve true scientific understanding, the paper proposes potential pathways to achieve this goal.

In summary, the main purpose of this paper is to re-emphasize the importance of scientific understanding and to inspire physicists, chemists, biologists, and AI researchers to go beyond the status quo and focus on one of the core goals of science—scientific understanding, thereby driving revolutionary advancements in computer-assisted scientific understanding.