Reproducibility and scientific interpretation in the age of AI: consilience in biological systematics, ecology, and molecular biology

TL;DR

This paper discusses the challenges of reproducibility in biological sciences and explores how AI can enhance scientific validity and interpretation through integrated, transparent approaches.

Contribution

It highlights the role of AI in improving reproducibility and interpretation in biology, emphasizing methodological rigor and dataset quality.

Findings

AI offers tools to mitigate bias and analytical variability

Reproducibility remains challenging due to interpretative choices

AI can reinforce scientific validity without hindering progress

Abstract

Achieving complete reproducibility in science, particularly in research fields such as biodiversity, is challenging due to analytical choices, bias and interpretation. Here, we examine examples of reproducibility in biological systematics, ecology, and molecular biology. To mitigate the impact of interpretation and analytical choices, Artificial Intelligence (AI) has provided potential tools. In the present work, while emphasizing the need for methodological rigor and transparency, we acknowledge the role of interpretation in activities such as coding biological characters and detecting morphological patterns in nature. We explore the opportunities and limitations associated with the synergy between big data and AI in molecular biology, emphasizing the need for a more comprehensive and integrated approach based on dataset quality and usefulness. We conclude by advocating for AI-based tools to assist biologists, reinforcing consilience as a criterion for scientific validity without hindering scientific progress.