By Sylvester Thompson
Dr Toyin Philips, Artificial Intelligence (AI) research expert, says AI is changing scientific practice through its tools, which support literature mining and meta-analyses.
Philips, who spoke in an interview with the News Agency of Nigeria (NAN) on Monday in Abuja, said this has been achieved by screening and summarising thousands of papers which helps researchers connect findings across disciplines and spot emerging trends.
NAN reports that literature mining is a specialised form of text mining that focuses on extracting meaningful information and discovering hidden relationships from large volumes of unstructured scientific publications and books.
Meta-analysis is a statistical method that combines results from multiple independent studies on a similar topic to produce a single, more powerful conclusion.
According to Philips, studies of publication data reveal that AI methods are employed across nearly all scientific domains, and are associated with more highly cited papers, indicating a shift in what constitutes cutting-edge methodology.
She stated that AI enhances faster data analysis, as AI systems can rapidly process and interpret massive datasets, revealing correlations and structures that are difficult or impossible to detect by hand.
“In genomics and biomedical research, machine learning models scan DNA sequencing or cell images to identify disease markers and early cellular changes, enabling earlier diagnosis and more targeted therapies,” she said.
Philips, who is Director of Research at Moonbeam Technologies, an AI firm specialising in general artificial intelligence solutions, noted that AI is increasingly used to generate hypothesis.
“In contemporary society, AI is being used to explore chemical or material spaces, and design experiments, expanding what is practically testable in the laboratory.
“Robot-AI systems can autonomously run and refine experiments in chemistry and materials science, learning which conditions are most promising and accelerating discovery cycles by large factors,” she said.
The expert noted that in fields where real-world experiments are expensive or infeasible, such as climate sciences, astrophysics, and particle physics, AI-driven models improve simulations and forecasts of complex systems.
She said AI helps correct biases and errors in existing physical models, citing example such as weather prediction, which leads to more accurate, actionable forecasts that can directly affect policy and safety decisions.
On risks and ethical challenges, Philips noted that because AI systems learn from existing data, they can inherit and amplify biases, which is a serious concern in areas like medicine, social science, and environmental policy. (NAN)(www.nannews.ng)
Edited by Ekemini Ladejobi
