0
On April 21, 2026, Reed City University’s School of Natural and Applied Sciences (FONAS) held its annual faculty lecture at the International Conference Center, featuring timely and nationally relevant talks on the intersection of artificial intelligence and scientific responsibility. The event attracted a distinguished and diverse audience, including academics, students, visiting scholars, professors emeritus, and the general public from across the university community. Lectures are now also accessible virtually, allowing for wider participation beyond the physical venue. The keynote address was delivered by Professor OFW Onifade of the University of Ibadan, who spoke on the theme “When Machines Learn Science: Ethical Limits of Machine Learning in Science.” His presentation addressed central and increasingly urgent questions in contemporary research. It was about where machines stop and scientists start, and especially what are the ethical issues and concerns regarding the use of machine learning and artificial intelligence in global science.
In a comprehensive and thought-provoking commentary, Professor Onifade traced the rapid integration of machine learning into virtually all areas of scientific research. From genomics and biomedical research to environmental monitoring, data analysis and presentation, toxicological analysis, and advanced physics, machine learning systems are now helping to process complex datasets, identify patterns, and even propose hypotheses.
He pointed out that the scientific method itself is undergoing subtle but significant changes. Modern science, traditionally based on human-driven observation and experimentation, is increasingly shaped by algorithmic processes that can reveal relationships at previously unimaginable scales. While Professor Onifard acknowledged that these advances were a major step forward, he cautioned that increased reliance on machine learning tools must be accompanied by a corresponding level of critical scrutiny. He argued that accelerated discovery should not come at the expense of conceptual understanding.
A central pillar of the lecture was the ethical aspects of machine learning in science. Professor Onifade emphasized that, contrary to popular perception, algorithms are not inherently neutral. They are formed by the data on which they were trained and by the developers’ assumptions and design choices.
This reality poses a risk of bias, sometimes subtle but with potentially significant consequences, especially in sensitive areas such as medicine, environmental policy, and public health research. For example, biased datasets can lead to flawed predictions, misinformed interventions, and unfair outcomes.
The talk raised important questions about accountability in an era where decision-making is increasingly mediated by intelligent systems. Professor Onifade asserted that responsibility for scientific achievements cannot be transferred to machines. Rather, it remains the duty of scientists to examine, verify, and interpret the output of algorithms within an appropriate ethical framework.
Another major concern raised during the lecture was the challenge to reproducibility, which is the basis of reliable scientific practice. The use of complex machine learning models poses a significant barrier to transparency, as many operate as opaque “black boxes.”
If the inner workings of a model cannot be easily explained or reproduced, it becomes difficult for other researchers to validate the results. This not only undermines the credibility of individual studies, but also poses a broader threat to public trust in science.
Professor Onifade called for a more transparent and interpretable model and the adoption of strict documentation standards. He emphasized that reproducibility must remain a non-negotiable principle even as scientific methodologies evolve.
The impact of these developments is particularly evident in biology, where machine learning is already reshaping research and practice. Applications range from genome sequencing and disease surveillance to antimicrobial resistance research and ecological modeling.
Addressing participants from a variety of scientific backgrounds, Professor Onifade emphasized that while machine learning can reveal patterns in vast biological datasets, it lacks the contextual and mechanistic understanding that human expertise provides. For example, biological systems are inherently complex, and their interpretation requires not only computational insight but also deep domain knowledge. He therefore advocated a balanced approach. It integrates computational tools with traditional scientific reasoning to ensure that conclusions are accurate and meaningful.
Looking to the future, the lecture emphasized the need for a paradigm shift in science education. As machine learning becomes an integral part of research, it becomes increasingly important to provide students with the skills necessary to engage seriously with these technologies.
Professor Onifade emphasized that AI literacy must go beyond technical proficiency. This should include an understanding of the ethical, methodological, and philosophical implications of using intelligent systems in scientific research.
He pointed out that higher education institutions and the National Universities Commission (NUC) have a vital role to play in shaping curricula that prepare the next generation of scientists not only to use machine learning tools, but to question and improve them.
The 2026 Annual Faculty Lecture will make a significant contribution to the ongoing conversation nationally and globally about the role of artificial intelligence in society. The event fostered interdisciplinary dialogue on the future of scientific practice by bringing together a broad audience including academics, students, honorary scholars, and virtual participants. Professor Onifade’s lecture was a powerful reminder that while machine learning has great potential, it also requires careful governance. To maintain the integrity of science, the boundaries between human judgment and machine capabilities must be clearly defined.
Finally, Professor Onifade reiterated that machine learning should not be seen as a replacement for human capabilities, but as a powerful extension of them. The nature of critical thinking, ethical reasoning, and intellectual curiosity in science remains fundamentally human.
As scientific tools become more sophisticated, the responsibilities of scientists increase accordingly. Ensuring that innovations are aligned with ethical standards is critical to maintaining public trust and advancing knowledge for the collective good.
The talk ended with a call to action to embrace technological advances while adhering to the principles that define scientific inquiry. In an age where machines can learn, it is ultimately the responsibility of scientists to ensure that knowledge continues to be guided by wisdom.
Emeka (PhD) writes from the Faculty of Biological Sciences, Lead City University, Ibadan.
