Why AI is not replacing scientists

AI is transforming how scientists tackle some of the world's most complex challenges. As these technologies become more powerful, the role of scientists is becoming even more important.

Waterloo Science+AI postdoctoral fellows Mohsen Rezaeian, Lucie Delobel and Amer Marwan El-Samman are working at the forefront of AI-enabled research across health, environmental and physical sciences. They discuss why AI is not replacing scientists, what students need to succeed in an AI era, and the opportunities and challenges these tools bring to scientific discovery.

To thrive in an AI era, students will need to be AI-literate. This doesn’t mean using AI for everything. The basis of scientific discovery is formed on asking questions and designing experiments to answer them. This role of a scientist will never change. That is why teaching students how to use AI as a tool without depending on it is important.

Students can use AI to analyze information, identify patterns, and explore potential research directions. However, they are taught to form their own hypotheses, design their own experiments, assess evidence and develop their own conclusions using critical thinking and problem-solving skills, backed by a science education. In Waterloo Science, students will graduate knowing how to responsibly use AI with a guiding set of ethical principles taught here at Waterloo.  

AI is accelerating the pace of scientific research, making information more accessible than ever. However, with greater access to information comes more responsibility. There is a quote that feels especially timely in the age of AI: “The formulation of a problem is often more essential than its solution.”

AI does not replace scientific understanding. While AI can quickly generate data, understanding why something is happening is a uniquely human experience that comes from critical and creative thinking. Human qualities such as questioning assumptions, challenging conventional approaches and exploring entirely new perspectives are essential for scientific progress.

Rather than replacing scientists, AI is changing what it’s like to work in science. By automating repetitive tasks such as data processing, it opens more time for researchers to develop hypotheses, design experiments, and interpret results.

Soft skills such as leadership, collaboration, and the ability to explain complex topics remain essential. While in school, students should focus on gaining hands-on experiences in the lab and in the field to show that they have developed skills that can’t be automated. Those who graduate with the ability to use AI as a tool in their scientific work will be competitive candidates in the workforce.  

Relying too much on AI is a big concern in the scientific community. One concern is that if students overuse AI, they may miss critical learning experiences that come from making mistakes, working through complex concepts, and solving problems. It could also narrow the range of questions scientists explore. Scientific work is more about understanding how and why an answer was formed than the answer itself, asking original questions and investigating problems where data is limited.

AI can also be biased and misinformed, producing inaccurate or misleading information from outdated facts, societal or cultural biases, and errors present in the datasets it is trained on. AI systems are also highly prone to sycophancy, meaning they will tell the user they’re right even when they’re wrong or misinformed.

Scientists have a responsibility to treat AI responses as a starting point rather than a final source. We teach that AI is a generative tool, not an archive of absolute truth. Our human expertise is essential for verification, cross-referencing, and ethical oversight to prevent misinformation.

The Faculty of Science is advancing learning and research initiatives that equip students and faculty to use AI responsibly. Learn more about AI in Science.