Artificial intelligence (AI) is used to help write emails and create documents for several purposes globally. Since the user-base has grown exponentially, the ability for large-language models (LLMs) like ChatGPT and Claude to learn from user preferences has grown as well . Interestingly, this learning doesn’t start and end with emails. In fact, 84% of scientists claim to use artificial intelligence tools in their work (Geena De Rose, WILEY).
Advancements in the artificial intelligence science space have been growing enormously. Earlier this year, researchers published their results: creation of “The AI Scientist”. An artificial intelligence tool they describe as able to “create research ideas, write code, run experiments, plot and analyse data, write the entire scientific manuscript, and perform its own peer review.” This AI scientist was shown to be very thorough, with its generated manuscript passing the first round of peer review at a top tier machine learning conference (Lu et al., 2026).
The idea of this artificial scientist is not the first of its kind. Many creators of artificial intelligence are coming up with their own science specialized software (for example, Claude Science and Benchsci’s EMET). These scientific tools are being adopted more and more openly, as artificial intelligence finds its scientific niche, but what is important to remember is that these scientific tools can inherit the bias its creators harbor. For example, a healthcare risk-score algorithm was found to apply race-based adjustments to the risk scores, systematically underestimating how sick black patients were as compared to their white counterparts. Luckily, there are several prominent figures rising to combat this injustice. Dr. Joy Buolamwini founded the “Algorithmic Justice League” to hold tech companies accountable for racial and gender discrimination found built into AI. While AI’s future is constantly evolving, what is undoubtedly true is that it will continue to change the landscape of science and scientists. What we must also consider is how our own internal biases may influence the devices, or new technologies that we create.
