Artificial intelligence and pharmacologists: threat or opportunity?

With some marketing coups, artificial intelligence (AI) has entered (for now) the lives of many mind workers. Among them: pharmacologists and biomedical researchers in general. AI is multifaceted and includes, for example, machine learning and data mining. In this regard, we should expect AI to greatly aid pharmacological research and facilitate the discovery of new drugs and new mechanisms of action, thereby easing and accelerating the overall pipeline. It is easy to echo the excitement that exists in radiology, where AI is both welcomed and feared.
However, conducting experiments and then writing or reviewing a scientific manuscript entails responsibilities that only humans can assume. The critical thinking and evaluation required to develop new directions in research and peer review are outside the realm of generative AI and AI-enabled technologies, and there is a risk that the technology will lead to incorrect, incomplete, or biased conclusions. With respect to scientific publications, these considerations, together with the principle that submitted manuscripts should be treated as confidential documents, underlie the Generative AI guidelines for reviewers and editors. Indeed, it is tempting to upload the manuscript or portions of it to a generative AI tool because there is no guarantee of where the material will be sent, stored, or viewed, or how it may be used to train the model in the future, and this may violate the confidentiality, property, and/or privacy rights of the authors.

Download the pdf below to read more