Seminar summary - Biological data have expanded in breadth and depth over the last decade, leading to larger datasets that hold the potential to answer questions at greater resolution. Despite this growth, some datasets remain inherently limited; including many derived from specimens from natural history collections. In this talk, Bryan Carstens will discuss the potential of Artificial Intelligence approaches to improve biological datasets. One example describes the application of transfer learning to improve the genomic data collected from specimens where the sampling is inherently limited, such as from extinct species. A second describes the use of neutral networks to automate pseudo-landmarks in three dimensional geometric morphometrics. Taken together, this work suggests that AI has the potential to enhance biological data, leading to increased inferential capabilities.