Talk Summary:
Advances in computational capacity have dramatically dropped the cost of measurement, allowed orders of magnitude increases in the aggregation of data, and are enabling computer scientists to test the limits of machine learning. Recent advances, such as transformers, are blurring the domains of prediction and creativity, challenging the unique role of human contributor and representing a fundamental turning point in humanity’s relationship with the world. The domain of science, which in the modern age has positioned itself as the extreme upper echelon of human intelligence, is not immune to this new dynamic. The relative role of human and machine in advancing science has been the same for centuries: technological innovations providing new observations, enabling data storage and communication, and performing numerical calculations based on human-generated theory. This has been a largely human-led endeavor. The current pace of developments in the domain of artificial intelligence portrays a future for science that is less human-led and more human-machine cooperative. The foundations for this journey are set and the outcome is inevitable. Science will progress in a human-machine hybrid paradigm wherein hypothesis generation is not solely the domain of humans. This fundamentally is enabled by the scale-up of digital means for capturing cause, effect and confounder associations such as directed-acyclic-graphs (DAGs) (Breiman 2001, Daoud and Dubhashi 2023) and the progression of large language models (LLMs) that support more precise capture of structured human hypothesis articulation. (Johnson et al. 2023) Capturing these hypotheses in a similar form while benefiting from regularization achieved through large language model technology will break down traditional barriers to sharing of hunches. (Johnson 2010) Taken together, these advances will open a period in which cooperative ideation can occur between and among humans and machines at an unprecedented pace. Operating in this environment, scientific discovery will accelerate and open a new period of human advancement.

Presenter Biography:
Dr. David Jaffray is a senior vice president and chief technology and digital officer (CTDO) at The University of Texas MD Anderson Cancer Center and a Full Professor in Radiation Physics and Imaging Physics. Before joining MD Anderson, Dr. Jaffray served as executive vice president for Technology and Innovation at the University Health Network in Toronto. He designed and led UHN’s digital transformation. In the 17 years there, he also served as Head of Medical Physics, vice chair of Research for the University of Toronto’s Department of Radiation Oncology, founding director of the STTARR Innovation Centre, and founding director of the Techna Institute. He was a Full Professor in the Departments of Radiation Oncology, Medical Biophysics, and IBBME at the University of Toronto and was active in strategic planning, teaching, and graduate student supervision. Dr. Jaffray holds 47 patents and has authored >300 peer-reviewed publications in topics related to cancer, including, the development of new radiation treatment machines, exploring the fundamental limits of imaging system performance, the development of novel nanoparticle formulations for improved detection of cancer, and challenges in global health. He has received many honors, including the Sylvia Sorkin-Greenfield Award, the Farrington Daniels Award and the Sylvia Fedoruk Award. In 2018, he received the Gold Medal from the American Society for Radiation Oncology and was invited to join the US National Academy of Inventors in 2022. Dr. Jaffray has led the development of a variety of commercial products, including software and hardware for safe, high-quality cancer care and including the development of cone-beam CT guided radiation therapy. Dr. Jaffray earned his B.Sc. in physics from the University of Alberta, and his Ph.D. in medical biophysics from the University of Western Ontario. He is also Board Certified in the discipline of Medical Physics by the American Board of Medical Physics.

Author