AI, AI, Oh! Task Force Prepping Medical School Community to Adapt to New Tools
This sentence, or any part of this article for that matter, easily could have been written by artificial intelligence (AI) tools like U-M GPT or U-M Maizey.
How tools like U-M GPT, U-M Maizey and many others are being used – and how AI can be transformative in enhancing research, teaching and learning, reducing administrative burden, and preparing learners for future challenges in health care – centers the work of a new Medical School Generative AI (GenAI) Task Force.
The task force is working to forge guiding principles to direct the medical school’s approach to GenAI across the research, education and academic administrative missions for all faculty, staff and learners. It will also guide the specification of new IT infrastructures, which are allowed by Michigan Medicine Information Assurance and Compliance offices, and are consistent with best practices of data use.
‘Beyond human intelligence’
The medical school must focus on the ethical and responsible use of AI, while preserving security and privacy – and it needs a realistic plan to determine which tools faculty, staff and learners are allowed to use, according to Brian D. Athey, Ph.D., chair of the Department of Computational Medicine and Bioinformatics. Athey co-chairs the task force with Matthew C. Comstock, M.B.A., MHSA, medical school chief operating officer; and Joan A. Keiser, Ph.D., a research strategy advisor to the Medical School Office of Research.
“What is really fascinating is that, in just a couple of years, we will have tools and capabilities that are going to be beyond human intelligence. It is happening fast, and it is absolutely astonishing,” Athey said. “There’s no question that these tools and capabilities are going to augment and run alongside our conventional (teaching and research) methods, and what we want to do is explore the potential depth of their uses — safely and securely.”
The task force’s work will continue through the end of 2024 and ultimately produce a series of reports to help guide leadership on how best to adopt and to adapt to this new reality.
“There are so many near-term opportunities for deploying Generative AI tools across our academic missions; however, the potential ethical, safety, security and privacy issues these resources carry along are daunting,” Keiser said. “We plan to provide our school leadership with a set of guiding principles for future decision-making and a roadmap to get us started on this journey.”
Added Comstock: “We are excited to come together to address GenAI — an area of critical importance for our future. We are grateful to all of the co-chairs, champions and task force members who are studying in this area with a goal of providing much-needed guidance around GenAI for the medical school community.”
Potential enhancements
Several champions are guiding working groups in key areas: research; education; academic administrative systems; and IT infrastructure, IA and compliance, and data.
As these groups continue their work, they do so acknowledging that while GenAI is still in its infancy — and many questions remain about how it will be utilized long-term — the medical school can benefit from its use in enhancing learning and personalization; improving diagnostic and clinical skills; streamlining administrative processes; addressing knowledge gaps; and preparing for the future of medicine.
Senior Associate Dean for Education Louito C. Edje, M.D., MHPE, FAAFP, a champion of the working group studying generative AI in the educational setting, said it is critical that Michigan Medicine embrace AI and the opportunities it allows for learners at all levels.
“The dictum in medicine, ‘First do no harm,’ typically refers to patients. For me, a medical educator, this dictum also applies to learners,” Edje said. “As our learners move forward in their professional identity formation and other learning, it would be harmful for us to stand still and not take advantage of all the educational opportunities in the rapidly growing generative AI arena.”
Incorporating the tools
Diana M. Witowski, M.B.A., associate chief financial officer for the medical school and a champion of the academic administrative systems workgroup, said the school must continually identify and assess emerging technologies, like AI, for opportunities, risks, impacts and limitations in the service of achieving its missions.
“Generative AI is currently in a highly dynamic phase, where high expectations are being put to the test in real-world scenarios. Competitors are racing to bring innovations to market, and established companies are integrating these innovations into their existing products,” she said. “Our ongoing challenge is identifying what will change and what will remain constant.”
There is a knowledge article publicly available from HITS, “Using Artificial Intelligence (AI) Tools at Michigan Medicine,” that provides an overview of how to incorporate AI tools using established processes by working with HITS or another Trusted IT Service Provider.
The article also includes links to the Guideline on Generative Artificial Intelligence Tools and Michigan Medicine Appropriate Clinical Use of Generative Artificial Intelligence Tools Policy, as well as a helpful infographic on using AI securely. It highlights the necessity to be vigilant with data input into AI systems, taking measures for using AI safely, and everyone’s responsibility to adhere to rules about data protection, copyright and liability.
(This article is from Michigan Medicine Headlines)