March 2024

Dear alumni, friends, and supporters,

As one of the nation’s preeminent public research universities, the University of Michigan is continually focused on how best to leverage cutting-edge technologies to tackle bold research challenges, provide exemplary health care, and educate the next generation of Leaders and Best. Fewer technologies have greater potential to both improve our work processes and present significant risks than generative artificial intelligence.

GenAI, which includes models such as ChatGPT for text and DALL-E for images, refers to technologies that can generate new content or data that are similar to but distinct from the data it was trained on. At the simplest level, GenAI models tasks that typically require human intelligence. It can learn patterns, structures, and features from input data and create content with similar characteristics. The technology can be used to craft text, music, and images that mimic human creation. 

For a quick example, here is what U-M GPT, our own AI model, produced when prompted to write a haiku about the Michigan Wolverines football team:

Autumn leaves descend,
Maize and blue warriors clash,
Hail to victors bold.

It won’t be winning any poetry contests, but it’s not bad for an instantaneous production. To help navigate the unprecedented scale and speed of this technology, we established the Generative Artificial Intelligence Advisory Committee. This group – comprising faculty, staff, and students across all segments of U-M – is assessing the opportunities and challenges, as well as advising campus leadership and the wider community on how U-M can navigate responsible technological development and integration. 

A goal of innovation and integrity

This video features Ravi Pendse, VP of information technology and chief information officer, explaining U-M’s adoption of GenAI technologies.

Harnessing the tech

The committee’s initial recommendations included establishing a University-wide initiative to leverage GenAI in developing tools and methodologies for AI-augmented education and research. The committee also focused on best-practice standards to ensure privacy protection, data-use controls, and research integrity when using GenAI.

To implement these recommendations, we have begun making strategic investments to harness the technology for our campus community and be a leader not only in the field but also in its ethical use.

We collaborated with Microsoft to create our ITS AI Services, which launched last year and enabled everyone at U-M to access and experiment with an innovative suite of GenAI tools, including U-M GPT. While many higher education institutions are offering GenAI education programs, U-M is believed to be the first major university in the world to offer a custom AI platform for its entire community.

Our faculty is developing more than 35 short online courses designed to build learners’ GenAI skills and competencies. Most of the courses – including “Generative AI in Business: A Tactical and Strategic Guide” and “AI for Lawyers” – will launch in July.

At Michigan Medicine, doctors and researchers are using artificial intelligence to help diagnose brain tumors, generate new antiviral molecules, and analyze cardiac signals to better predict adverse cardiac events.

Used ethically, generative artificial intelligence is a tool that has the potential to accelerate curriculum development for faculty, learning for students, and research for the institution. At the University of Michigan, we are well-positioned to harness this transformative technology while ensuring a continued commitment to fairness, access, and equity.

Sincerely,
Santa J. Ono
President, University of Michigan