Provost’s Advisory Council on Artificial Intelligence (PAC-AI)

The Council will provide advice on strategic AI matters including, but not limited to, those described below. As the name implies, PAC-AI will not be charged with performing any oversight, coordination, or executive responsibilities.

Members

  • Tracy Shinn, Associate Vice Chancellor for IT and Chief Information Officer, Chair  
  • William ("Bill") Allison, Chief Technology Officer (liaison to the Technology Foundations Committee) 
  • Trevor Darrell, Professor, Department of Electrical Engineering and Computer Sciences 
  • John DeNero, Associate Teaching Professor, Department of Electrical Engineering and Computer Sciences 
  • Salwa Ismail, Associate University Librarian for Information Technology and Digital Initiatives 
  • Ken Lutz, Director of Research IT 
  • Massimo Mazzotti, Professor, Department of History
  • Zach Pardos, Associate Professor, School of Education 
  • Eugene Whitlock, Chief People & Culture Officer and Associate Vice Chancellor for Human Resources

Charge

PAC-AI will provide thought leadership in a broad range of issues, such as the following: 

Instruction 

  • How can AI be integrated into instruction to improve pedagogy?
  • How can AI help increase access to courses and instruction?
  • What are the best ways for students to use AI as an aid to learning?
  • What uses of AI should not be permitted to ensure integrity and appropriate learning? 

Research

  • What kinds of investment in technology and training are needed to ensure researchers are able to take appropriate advantage of advances in AI in their research?
  • What ethical guidelines, beyond those that already exist for research, are necessary to govern the use of AI in research? 

Operations

  • How can AI help automate the operations of the University to make them more efficient?
  • How can AI help people navigate our bureaucracy?
  • How can AI help advise people, including students?
  • What are appropriate investments in AI for the above purposes and how do we forecast the ROI on such investments?

Infrastructure

  • What are the principles that might govern the purchase of AI software or subscriptions to AI platforms (e.g., ChatGPT 4)?
  • What are the infrastructure demands associated with the increased use of AI?
  • To what extent does new infrastructure need to be designed to accommodate an increased use of AI? 

Policy

  • How do we ensure that uses of AI are consistent with the needs to ensure access to all members of the campus community and the various consent degrees governing accessibility?
  • What protections are needed to protect privacy and what might privacy concerns be?
  • What uses should be prohibited, tightly circumscribed, and/or tightly governed (e.g., uses of facial recognition software)?
  • What protections are needed to protect intellectual property, both of University affiliates and that of non-affiliates?
  • What sharing of our data should we permit to allow better training of widely used AI programs, who should govern data sharing, and what remuneration should be sought when sharing data outside the University? 

Envisioning the Future

  • Much of the discussion of AI (including above) is premised on AI doing what humans now do. What can AI allow us to do that we can’t currently do because it is currently infeasible technically or economically and how might we identify those things?
  • The greatest productivity gains from new technology are often because it allows us to do things in very different ways than the existing technology permitted. How can we re-envision processes to achieve transformational gains? 

View charge letter - May 21, 2024