Courtesy Announcement

Call for Proposals

As part of OpenAI’s recently announced NextGenAI research partnership, the University of Michigan will launch a collection of new projects advancing the state of art and practice in artificial intelligence, covering a range of topic areas and investigative approaches.

To identify the initial set of projects, we solicit brief proposals for focused research efforts. We will consider projects in all areas of AI, including core technology, innovative applications, and societal implications. We aim to support projects covering a range of goals, from a range of disciplines, including extensions of current research threads as well as entirely new directions. Selected projects will be granted general research funds, funds specifically for computation, and credits for access of OpenAI models and tools.

Eligibility

All tenure-track and research-track faculty of the University of Michigan—Ann Arbor are eligible to serve as principal investigator. Each project must designate one PI and optionally additional co-PIs. A faculty member may participate in at most one proposal submitted in response to this Call.

Topics

This program is open to proposals for research to advance AI, in any of several ways:

Core technology. Techniques that improve fundamental AI capabilities, in terms of performance, reliability, or cost-effectiveness. New methods for evaluating advanced AI algorithms and systems.

Innovative applications. Exploiting cutting-edge AI models or techniques to solve challenging problems, develop impactful services, amplify creative processes, or otherwise demonstrate novel abilities.

Societal implications. Understanding social and economic impacts of AI developments. Establishing practices and creating systems that promote development of AI that is safe and beneficial. 

Project Size and Scope

This program will support 12-month projects in two size categories: small (up to $50K in general research funds) and medium (up to $100K). These figures include direct and indirect costs (56% IDC rate), per standard U-M policy. We anticipate selecting five small and two medium projects in the first stage of this program. Allocation of computation funds and OpenAI access credits will be commensurate with project needs. Expected start date for projects is the beginning of fall term 2025.

Selection Criteria and Process

Proposals will comprise brief statements of the goals and approach of the project, with additional details on faculty investigators, computational and model access needs, and budget. Proposals must be submitted via this form by 24 April 2025.

Submitted and eligible proposals will be reviewed by an internal committee of U-M faculty and staff. We expect to receive a large number of high-quality proposals. The review committee will assemble a short list of finalists from among these, and enter discussion with OpenAI representatives about the short-listed proposals. During this stage we may seek further information from finalist PIs. Selection of the portfolio of funded projects will be a joint decision of OpenAI and the U-M review committee. We intend to provide provisional notifications by 6 June 2025.

Proposals will be evaluated for their potential impact on AI-related knowledge and practice, novelty of approach, and their prospects for demonstrating advancements within the project horizon. We will also assess how well they leverage OpenAI collaboration, including access to frontier AI models. The overall selection is of a portfolio, so will expressly attempt to include projects showcasing a breadth of topics and approaches.

Questions? Contact:
Michael Wellman 
Lynn A. Conway Collegiate Professor of Computer Science & Engineering
Research: Strategic Reasoning Group