Request for Information: Pathways to AI-enabled research

Transforming scientific work with tools and technology

Technology has repeatedly demonstrated its ability to transform scientific discovery and the scientific process. Throughout human history, technological advancements — from printing, to computers and the internet, to modern scientific software — redefined each step of the research cycle. Research software, in particular, has reshaped scientific work and computational reproducibility over the last two decades. Applied thoughtfully, we believe artificial intelligence now has the power to catalyze a new generation of tools for science and ways of doing science. The Open Science program at Chan Zuckerberg Initiative (CZI) and Alfred P. Sloan Foundation Technology Program (Sloan) are requesting community input about existing applications of AI to transform scientific work that are consistent with open research practices.

Through years-long programs like Better Software for Science and Essential Open Source Software for Science, both organizations have invested in open source technology and computational tools that fuel scientific discovery, as well as in practices that support the production, maintenance, and broader adoption of research software. Both funders are also conducting user research to understand the uptake of generative AI in different areas of science, such as CZI’s collaboration with Ithaka S+R to survey biomedical researchers’ perceptions and practices with generative AI. As new AI models, methods, and applications to drive domain-specific scientific discovery are built (such as the development of foundational models to explore the molecular underpinnings of human health and disease), we also expect the future of scientific work and reproducible research to be impacted and transformed by these technologies.

To address this rapidly developing landscape and inform our programmatic priorities, CZI and Sloan are soliciting information from the community of scientific tool developers and practitioners via a public Request for Information (RFI).

This RFI aims to identify existing tools that harness AI to:

• increase reproducibility and transparency in the research process
• offset the cost inherent in participating in open research practices, or more generally increase the productivity of scientists
• facilitate the re-use of resources, thereby enhancing the efficiency of research practices
• expand access and reduce barriers to participation in research, particularly for members of underrepresented communities, thereby increasing the diversity and inclusion of participating researchers

We are interested in learning about tangible tools and applications for AI-enabled research practices, with an eye toward principles of open science. Focus areas for this RFI include (but are not limited to) the use of AI in:

• conducting exploratory data analysis
• the development and maintenance of open source research software
• creating and testing reproducible workflows
• querying and aggregating complex, multimodal data from repositories
• reviewing and summarizing research results
• drafting narrative summaries from existing experimental results
• extracting research entities from text and recommending missing identifiers
• sharing and curating documentation and metadata for protocols and methods, code, and research data

This list is not exhaustive — we hope to learn about applications of AI beyond that which we’ve already imagined. Regardless of the type of application, the AI tool submitted should increase access to scientific resources to encourage facilitation of a diverse, equitable, and inclusive global scientific community. We are especially interested in approaches that anticipate and define measures to mitigate unintended and harmful uses and risks of this technology.

This is a Request for Information (RFI), not a Request for Proposals or Funding Applications.

Please note that this RFI is not tied to an existing funding program at CZI or Sloan. Read “Frequently Asked Questions” below for more information on how we will use submissions.

Submitting a response

Submissions are due by May 30, 2024 at 17:00 Pacific Time. The RFI form requests the following information:

• Title of submission
• Contact information about the submitter (name, email, affiliation)
• Describe the AI-enabled tool or application (200 words maximum)
• What is the value proposition and/or potential outcomes of your AI-enabled tool for facilitating the scientific process? (200 words maximum)
• How will the tool support open science and/or expand access to the scientific process? (200 words maximum)
• (optional) How does the tool mitigate harmful uses or risks associated with the technology? (200 words maximum)
• (optional) Progress made to date (200 words maximum)
• (optional) Additional links to relevant materials

Confidentiality and Privacy

CZI and Sloan will independently process your data in accordance with their respective policies (for example, see CZI’s Privacy Policy). CZI and Sloan may use the information submitted to inform emerging strategies for AI in our organizations. By submitting, you agree NOT to submit confidential, business sensitive, proprietary, or privileged information, or information that would infringe upon third-party rights, such as their intellectual property or privacy rights. We may share your contributions with external experts or partners to gain insights into challenges and opportunities. Submitted materials will not be returned to submitters. CZI and Sloan will honor the attribution and intellectual property of ideas submitted to this RFI and may contact the submitter to learn more about the proposed ideas.

Frequently Asked Questions

Is this a request for proposals or applications?

This is not a request for either proposals or applications for funding. This is a request for information that will help CZI and Sloan understand how to best develop program strategies, including future funding priorities.

What types of tools and activities are out of scope for this RFI?

We are not interested at this time in submissions related to ideas that:

• focus exclusively on training and skill development for AI in science — the focus of this RFI is on tools and applications, not AI literacy or capacity building for scientists.
• address specific scientific research questions or narrow disciplinary contexts — while domain-specific examples are welcome in the submission, we want to learn about tools and applications that address generalizable use cases and workflows in the research process.
• relate to the summarization, synthesis, and discoverability of the scholarly literature — we are interested in use cases that are upstream of the scholarly dissemination process. Tools and applications that relate to published outputs are increasingly served by the publishing and technology industries, thus are less relevant to our strategic considerations in philanthropy.

What does “open” in this call mean? Do all data, models, or application source code associated with a submission need to be publicly available and open licensed?

CZI and Sloan are committed to principles of open science. We recognize the landscape of AI-enabled technology is rapidly developing, with best practices in relation to open science continuing to change. We also understand that some of the most popular models (specifically LLMs) are only open to some degree. Ideas leveraging off-the-shelf LLMs and generative AI services are in scope for this call, as long as they explain how the proposed tool or application supports open science and/or helps expand access to the scientific process.

Who can submit?

Any person, organization, institution, or group can submit a response. We welcome submissions from for-profits, nonprofits, labs, and academic institutions.

Can I submit more than one response?

Yes, you are welcome to submit multiple responses to the RFI for discrete tools or applications.

When are submissions due?

All responses to the RFI must be submitted by May 30, 2024 at 17:00 Pacific Time.

Who will see my submission? Who will review the RFI responses? Will my submission be kept confidential?

After the RFI submission period closes, CZI and Sloan staff will compile and review the RFI submissions. There are no “review criteria” for submissions. Instead, we aim to learn from each submission and glean insights from aggregating all submitted responses. We may also engage external experts and partners who can help us better understand and identify key opportunities. Read the “Confidentiality and Privacy” section above for more information about who will have access to submitted responses.

What can I expect after submission? How will you use the information?

Our goal is for the RFI submissions to deepen our understanding of existing tools and applications for AI-enabled research practices. We plan to use RFI responses to inform and shape our program priorities, including potential future funding opportunities. Read the “Confidentiality and Privacy” section above for more information about how submissions will be used.

Do I need to respond to the RFI to qualify for any potential future funding related to these topics?

No. Responding to this RFI is voluntary and will not be a prerequisite of eligibility for any funding opportunities from CZI or Sloan that may arise in the future related to AI.

I have a question that isn’t answered here. Who can I contact for more information?

If you have additional questions about the RFI that are not addressed already, contact openscience@chanzuckerberg.com. Please note that we will not accept submissions via email, only through the RFI form.