Three ways Generative AI can give Research Managers “superpowers”

Author: Saskia Hoving

As with almost every other industry, generative AI is causing a storm within the research community. In today's fast-paced environment, AI offers powerful new opportunities to those who work in a research environment. However, the extent of these opportunities is still unclear, with many unsure about how AI can help them, and how it can be used responsibly. In this blog - the first in our new series outlining potential AI use cases for research professionals - we explore the significant ways AI can enhance the research management profession, from producing concise research summaries to handling repetitive tasks.

1. Exploring possibilities and experimenting with AI

Research managers can harness AI tools in various ways to enhance their work. These examples are some of the low-hanging fruit that some research managers have gotten started with, even with off-the-shelf products and services.

• Seeing patterns in the noise
  One of the main engines for generative AI are systems called generative pretrained transformers (GPT). They work in part by learning patterns in the dataset on which they’re trained and using patterns to make predictions. Those “predictions” include text output (predicting what word comes after another word), images (based on pixel patterns), etc. — this is where the output comes from.
  
  This makes GPTs very powerful at seeing patterns in a mass of data that human eyes tend to miss, which. can help research managers make more informed and better decisions. For example, research managers can use AI to better understand their historical data, to find connections between past successful projects to inform future resource allocation. Another use-case could be analysing the ways previous grant funding has been used, and applying these insights when assembling evidence for new funding applications, or for maintaining and increasing existing grants.

• Synthesising and summarising research
  Research offices can also use AI to get plain English distillations and summaries across a range of research sources. This can help make executive summaries, cross-output “abstracts,” and other shorter summaries. Research managers and research offices can then put together concrete, but easy to understand, “bottom lines” of research effectiveness that stakeholders like Trustees can quickly grasp.

• Producing first drafts, converting research into plain language
  While AI text output often doesn’t meet standards for final drafts, research office staff can use it to generate first drafts of things like reports, grant proposals, presentations, and more. AI can be especially helpful in transforming data into plain language, saving significant time with analysis and ideation, so staff can focus on what needs to be analysed and reported on, rather than creating the raw first draft. It’s then easier to edit, rewrite, and refine from these first drafts. AI can also rewrite specialist research — including articles, book chapters, or any other research output — into plain language, so that academics from other disciplines or the wider public can understand it.

• Automating repetitive tasks
  As with many other industries, research managers can use AI to help with a range of routine and menial tasks, from data entry to email writing. This frees up time to work on higher-level tasks that require greater creativity, imagination, and new thinking. It is even possible to get access to the underlying technology via licensing arrangements, allowing research offices to build bespoke tools to fit a particular institution, organization, or department. This can help leverage these tools to fit specific administration workflows.

2. Making AI work for everyone

Research offices can also play central roles in AI use across the research enterprise. That means doing things like:

• Ensuring all teams across the institution have access to the kinds of AI tools that they’ll benefit from.
• Training all teams — particularly those wary of the technology — in how to get AI to work for them, and how to use it responsibly.
• Developing an approach for your institution’s responsible use of AI.

AI tools can help support everyone in the research enterprise. If these tools can help research managers translate data and generate first drafts for their reports, then clearly it can help researchers do the same; but only if they have the training in how to use them best.  AI tools and services are not yet at the point where they can run on their own — and in fact might never be - and research managers have an opportunity to really be at the forefront of how humans and technology can work together to benefit the research enterprise.

3. Empowering and shaping policy for a responsible future

As with most new technologies, AI has risks that need watching and mitigating. As research managers experiment with AI, and encourage researchers to do the same, they should also strive to better understand these risks and educate others on best practices.

This could include providing guidance on using “free” software, which often comes with data security risks, or setting up frameworks for human checking of AI-generated information to stay on top of any potential biases or inaccuracy.

Making use of AI tools effectively and responsibly can help research managers extend their capabilities — like a lever can help move a heavy weight — so they can focus more on the creative, innovative, and human aspects of research. 

The changing role of the researcher manager

Leveraging and integrating AI into the research office is just one part of how the research office is changing. Read more about other changes in Springer Nature’s white paper, “Global Perspectives on Research Management,” and also about the growing role of collaboration in research management in the white paper, “A Changing Landscape in Collaboration.”

Related content

• Navigating the evolving role of research managers: Insights from INORMS Council members
• Global research collaboration strategies that support researchers
• The Continuing Rise of Collaboration
• AI and Science Publishing: “Cutting through the clutter has never been more important”
• Can AI help us manage information overload?

Don't miss the latest news & blogs, subscribe to The Link Alerts!