Master the R Series Bootcamp - October 17, 24, 31 and November 7. Registration is now open.

Tips for Using AI in Grantwriting

By Bouvier Grant Group

We stay current on NIH happenings and would be delighted to keep you informed.

Guest Post by Becky Miro

In February 2024, we brought you a first installment regarding AI and grantwriting courtesy of colleague and guest writer Dr. Agnella Izzo Matic.

As you may imagine, the topic of AI as a whole and, similarly, AI as it relates to grantwriting and scientific writing is evolving.  For a bit of background into what AI is, we point you to Dr. Matic’s post.

In this blog post we’ll focus on highlights and tips gleaned from some recent literature regarding grantwriting and AI. We hope this content gives you realistic expectations for how AI can help your grant writing and where you need to proceed with caution.

For busy researchers or clinicians, the idea of using AI to generate content and ideas can be tempting. AI may also be helpful for researchers whose native language is not English. And, indeed, there are benefits to incorporating the use of AI into the grantwriting process as long as its done with knowledge and awareness of what it can and can’t do.

Seckel et al. (2024) recommends having an initial draft of your application before engaging AI.1 This way, the idea, draft, data, and so on are yours. Najafali et al. (2023) tested ChatGPT’s ability to write the Specific Aims for an NIH R01.2  ChatGPT generated content guided by very specific prompts from the researchers.

Najafail et al. did not discuss the legitimacy of the content generated. As you may have already heard, one of the problems with AI is that it has led to issues of plagiarized content. 

Once you have the initial drafts of the application sections ready, use AI to refine your content. It’s possible that more experienced researchers could more successfully harness the power of AI because they have more experience with grant writing and peer review processes and therefore can input more effective questions/prompts. Not only may a less experienced researcher obtain less effective results from AI, but we argue that they are also missing out on developing necessary skills as a researcher if relying on AI. Having a (human) colleague read your work and provide feedback can bring a wealth of information and possibly build collaboration. 

AI can do more than generate words. AI tools like DALL-E-3 or Midjourney can help generate images. Type in a description of what you have in mind or upload sample images. 

Always check resources used by AI. ChatGPT has been updated to search for recent information, eliminating the knowledge base gap that existed when it searched only through 2021. However, your application should include seminal works For example, you want to be sure that your application references recent works by leaders in your field, including possible reviewers of your application.

It’s imperative to thoroughly check AI-generated content – both words and images –  as AI can present erroneous or inaccurate information. AI-generated content can also be biased. Users should also be careful with proprietary information.

One last thing: Sponsors are addressing the use of AI in submissions. There is a range of acceptability regarding AI among sponsors, so be sure to check your sponsor’s policies. Currently, NIH’s only policy regarding AI is that it does not allow the use of AI in the peer review process.

For those wishing to explore the possibilities of AI for grantwriting purposes, a suggestion is to test the possibilities when the stakes are low, not when you are racing to meet an application deadline.

AI has many benefits to the grantwriting process and it’s expected that this will continue to evolve. However, for the foreseeable future, it’s clear that grant applications will continue to need that human touch.

References

1. Seckel E., Stephens B.Y., Rodriguez F. (2024) Ten Simple Rules to Leverage Large Language Models for Getting Grants. PLOS Computational Biology, 1 March 2024. doi: https://doi.org/10.1371/journal.pcbi.1011863.

2. Najafali D., Hinson C., Camacho J.M., Galbraith L.G., Gupta R., Reid C.M. (2023) Can Chatbots Assist with Grant Writing in Plastic Surgery? Utilizing ChatGPT to Start an R01 Grant.  (Letter to the Editor). Aesthetic Surgery Journal 43(8) NP663-NP665.

Becky Miro Bouvier Grant Guest Author

Author:
Becky Miro

This guest post was written by Becky Miro.

Rebecca Miro, PhD, CP, CRA has 25 years of experience in research administration and research coordination, primarily in the higher education setting. She attained the Certified Research Administrator (CRA) credential in 2004. Dr. Miro is also a certified prosthetist with an interest in issues faced by women with limb loss. She received a grant from the American Orthotic & Prosthetic Association (AOPA) to conduct a systematic literature review and retrospective data analysis regarding female amputee issues. Dr. Miro completed her PhD in Curriculum and Instruction with an emphasis in Higher Education Administration at the University of South Florida.

Categories:
Related posts

You May Also Be Interested In

We read all NIH notices for our clients. When you join our mailing list, we’ll pass along important changes directly to your inbox, as well as opportunities to improve your grantsmanship skills.
Primary Position
Lead Source

Wait!

Subscribe to our monthly newsletter for the latest NIH news, grantwriting tips, and more.

NIH October 2023 Newsletter cover