The AI revolution is well underway, and its transformative impact is being felt across virtually all industries. From transportation and marketing to healthcare and finance, Artificial Intelligence (AI) is revolutionising the way we work.
It is natural that universities, researchers, and scholarly publishers are also experiencing this AI-driven transformation. For example, in the world of scholarly publishing, we are seeing the application of AI in various workflows, from easing manuscript submission processes, generating article summaries and improving plagiarism detection. In research, AI is used to analyse large amounts of data, automate repetitive tasks and identify errors in writing and research. Respondents to a survey from Nature reported using AI tools to help write code and papers, and highlighted the particular benefit of large language models to researchers who need to use English where this is not their first language. Over half of respondents highlighted benefits such as faster ways to process data, quicker computations and saving time and money.
At Research Consulting, we’ve been exploring the use of AI tools for a few years now to understand where and how they can assist us in our work. In this blog, I share some insights from our journey and what we expect to do next.
Our AI journey
As a starting point, we have focused on tools that offer transcription, language processing, content generation and analysis technologies. Our journey with AI began with the use of transcription and language processing software, Otter, which allows users to upload meeting recordings, generate full-text transcripts and assign parts of the transcript to specific speakers. Otter has significantly reduced the time we spend preparing and cleaning transcripts, allowing us to dedicate more time to analysis and generating insight for our clients.
An overview of the AI tools we trialled, including strengths and weaknesses
Following the success of Otter, we wanted to see how AI tools might support us in other areas of our work, such as literature reviews. To explore this potential, we commissioned a student project at the University of Nottingham in 2022, asking students to identify and test AI tools that we might adopt as a company, specifically in relation to literature reviews and analysis. The students identified Genei – a multi-document summarisation tool.
Since then, we have added generative AI tools like ChatGPT and Jasper to our toolbox, which are particularly helpful in supporting our marketing activities. When coupled with our use of Canva for visuals and social media communications, these tools have elevated our ability to engage clients and partners.
However, while there are numerous benefits to the use of AI tools, we are also aware of the negative impacts and risks that AI poses, especially regarding trust and transparency. As highlighted in an article from WIRED, the use of AI in academic journals, particularly where this is not made clear, has a number of ethical concerns including plagiarism, and fake or fraudulent data and references. This was reflected in the responses to Nature’s survey, with 55% of respondents reporting that AI tools made fraud easier and 53% stating that improper use can lead to irreproducible research. At Research Consulting, we are currently managing these risks by limiting our use of AI to data cleaning, copyediting, and supporting marketing materials, while exploring how to communicate to our clients when and how AI is being used, drawing on examples such as this policy from WIRED.
Exploring upcoming innovations
We have now tried and tested several AI tools across many aspects of project delivery and business activities. While our experience of these tools is still fresh, we do see great promise in the future. Going forward, we want to find out more about how we can automate our literature searches and reviews, and we are planning to trial Elicit and Litmaps to see what they can offer. We are also keeping an eye on emerging tools, to understand how they may be applicable to us as consultants. One example of this is MS Copilot – a next-generation AI tool combining large language models with data already held in our Microsoft 365 apps – and we are looking forward to its release in the coming months.
In a rapidly changing landscape, we look forward to continuing to innovate and create solutions that meet our clients’ needs and exceed their expectations.
