4 responses to “Using AI in Open Research: Possibilities and Challenges”

  1. Karina

    AI tools could be used positively to help with summarizing major themes across multiple papers for a literature review. However, it’s possible that AI could miss new or recent papers in its assessment while hallucinating information if its output is not thoroughly checked/verified.

  2. Barbara Clerihue

    I think that in the early stages of a project, AI can significantly enhance the literature review and background research process by rapidly scanning and processing large volumes of academic publications. Some specialized AI systems can identify relevant papers, summarize key findings, and highlight gaps in the existing research. This helps researchers by making much of this work faster and more “efficient”.
    However, use of AI risks overlooking subtle insights in papers, missing nuances like tone, skepticism, or methodological caveats that human researchers would catch. Additionally, AI often relies on existing databases that may be skewed toward English-language or high-impact journals, which can under-represent niche or non-Western research and create blind spots in the literature review. Some AI tools are not updated in real time and may miss recent studies, potentially leaving gaps in the most current research landscape, and finally AI can hallucinate which – without thorough review and analysis – could insert false information in to a lit review.

  3. Marisa

    I can see how AI could be used to help with research dissemination. AI could be used to help adjust language to be more suitable to individuals without the same scientific background. We could ask AI tools like ChatGPT or Microsoft Co-Pilot to help summarize the research using language that is more accessible to a broader audience. However, as mentioned by the user above, AI models can hallucinate or provide inaccurate information and should always be reviewed for it’s accuracy before sharing!

  4. Nina

    I could see AI tools quickly being integrated in all stages of the research lifecycle. At this point, I think the greatest strength of tools such as ChatGPT in research is in literature review. GenAI seems to be able to quickly generate an outline of a topic although it can fail somewhat in terms of providing details. As a starting point for a lit review I think it could be very helpful. As others mentioned, the bias involved in training GenAI tools could present a problem. Also, accurately acknowledging the source of information obtained through AI tools could be a challenge, and even an ethical concern if the intellectual property used by the AI is not intended for this type of use.

Leave a Reply