BIP AI – AI in research: Opportunities and challenges

See Blended Intensive Programmes for background and context.

Other presentations I delivered during this BIP for AI in education and research:

• Artificial intelligence in higher education: Opportunities and challenges.
• Workshop 1: AI fundamentals and prompting.
• Workshop 2: AI for organisation and communication.

---

Presentation title – Artificial intelligence in research: Opportunities and challenges

Download the presentation.

In this presentation (06 May 2024) as part of the BIP for AI in education and research, Guillem Jabardo and I discuss the opportunities and challenges that arise in the application of artificial intelligence in research.

The advent of large language models (LLMs) and generative AI has the potential to profoundly impact academic research practices. Unlike traditional information retrieval systems, generative AI models generate original text outputs word-by-word based on the input prompt, without retrieving from a fixed database. This allows generative AI to act as an “expertise abundance” tool, capable of providing knowledgeable responses and creative ideas across diverse domains.

This presentation explores how generative AI can support the research process, from question formulation and literature review to methodology design, data analysis, summarisation, and writing. It demonstrates a variety of AI tools tailored for different research tasks, including research question generators, literature assistants, methodology recommenders, summarisation models, and writing aids. Hands-on activities allow participants to experience using these tools firsthand.

While extremely powerful, generative AI has limitations including potential hallucinations, biases, and lack of grounding that necessitate critical review. Ethical considerations around academic integrity, equitable access, and transparent AI usage are discussed. The ability of generative AI to augment human cognition represents a paradigm shift for academia. This highlights the need for principled integration of these technologies into research workflows. Proactive research on maximising the benefits while mitigating risks is crucial for responsible AI development.