BEGIN:VCALENDAR
PRODID:-//University of Otago//Events Calendar//EN
VERSION:2.0
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-WR-CALNAME:University of Otago Calendar
BEGIN:VEVENT
UID:https://www.otago.ac.nz/news/events/biochemistry-seminar-professor-mik-black
URL:https://www.otago.ac.nz/news/events/biochemistry-seminar-professor-mik-black
DTSTART;TZID=Pacific/Auckland:20260721T120000
DTEND;TZID=Pacific/Auckland:20260721T130000
SUMMARY:Biochemistry seminar: Professor Mik Black
DESCRIPTION:Large language models and agentic coding tools for genomic data analysis: What are the impacts on our research and teaching?In late 2022, ChatGPT became the first Large Language Model (LLM) to "go viral", gaining widespread public attention and kicking off a multi-billion dollar Generative Artificial Intelligence (GenAI) arms race between some of the world's largest corporations. Despite my best efforts over the past four years, ignoring GenAI does not seem to be a viable option. Like it or not, the increasingly sophisticated tools being produced are already having an impact on our research and teaching, and failure to either adapt or adopt threatens to decrease our research competitiveness and disadvantage our students. Although much of the discussion within the University about GenAI has focused on students' use of these tools to assist with assessment tasks, less time has been spent looking at how GenAI, and LLM-based tools in particular, can be used to enhance research. An obvious example of an AI-fuelled scientific success is AlphaFold3, a deep learning model for protein structure prediction that has rapidly become an indispensable research tool. Few would doubt the usefulness of AlphaFold3, but when it comes to GenAI-based tools, questions remain about both their utility, and their acceptable and responsible use in research and teaching. As "word prediction engines", one area where LLMs excel is in the generation of computer code. In this presentation I will look at LLM-based tools that are available to assist with writing code, from chat-based interfaces through to more sophisticated agentic tools and will demonstrate how these can be used to assist with both simple analysis tasks, as well as the interrogation of large-scale genomic data sets. I will also discuss how we can start to introduce these tools to our students, focussing on the skills we need to teach to ensure that curiosity, critical thinking and scientific rigour remain at the core of our work. 
LOCATION:Biochemistry Seminar Room G.13 (BIG13), 710 Cumberland Street, Dunedin
END:VEVENT
END:VCALENDAR