21 September 2026 | 16:00–17:00
Neelsie Road House Cinema, Stellenbosch University | Online
Artificial intelligence is rapidly changing how we work, think, research and make decisions. But as AI capabilities accelerate, an important question remains: how do we know what to trust?
In this NITheCS Colloquium, Francesco Petruccione explores the rapid evolution of large language model (LLM)-based AI, from the launch of ChatGPT to today’s increasingly capable systems contributing to frontier mathematics and scientific research.
The pace of change has created a familiar challenge: our tools are advancing faster than the systems, habits and institutions designed to support them.
The talk will examine what this transformation means for working scientists and researchers — from changing research practices and verification methods to the institutional and geopolitical forces shaping AI.
Drawing on his own experiences using AI in research, Petruccione will share examples of both the opportunities and the risks: instances where AI tools helped uncover issues that peer review had missed, as well as situations where they came close to producing misleading conclusions.
Rather than offering simple answers, the colloquium will focus on the questions researchers should be asking as AI becomes increasingly embedded in scientific work:
- What can we trust AI systems to do?
- How should researchers verify AI-generated results?
- What happens when our tools become more capable than the workflows around them?
- How should universities and scientific institutions adapt?
- What does the rapid development of AI mean for the future of research?
Join us for a timely conversation about AI, scientific research, trust, verification and the changing nature of intellectual work.
Audience: Academic staff, postgraduate students and undergraduate students
Event type: NITheCS Colloquium
Venue: Neelsie Road House Cinema, Stellenbosch University + Online
