LSE: The Ballpark | Is AI thinking about threatening humanity? with Dr Alessandra Buccella
Can AI systems go rogue? How and why might they be harmful to humans and society? Can AI think and understand in the way that people do? What is really going on behind the screen when you ask an AI bot a question or to perform a task for you?
To help find answers to these questions, for this episode of our series on AI and the US, the Phelan US Centre spoke to Dr Alessandra Buccella, assistant professor at the University at Albany, State University of New York. Dr Bucella studies the philosophy of cognitive neuroscience – how people think and how they perceive the world. Her work examines how this intersects with AI and the effects and risks that AI systems may pose to society. Their conversation covers how AI systems work, whether they can think, what that means for society, and whether AI actually can "go rogue" in a way that could harm humanity.
This episode was produced by Chris Gilson and Avan Fata.
Further reading and resources
- Buccella, A. (2026). Interaction, not dissection: reframing AI interpretabilityEuropean Journal for Philosophy of Science, 16(4). doi:10.1007/s13194-026-00800-6.
- Magnus, P.D., Buccella, A. and D’Cruz, J. (2025). Chatbot apologies: Beyond bullshit. arXiv.
Take the Ballpark Listener survey and enter the prize draw for £250 in vouchers!
The Ballpark will be ten years old in 2026, and we want to hear from you to make the podcast even better, so we’re running a listener survey.
- Fill in our listener survey – it only takes a few minutes – here: https://forms.office.com/e/Vcj8V8uGM1
- Voucher prize draw terms and conditions are available here: https://www.lse.ac.uk/united-states/the-ballpark/terms-and-conditions-for-the-ballpark-listener-survey-voucher-prize-draw