Recursive Self-Improvement: When AI Builds Itself
8 July 2026 · 11:30 am–11:55 am · Refectory
Anthropic recently reported that AI is already accelerating its own development: more than 80% of the code the company ships is now written by its models. Pushed far enough, that trend points to systems capable of designing and training their own successors with little human input: recursive self-improvement. Is that a distant dream, an emerging reality, or an imminent danger? This panel brings together differing views to debate how close we actually are, what would have to be true for the loop to close, and what it would mean for safety, governance, and human oversight if it did.
Speakers
Toby Walsh
Chief Scientist, UNSW's AI Institute
40 years researching AI. Professor of AI.
Tiberio Caetano
Chief Scientist, Gradient Institute
AI scientist. 25+ years in AI: doing research, founding companies, mentoring students, data scientists and researchers, advising institutions. 10+ years in AI ethics/safety/governance.
Dane Sherburn
Co-Founder and CEO, P-Zero Research
Anna Goldsworthy
Dean, Elder Conservatorium of Music and School of Performing Arts, Adelaide University
Professor Anna Goldsworthy is a pianist, writer and academic. She is Dean of the Elder Conservatorium of Music and School of Performing Arts at the University of Adelaide, and Artistic Director-designate of the Australian National Academy of Music (ANAM). An award-winning author and performer, her recent Quarterly Essay, 'The God We Made: the Threat an Promise of Artificial Intelligence' examines the impact of artificial intelligence on art, education and human creativity, bringing a distinctive interdisciplinary perspective to contemporary debates.
Audience Q&A
Ask a question or upvote others.
Loading questions…