An Unethical Optimisation Principle

An Unethical Optimisation Principle

Explore the concept of unethical optimization in AI systems through this 53-minute seminar led by Nicholas Beale, Heather Battey, and Robert S. MacKay from the Alan Turing Institute. Delve into a groundbreaking paper destined for PNAS publication that quantifies the risk of AI selecting unethical strategies when maximizing risk-adjusted returns. Examine the implications for financial markets, AI regulation, and the role of impartial spectators. Gain insights into the importance of carefully crafting objective functions to mitigate ethical risks in AI decision-making processes.