Towards shutdownable agents via stochastic choice
Elliott Thornley (Global Priorities Institute, University of Oxford), Alexander Roman (New College of Florida), Christos Ziakas (Independent), Leyton Ho (Brown University) and Louis Thomson (University of Oxford)
GPI Working Paper No. 16-2024
Some worry that advanced artificial agents may resist being shut down. The Incomplete Preferences Proposal (IPP) is an idea for ensuring that doesn’t happen. A key part of the IPP is using a novel ‘Discounted REward for Same-Length Trajectories (DREST)’ reward function to train agents to (1) pursue goals effectively conditional on each trajectory-length (be ‘USEFUL’), and (2) choose stochastically between different trajectory-lengths (be ‘NEUTRAL’ about trajectory-lengths). In this paper, we propose evaluation metrics for USEFULNESS and NEUTRALITY. We use a DREST reward function to train simple agents to navigate gridworlds, and we find that these agents learn to be USEFUL and NEUTRAL. Our results thus suggest that DREST reward functions could also train advanced agents to be USEFUL and NEUTRAL, and thereby make these advanced agents useful and shutdownable.
Other working papers
High risk, low reward: A challenge to the astronomical value of existential risk mitigation – David Thorstad (Global Priorities Institute, University of Oxford)
Many philosophers defend two claims: the astronomical value thesis that it is astronomically important to mitigate existential risks to humanity, and existential risk pessimism, the claim that humanity faces high levels of existential risk. It is natural to think that existential risk pessimism supports the astronomical value thesis. In this paper, I argue that precisely the opposite is true. Across a range of assumptions, existential risk pessimism significantly reduces the value of existential risk mitigation…
Quadratic Funding with Incomplete Information – Luis M. V. Freitas (Global Priorities Institute, University of Oxford) and Wilfredo L. Maldonado (University of Sao Paulo)
Quadratic funding is a public good provision mechanism that satisfies desirable theoretical properties, such as efficiency under complete information, and has been gaining popularity in practical applications. We evaluate this mechanism in a setting of incomplete information regarding individual preferences, and show that this result only holds under knife-edge conditions. We also estimate the inefficiency of the mechanism in a variety of settings and show, in particular, that inefficiency increases…
Is In-kind Kinder than Cash? The Impact of Money vs Food Aid on Social Emotions and Aid Take-up – Samantha Kassirer, Ata Jami, & Maryam Kouchaki (Northwestern University)
There has been widespread endorsement from the academic and philanthropic communities on the new model of giving cash to those in need. Yet the recipient’s perspective has mostly been ignored. The present research explores how food-insecure individuals feel and respond when offered either monetary or food aid from a charity. Our results reveal that individuals are less likely to accept money than food aid from charity because receiving money feels relatively more shameful and relatively less socially positive. Since many…