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 does not 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
The long-run relationship between per capita incomes and population size – Maya Eden (University of Zurich) and Kevin Kuruc (Population Wellbeing Initiative, University of Texas at Austin)
The relationship between the human population size and per capita incomes has long been debated. Two competing forces feature prominently in these discussions. On the one hand, a larger population means that limited natural resources must be shared among more people. On the other hand, more people means more innovation and faster technological progress, other things equal. We study a model that features both of these channels. A calibration suggests that, in the long run, (marginal) increases in population would…
Dispelling the Anthropic Shadow – Teruji Thomas (Global Priorities Institute, University of Oxford)
There are some possible events that we could not possibly discover in our past. We could not discover an omnicidal catastrophe, an event so destructive that it permanently wiped out life on Earth. Had such a catastrophe occurred, we wouldn’t be here to find out. This space of unobservable histories has been called the anthropic shadow. Several authors claim that the anthropic shadow leads to an ‘observation selection bias’, analogous to survivorship bias, when we use the historical record to estimate catastrophic risks. …
Doomsday and objective chance – Teruji Thomas (Global Priorities Institute, Oxford University)
Lewis’s Principal Principle says that one should usually align one’s credences with the known chances. In this paper I develop a version of the Principal Principle that deals well with some exceptional cases related to the distinction between metaphysical and epistemic modality. I explain how this principle gives a unified account of the Sleeping Beauty problem and chance-based principles of anthropic reasoning…