arXiv subjects

David Thorstad

Publications and source records attributed to David Thorstad.

3 recordsLinked to original sources

arXiv2026

Instrumental convergence and power-seeking

Recent years have seen increasing concern that artificial intelligence may soon pose an existential risk to humanity. One leading ground for concern is that artificial agents may be power-seeking, aiming to acquire power and in the process disempowering humanity. I show how the argument from power-seeking rests on a strong version of a claim known as the instrumental convergence thesis. I explore leading defenses of the instrumental convergence thesis and argue that none establishes the thesis in a strong enough form to ground the argument from power-seeking. I discuss implications for longtermism, the governance of artificial intelligence, and the methodology of studying risks posed by artificial agents.

cs.AI↗

arXiv2026

Revisiting the shutdown problem

A key premise in leading arguments for existential risk from artificial intelligence is that malfunctioning artificial agents could not be easily shut down. This motivates the catastrophic shutdown problem of ensuring that agents can be shut down before causing an existential catastrophe. A range of arguments and theorems are offered to suggest that solving the catastrophic shutdown problem is difficult, bolstering arguments for existential risk and motivating a search for solutions to the catastrophic shutdown problem. This paper argues for two conclusions. First, existing arguments require substantial additional assumptions and evidence to play the desired role in motivating existential risk concerns. Second, concern for the catastrophic shutdown problem has led to technical solutions that impose a high safety tax on model performance. These results redirect research towards the identified assumptions and provide guidance for technical alignment strategies.

cs.AI↗

arXiv2023

Cognitive bias in large language models: Cautious optimism meets anti-Panglossian meliorism

Traditional discussions of bias in large language models focus on a conception of bias closely tied to unfairness, especially as affecting marginalized groups. Recent work raises the novel possibility of assessing the outputs of large language models for a range of cognitive biases familiar from research in judgment and decisionmaking. My aim in this paper is to draw two lessons from recent discussions of cognitive bias in large language models: cautious optimism about the prevalence of bias in current models coupled with an anti-Panglossian willingness to concede the existence of some genuine biases and work to reduce them. I draw out philosophical implications of this discussion for the rationality of human cognitive biases as well as the role of unrepresentative data in driving model biases.

cs.AI↗