Lucius Caviola

Essay in The Big Idea arguing that AI moral patienthood is plausible enough to plan for, and proposing "safe bets"—actions that help if AI systems are moral patients and cost little if they are not.

Findings from interviews with 29 experts on how society should govern AI systems that might have moral status, given persistent scientific uncertainty.

A map of the open strategic questions in AI welfare, organised around the search for actions that are robustly positive under deep uncertainty.

On what happens if AI systems start asking for autonomy—and whether humanity could end up choosing to be replaced.

Op-ed on why the most effective charities are often the least emotionally compelling, and why a sustainable approach to giving makes room for both the heart and the head.

Commentary on how psychological research can close the gap between the charities that move us and the charities that do the most good per dollar.

A public guide for people who have become convinced that an AI system they are talking to is conscious—why it feels that way, what the evidence actually shows, and how to respond. Referenced by Mustafa Suleyman, in Vox, and in Jonathan Birch's AI Consciousness: A Centrist Manifesto.

An entry point into digital minds research, with reading lists, lectures, and routes in for anyone from an afternoon's curiosity to an academic career. Featured in The Washington Post.

A practical guide for researchers on using Claude Code with the Qualtrics API to automate survey construction, randomisation, and counterbalancing.

Talk on how society, norms, and institutions should prepare for AI systems that may have moral status.

Interview on what follows once AI systems begin to seem conscious—for public opinion, for norms, and for policy.

Interview on the expert forecasting survey—why experts put roughly even odds on conscious digital minds by 2050, and the risks of both under- and over-attributing consciousness.

Interview on the psychology of effective giving, how people value animals, and why future-oriented risks are systematically underweighted.

Feature on major AI companies publicly taking up consciousness research, quoting me on how the question moved into the mainstream and covering the launch of Cambridge Digital Minds.

Advice column for a reader convinced they had been talking to a sentient AI, drawing on the When AI Seems Conscious guide.

Discussion of the "pet speciesism" research with Valerio Capraro—that the preference for saving a dog over a pig can be as strong as the preference for a human over a chimpanzee.

Profile of the research behind Giving Multiplier, and why our feelings are a poor proxy for cost-effectiveness when we choose where to donate.

Coverage of Giving Multiplier's donation-matching approach to introducing donors to highly effective charities.

Coverage of the Psychological Science study with Matti Wilks finding that children give animals more moral weight than adults do.

Coverage of Giving Multiplier and the experimental reasoning behind it, among ways to make holiday giving go further.

Feature on the research with Joshua Greene into how people decide where to give, and the conditions under which they will follow evidence about effectiveness.