If AI can be conscious, then this will have major impacts on how to build AI (e.g., maybe AI labs avoid building the kinds of AI that are conscious), how to treat AI (see, e.g., "Sharing the World with Digital Minds and "Propositions Concerning Digital Minds and Society"), and perhaps even how we view our place in the universe as conscious biological beings.

One way of operationalizing the question of whether (and when) conscious AIs will be built is to ask whether leading thinkers on AI consciousness, otherwise known as artificial sentience or digital sentience, will believe we have conscious AI by given dates in the future. This is what we do in this question and in some others in the series (e.g., this one).

Additional background: there has recently been increased discussion of AI consciousness within the effective altruism (EA) community. Rob Long is one of the key people thinking about artificial sentience within this community: earlier this year he appeared on the 80,000 Hours podcast episode "Robert Long on why large language models like GPT (probably) aren't conscious", and co-authored "Moral Consideration for AI Systems by 2030".

EA funders can and have moved hundreds of millions of dollars into what they view as the most important problem areas in the world. Examples include global poverty, animal welfare, and AI safety. Whether or not AI consciousness (and related things like AI welfare and AI rights) becomes a highly funded area within the next few years will likely depend—at least in part—on to what extent AI consciousness is possible, and how soon it might arrive, in the view of key thinkers on AI consciousness within the EA community.