AI and Illusionism as a Risk to Altruism - and new musings about the Fermi Paradox Florian Habermacher (Department of Economics, University of St. Gallen, Zurich, Switzerland) C2
We link the increasing prevalence of ever more complex Artificial (General) Intelligence (AI) to society's views about the nature of consciousness. We discuss effects the shifts in these views may have, on how people experience the world, and especially on how people deal with each other. Specifically, we consider the hypothesis that the spread of AI may lead to widespread adoption of illusionism, and that this poses a serious threat to altruistic behavior and therefore to the modern society at large, thus, that there is an AI risk of a type hitherto neglected in the literature. To analyse the plausibility of the hypothesis, we consider four separate questions: First, (i) does progress in computation and AI suggest that illusionism - the often belittled theory that consciousness as phenomenal experience - is an illusion created by our brain (i.e. that it is essentially standard computation rather than any intrinsically valuable feeling unexplained by basic physics) - is true? Second, and somewhat independently, (ii) will some type of illusionism plausibly - even if maybe mistakenly - become popular among laypeople and/or key stakeholders, as advanced AI starts to shape everyday life and suggest things akin to 'our' consciousness can be created and explained by raw compute? Third, (iii) could variants of illusionism justify a reduction of altruistic concerns, as, after all, 'feelings' of others would be mere computation, while as individuals we still have our own hardwired aims? And fourth, (iv) does the adoption of an illusionist view put human altruistic feelings and behavior at risk in practice (somewhat independently of (iii))? The expectation that illusionism may become more prevalent, and that this poses a serious risk to human altruism, first suggests that there is an unprecedented importance of policies keeping egoistic behavior in check. More speculatively, we potential links to the Fermi Paradox. And we find that in some sense AGI could be said to undermine human societiies even before it emerges.