Jecker et al critically analysed the predominant focus on existential risk (X-Risk) in artificial intelligence (AI) ethics, advocating for a balanced communication of AI’s risks and benefits and urging serious consideration of other urgent ethical issues alongside X-Risk.1 Building on this analysis, we argue for the necessity of acknowledging the unique attention-grabbing attributes of X-Risk and leveraging these traits to foster a comprehensive focus on AI ethics. First, we need to consider a discontinuous situation that is overlooked in the article (...) by Jecker et al. This discontinuity refers to the phenomenon where X-Risk is perceived as dominating the discourse, yet contrary to expectations, it does not lead to a significant allocation of social resources for specific risk management and practical initiatives. In both the specific realm of ethical AI initiatives and the broader scope of AI risk management, the responses to X-Risk do not have the advantage of prioritising the allocation of resources over other related risks.2 This discrepancy suggests that, in terms of actual social resource allocation, X-Risks do not receive commensurate resources relative to the attention they attract. Unlike other types of risks, X-Risk is perceived as a distant threat, which does not correspond with actual initiatives in terms of media exposure. Despite the prominence of the longtermism view in various media and public discourse, the X-Risk of AI often serves merely as a cautionary note or opinion about the current situation. This suggests that concerns about an AI-driven catastrophe have not been effectively translated into practical initiatives. The gap between the attention drawn and …. (shrink)

In their article,1 Jecker et al highlight a widespread and hotly debated issue in the current application of artificial intelligence (AI) in medicine: whether we should develop more powerful AI. There are many perspectives on this question. I would like to address it from the perspective of the fundamental purpose of medicine. Since its inception, medicine has been dedicated to alleviating human suffering and ensuring health equity. For thousands of years, we have made great efforts and conducted many investigations to (...) eliminate human suffering and have achieved remarkable results. However, the reality of achieving health equity falls far short of expectations. The WHO considers health equity to be a fundamental human right, meaning that there should be no unjustified, avoidable or remediable differences among the population as a whole, whether those differences result from social, economic, demographic or geographic factors, or from other forms of inequality such as gender, race, disability or sexual orientation.2 Health equity means reducing and ultimately eliminating disparities in health outcomes between vulnerable groups and others, caused by health risk factors and difficulties in accessing quality health services. This means that we must strive to reduce structural social bias. It is well known that the traditional healthcare system is riddled with various health inequities. …. (shrink)

I agree with Jecker et al that “the headline-grabbing nature of existential risk (X-risk) diverts attention away from immediate artificial intelligence (AI) threats…”1 Focusing on very long-term speculative risks associated with AI is both ethically distracting and ethically dangerous, especially in a healthcare context. More specifically, AI in healthcare is generating healthcare justice challenges that are real, imminent and pervasive. These are challenges generated by AI that deserve immediate ethical attention, more than any X-risk issues in the distant future. Almost (...) 50 years ago, John Knowles edited a volume titled Doing Better and Feeling Worse: Health in the United States. We are ‘doing better’ because numerous advances in medical technologies are saving more lives and improving the quality of our lives. But we are ‘feeling worse’ because the additional costs are unsustainable, are threatening funding for other social goods and are increasing injustices regarding the allocation of those resources. This is precisely the situation we are faced with regarding AI in medicine today. Exploding healthcare costs have generated increasing pressure to control those costs, often with unjust consequences. AI is touted as a critical mechanism for controlling healthcare costs, often with little thought given to justice-relevant consequences. Consider, for example, …. (shrink)

Jecker et al 1 offer a valuable analysis of risk discussion in relation to Artifical Intelligence (AI) and in the context of longtermism generally, a philosophy prevalent among technocrats and tech billionaires who significantly shape the direction of technological progress in our world. Longtermists accomplish a significant justificatory win, when they use a utilitarian calculation that pits all future humanity against concerns about current humans and societies. By making this argument, they are able to have abstract (and uncertain) benefits for (...) an infinite group of people, outweigh concrete harms for current people—a nifty trick given it also often seems to align with their own personal benefits related to status, power and profit-making. In the AI space, specifically, this has worked well. As Jecker et al 1point out, while AI leaders have spoken openly and loudly about existential risk (X-risk) that conjures thoughts of sci-fi fantasies rather than realities (a strategy that also serves to hype up their work well beyond its current capacity), it fails to engage with concrete and less existential risks that have significant impacts. These risks are inconvenient to development in the Silicon Valley model that requires people to ‘move fast and break things’ rather than proceed carefully and thoughtfully. Not only that but …. (shrink)

We appreciate the helpful commentaries on our paper, ‘AI and the falling sky: interrogating X-Risk’.1 We agree with many points commentators raise, which opened our eyes to concerns we had not previously considered. This reply focuses on the tension many commentators noted between AI’s existential risks (X-Risks) and justice here and now. In ‘Existential risk and the justice turn in bioethics’, Corsico frames the tension between AI X-Risk and justice here and now as part of a larger shift within bioethics.2 (...) They think the field is increasingly turning away from ‘big picture’ questions new technologies raise and focusing on narrower justice concerns of less significance. They compare our paper’s emphasis on justly transitioning to more AI-centred societies with the approach of environmentalists fretting about human protection against climate change while losing sight of the need to protect the planet and all living things. Corsico doubts there is much point to pressing for ‘justice on a dead planet’, so too Corsico questions our concern with just transitions. Presumably, that matters little if intelligent life on Earth is destroyed. Corsico’s recommends bioethicists return to big questions, like: ‘Should we develop AI at all, given AI X-Risk?’. Yet this question is increasingly moot. The genie is already out of the bottle. The Future of Life Institute’s 2023 call for a temporary pause on training AI systems more powerful than ChatGPT fell on deaf ears, in part because freely accessible source codes enable anyone to train systems and create AI applications.3 The focus now must be managing the genie. While we take AI X-Risk seriously, …. (shrink)