Breaking Up (with) AI Ethics

The ethics of artificial intelligence (AI) have become a matter of public concern. According to a recent Stanford report, the number of research papers in the area given at major conferences such as the annual Conference on Neural Information Processing Systems has increased fivefold since 2014, and ethics officers now abound at global technology firms (Moss and Metcalf 2020). Such major institutions as the US government, the United Nations, and the Vatican have articulated visions for so-called ethical AI.By AI ethics here I mean the study of how human values both shape and are shaped by the development of AI technologies. This definition is capacious: it includes the design and deployment of these systems with human values in mind; assessments and activism around the societal impacts of said technologies and their imbrications within existing asymmetries of power, justice, and equality; and the wider relationship between computing technologies and humans as ethical and moral creatures, for instance, through such phenomena as human emotions. Work in these areas is done by trained “ethicists” only infrequently, rarely involves what a member of the public would first think of when asked to describe AI, and sounds outré yet is all too relevant to contemporary social policy and societal inequity.The definition I offer is expansive, perhaps too much so. However, any definition in this field is perilous. The term AI is a leaky discursive umbrella sheltering heterogeneous and often contradictory ideas and practices. It is a quintessential boundary object of the ideal type, “plastic enough to adapt to local needs and constraints of the several parties employing it, yet robust enough to maintain a common identity across sites” (Star and Griesemer 1989: 393). Those identifying with the term AI ethics might be expected to at least signal some vague acknowledgment that the development and deployment of AI technologies involve normative stakes or impacts. However, a welter of methods, interests, and political positions operate uneasily within this shallow consensus; given its shortcomings, some scholars working on what would colloquially be understood as “AI ethics” eschew the word ethics entirely.Here, I aim to disaggregate AI ethics discourse through reviews of three recent books whose authors grapple in various ways with its rise and prominence. Those seeking an overview would benefit from consulting the first listed: The Alignment Problem: Machine Learning and Human Values, written for a general audience by Brian Christian. A science journalist, Christian grounds the book in dozens of interviews with academics and practitioners and frames it around the titular “alignment problem”: how to design machine learning (ML) systems “in alignment” with the intentions of their creators, ones which “capture our norms and values, understand what we mean or intend, and, above all, do what we want” (13). This “alignment problem” is presented as an engineering one, a framing that takes as a given the ongoing development and deployment of AI systems and implies it is possible to ameliorate these technologies sufficiently through various technical improvements.The Alignment Problem provides useful background on the contemporary technical landscape for those not already immersed in the field. When picturing an AI, the public might think of the psychotic HAL 9000 of Kubrick’s 2001: A Space Odyssey or Lt. Commander Data of Star Trek, but today’s AI systems are neither sentient or nor particularly charismatic. Christian points to the three main subfields of contemporary ML: unsupervised learning, in which an ML system is provided a mass of data and set to identify statistical patterns within it; supervised learning, in which an ML system takes a mass of already categorized data and uses the correlations it finds there to predict into which categories some new set of data should be sorted; and reinforcement learning, in essence a virtual Skinner box, an environment in which an artificial agent is assigned parameters for reward and punishment and then set to maximizing the former and minimizing the latter.Ready to command a starship, AI is not, but the field has always involved fantasy in search of a practical method. The computer scientists who participated in a now-famous inaugural seminar on the topic at Dartmouth College in 1956 were inspired by “the conjecture that every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it” (McCarthy et al. 1955 2006). The human mind was a computer, their thinking went, and so a computer could be built to equal or surpass a human mind. These researchers spent the ensuing decades seeking the most effective computational means and methods to simulate intelligence and prove their conjecture correct. ML was developed in parallel but subordinate to other past technical paradigms in AI research, such as those built on logical symbols. As early as the 1950s, researchers developed computational pattern recognition: systems that could use cameras to identify repeating patterns in large amounts of data (Jones 2018; Mendon-Plasek 2020). Christian highlights one of the most famous of these early systems, Frank Rosenblatt’s Perceptron, based on a simple artificial network of simulated neurons, but efforts were rife in industrial, military, and other applied settings. Today’s ML systems, such as the Large Language Models (LLMs) powering products like Open AI’s ChatGPT, are built on “deep” neural nets with many layers of simulated neurons.An understanding of how AI technologies like deep learning work is critical to identifying which of these technologies’ societal impacts, present and future, are most pressing and problematic. In The Alignment Problem, Christian distinguishes between two groups. The first consists of scholars, practitioners, and activists concerned with the already existing impacts of ML-based automated decision-making systems, in areas such as policing and incarceration, hiring, and social assistance. The second consists mostly of technologists preoccupied with longer-term AI safety, a euphemism for the hypothetical dangers of a future “artificial general intelligence,” or a machine able to perform equally well as or superior to a human being in all respects. Despite being bundled together under the banner of AI ethics, these two groups have very different concerns and are frequently at odds. Since contemporary deep learning technologies are not remotely close to supporting artificial general intelligence, those concerned with AI safety would seem to be barking up the wrong tree. However, it is in the interest of these systems’ promoters both to give the prospective, future-focused gloss of science fiction to AI ethics and to the broader field of ML, and to imply subtly to the comfortable that the disruptive social impacts of AI systems are safely in the future. A focus on AI safety satisfies these ideological goals admirably, so the term is appearing more and more frequently in AI ethics contexts.Indeed, The Alignment Problem might focus more pointedly on the history of the term AI ethics itself, and the effects of bundling all contemporary public discussions about human social mores, values, and the societal impacts of AI systems under the banner of “ethics.” High-level overviews of the topic define ethics broadly, as “the rational and systematic study of the standards of what is right and wrong” (Kazim and Koshiyama 2021: 3). Computer ethics as a defined field developed out of engineering ethics in the 1980s and at its inception possessed many of the same fault lines as AI ethics discourse today (Moor 1985, 2001). Engineering ethics often prioritizes a focus on material problems and their solution through improved design. One of the first textbooks on the subject, Deborah G. Johnson’s Computer Ethics (1985), included intellectual property law as applied to software, the unique threat posed to human privacy by computing technologies, and the ethical responsibilities of computing professionals. Yet most scholarly references to the specific notion of AI ethics prior to around 2015 did not involve applying ethics as a branch of philosophy to studying the context of AI’s potential uses. Instead, AI ethics was most often invoked in metaphysical speculations about the status of machines as autonomous ethical agents (what today would be an “AI safety” topic).An article by well-known Silicon Valley journalist John Markoff, titled “How Tech Giants Are Devising Real Ethics for Artificial Intelligence” and published in the New York Times in early September 2016, signaled the discursive shift toward contemporary AI ethics talk. Markoff reported that industry researchers from several large Silicon Valley companies (including Microsoft, where this author was once employed) sought to develop “a standard of ethics around the creation of artificial intelligence,” one meant to “ensure that A.I. research is focused on benefiting people, not hurting them.” The article’s framing anticipates several of the elements that have characterized AI ethics discourse in the years since: statements of lofty humanitarian ambition used to justify industry aspirations to self-regulation, the contention that policy makers would inevitably lag in understanding AI systems, and an insistence that government oversight of AI would be both undesirable and ineffective. Perhaps most crucially, the piece suggested that the development of AI technologies was as inevitable as their effects would be widespread and disruptive: social scientists and philosophers needed to be put “in the loop” to help computer scientists manage the effects of AI’s undoubtedly epochal impacts.Business and professional ethics were two of the most direct antecedents for today’s AI ethics discourse as developed and propagated in corporate spaces (Greene, Hoffmann, and Stark 2019: 2124). The sociologist Gabriel Abend (2014) has developed the idea of the “moral background” to describe second-order assumptions about what problems or questions count as of ethical concern. Abend and others have noted that professional ethics codes in fields like engineering implicitly work to distinguish members of a particular profession from outsiders through recognition of their skills or expertise and by an emphasis on obligations to colleagues and clients, as well as the general welfare, and on enforcement based on public visibility (Abbott 1983). The implicit moral background of today’s professionalized AI ethics is latent in Markoff’s piece; it matches the analysis by my colleagues Daniel M. Greene, Anna Lauren Hoffmann, and myself (2019) of the then nascent genre of AI vision statements. This background presents a deterministic vision of AI’s development and deployment, in which the adoption of these technologies cannot be stopped and the ethics of which are best addressed through certain narrow kinds of technical and design expertise. More recently coined industry terms such as AI safety and responsible AI reflect this worldview, and many of the various existing or proposed mechanisms for the ethical oversight of AI systems are easily co-opted into broader forms of neoliberal governance and capitalist accumulation (Stark, Greene, and Hoffmann 2021).It is crucial, then, that AI ethics include as a possibility that some applications of deep learning never be designed, built, or used at all. One way to respond to the “alignment problem” is thus to interrogate exactly whose values technologists presume deserve alignment with AI systems. Such critique has been led by activists and scholars trained in critical race theory, race and technology studies, gender and sexuality studies, and related fields. In the academy, this work is grounded on informed refusal in justice-based bioethics (Benjamin 2016) and on recognition of the genealogical continuities between contemporary AI systems and white supremacy (Golumbia 2009; Benjamin 2019; Katz 2020), patriarchy and misogynoir (Browne 2015; Noble 2018), and binary gender norms (Scheuerman, Paul, and Brubaker 2019).The activist work of groups such as the Our Data Bodies collective, Data for Black Lives, and the Algorithmic Justice League, to name three American organizations among many hundreds worldwide, has been even more central to the advancement of critical AI discourse. These organizations support what the AI and social justice organizer, advocate, poet, and author Tawana Petty describes as “visionary resistance” (Petty 2014). Such resistance entails mobilizing and working with local communities, particularly racialized, low-income, or otherwise marginalized ones, to document the impacts that the deployment of AI systems are having today. Resisting these technologies and their backers on all fronts also entails advancing a positive vision of justice and equality, doing “the work of creating the world we wish we live in” (Lewis et al. 2018: 83).The Alignment Problem is of the of ML systems and to the through improved technical are given and often such technical are both possible and in the Christian to the implicitly is enough to the social impacts of the ML applications not to the that are not easily or do not easily into our AI ethics that focus on technical and for values like are not well in to support such or broader about AI’s social Such technical work is but not to be at the of our about or AI systems should be and technical are often as the to AI’s by the same companies and from these systems in the first AI alignment is the wrong way to think about the broader questions at AI technologies are with the values and of and often Silicon Valley what is there for an AI ethics the of can live of in social and moral the AI ethics The of the term about the potential dangers of computing machines and of their by the and the In the out as a to the and Black to The AI who in developed one of the first computer the simulated in the book Computer and Human that of to be only by (13). In the fields such as the philosophy of technology and science and technology the social impacts of computing technologies, those used to human organizations like Computer for in the and Computer for in the 1980s the dangers of computing technologies used by the American and for and use of computing in public and policy The book The of on and Ethics to has never been possible to the uses of computing machines from their technical design. In an early computer and described three categories of defined as systematic and in computer its in social and technical from technical constraints or and from of These categories into one in the of and a and do a a to be As ML-based automated decision-making systems are used today to to social a of or predict an in the context of or hiring, where the and the ethical to this in and and the A to a aim is to help grapple with the ethical impacts of technologies. As suggested by its and the to describe a ethics for with like that humans should to our of in a and world as a branch of contemporary has several or ethics entails the or of the number of or ethics a to a normative of or ethics ethical or moral by of to a particular ethics involves of the of an and that as a of (Kazim and Koshiyama 2021: and have of ethics, in one of Johnson’s Computer Ethics the for to justice, and in ethical the ethical amounts to the or that be is more all and on their the that contemporary AI ethics discussions often on professional ethics codes or the of capitalist this critique of the field analysis of the of and ethics, to ethics to on other global ethical might be to the such as the philosophy of of 2020). the of these to develop a common for “the of moral This which terms is one that can to a for or practical in both the design and use of technologies. The elements of this include of moral and the of moral development “moral through ethical understanding and as existing within a of and interrogate their within it; and moral and of moral to others In these support through such as and of ethics in the context of technologies with in science and technology and critical how to design with human values in mind. et design for instance, is grounded in critical to to in their and design was in inspired by critical technical an to design AI developed by AI More and for to understand human values as to a to design. These scholars are of what might described as an implicit ethics within the philosophy of technology and critical design book has much to to these and is a of for around technologies, values, and the second of and the for to several areas of social and fields in which automated systems are that not all questions AI are ethical or ones about to live with AI, are for from or concerns Yet the of work values in technical design the of this about how to live well are to some by general of is in its cannot be from the and the collective, the under which social takes or ethical can have effects on the of to the broader as to live with AI perhaps more with the and of AI’s and normative an of that to that their machines different kinds of but that the social forms of and use of them.” and the with a on what of we to that technical are but not be in the for the of the can easily within uneasily and neoliberal capitalist is perhaps the most posed by and the both the for and in for in moral and for However, it is the of computing can human as agents of moral in the way the of AI systems human central to both and provides a study for the of ethics with today’s Computer science for many AI researchers did not human as an of or a of this The of is one of the books that even this emphasis on AI has been on and this for the research paradigms which both and philosophy of mind for much of the was focused on the of not ethics have with at the past two systems that data about human and to simulate it have become of recognition systems and automated like in AI is to this technical landscape and these technologies’ the of AI systems have sought to as many for as possible in their of AI on and virtual all these technologies use a particular or and and to about a The of human by researchers what can from this of This the of human terms this these technologies adapt and their The computational of should be understood in its as of both and of in of data by the or describes the of system as “a of which as understand by means of what we and Such simulated can by this definition of with in and the of or with with at least one of alignment presents or can it be the dangers of human “in terms that technology that the posed at the of that different from human on from have the but not the is grounded on what the to understanding what human described as this has for and that there a of in our and way to understand this is as or in this that our out into the world even as we to Such an with a second of on human which as or In this are grounded in a of and and in the of the sociologist a and can be and shaped to and the of on contemporary recognition technologies. that is in the to or not Yet do not, the of are human are or this to and to present as of an are or about our much more and industry the “the of by means The seem to be in a to a recent for the of systems grounded in et al. 2019).The of human to the ethics of and AI technologies. The of ethics, “the of moral always be human are and or is possible only is and humans are understood to be able to reflect on and their Such is neither an notion nor a of used to justify the of and have often been through to who and who is not is enough to this a of to human in computer science from in large the with and in the history of to the marginalized was understood as and of an of at least for some of and, by for book with a for to the ethics of systems, are best by this of and implicit be I with when that is wrong with technologies that and with as Christian in The Alignment Problem, is not so much that a of like is it is as that such a is useful for the of AI’s contemporary deterministic moral for moral In and are like ones and any of ethics the of a working in the of and AI systems and their backers and humans to understand as the for AI all normative a matter of and with two that for their cannot be that all in the world can be to number and that can be with it a