Advances in artificial intelligence (AI) raise important questions about whether people view moral evaluations by AI systems similarly to human-generated moral evaluations. We conducted a modified Moral Turing Test (m-MTT), inspired by Allen et al. (Exp Theor Artif Intell 352:24–28, 2004) proposal, by asking people to distinguish real human moral evaluations from those made by a popular advanced AI language model: GPT-4. A representative sample of 299 U.S. adults first rated the quality of moral evaluations when blinded to their source. (...) Remarkably, they rated the AI’s moral reasoning as superior in quality to humans’ along almost all dimensions, including virtuousness, intelligence, and trustworthiness, consistent with passing what Allen and colleagues call the comparative MTT. Next, when tasked with identifying the source of each evaluation (human or computer), people performed significantly above chance levels. Although the AI did not pass this test, this was not because of its inferior moral reasoning but, potentially, its perceived superiority, among other possible explanations. The emergence of language models capable of producing moral responses perceived as superior in quality to humans’ raises concerns that people may uncritically accept potentially harmful moral guidance from AI. This possibility highlights the need for safeguards around generative language models in matters of morality. (shrink)

"These notes are about the process of design: the process of inventing things which display new physical order, organization, form, in response to function." This book, opening with these words, presents an entirely new theory of the process of design. In the first part of the book, Christopher Alexander discusses the process by which a form is adapted to the context of human needs and demands that has called it into being. He shows that such an adaptive process will be (...) successful only if it proceeds piecemeal instead of all at once. It is for this reason that forms from traditional un-self-conscious cultures, molded not by designers but by the slow pattern of changes within tradition, are so beautifully organized and adapted. When the designer, in our own self-conscious culture, is called on to create a form that is adapted to its context he is unsuccessful, because the preconceived categories out of which he builds his picture of the problem do not correspond to the inherent components of the problem, and therefore lead only to the arbitrariness, willfulness, and lack of understanding which plague the design of modern buildings and modern cities. In the second part, Mr. Alexander presents a method by which the designer may bring his full creative imagination into play, and yet avoid the traps of irrelevant preconception. He shows that, whenever a problem is stated, it is possible to ignore existing concepts and to create new concepts, out of the structure of the problem itself, which do correspond correctly to what he calls the subsystems of the adaptive process. By treating each of these subsystems as a separate subproblem, the designer can translate the new concepts into form. The form, because of the process, will be well-adapted to its context, non-arbitrary, and correct. The mathematics underlying this method, based mainly on set theory, is fully developed in a long appendix. Another appendix demonstrates the application of the method to the design of an Indian village. (shrink)

The question of whether humans represent grammatical knowledge as a binary condition on membership in a set of well-formed sentences, or as a probabilistic property has been the subject of debate among linguists, psychologists, and cognitive scientists for many decades. Acceptability judgments present a serious problem for both classical binary and probabilistic theories of grammaticality. These judgements are gradient in nature, and so cannot be directly accommodated in a binary formal grammar. However, it is also not possible to simply reduce (...) acceptability to probability. The acceptability of a sentence is not the same as the likelihood of its occurrence, which is, in part, determined by factors like sentence length and lexical frequency. In this paper, we present the results of a set of large-scale experiments using crowd-sourced acceptability judgments that demonstrate gradience to be a pervasive feature in acceptability judgments. We then show how one can predict acceptability judgments on the basis of probability by augmenting probabilistic language models with an acceptability measure. This is a function that normalizes probability values to eliminate the confounding factors of length and lexical frequency. We describe a sequence of modeling experiments with unsupervised language models drawn from state-of-the-art machine learning methods in natural language processing. Several of these models achieve very encouraging levels of accuracy in the acceptability prediction task, as measured by the correlation between the acceptability measure scores and mean human acceptability values. We consider the relevance of these results to the debate on the nature of grammatical competence, and we argue that they support the view that linguistic knowledge can be intrinsically probabilistic. (shrink)

Contemporary architecture is increasingly grounded in science and mathematics. Architectural discourse has shifted radically from the sometimes disorienting Derridean deconstruction, to engaging scientific terms such as fractals, chaos, complexity, nonlinearity, and evolving systems. That's where the architectural action is -- at least for cutting-edge architects and thinkers -- and every practicing architect and student needs to become conversant with these terms and know what they mean. Unfortunately, the vast majority of architecture faculty are unprepared to explain them to students, not (...) having had a scientific education themselves. Here is an architecture book by an architect/scientist, just in time to help architects in the new millennium. Alexander discusses many of the scientific terms arising in cutting-edge architecture, and explains them to those who don't have scientific training or advanced mathematical knowledge. We find discussions of the evolution of forms; the importance of process in design; iteration; genetic algorithms; sequences of transformations; different levels of scale (i.e. fractals); etc. They are explained here by an architect who is also a scientist, because he wants to change the way architects think and build. Alexander is not merely popularizing other scientists' results and making them accessible to architects: he is in fact presenting new and original scientific work that ties many of these concepts together in a way that will be useful to architects. (shrink)

The 4E approach in (philosophy of) cognitive science—based on ideas that the mind is embodied, embedded, extended, and enacted—is so diverse that it defies straightforward explanation. This paper considers the most ecumenical explanation of the extant concept of the 4E approach. Purported explanations of 4E based on contested definitions of cognition, contrasts with non-4E approaches, or essential and unifying features (including embodiment) either fail to capture the correct extension for the concept of 4E or lead to incoherence. The incoherence is (...) generated by a failure to abide by several desiderata for non-revisionist conceptual explanations: informativeness, non-contentiousness, and identification of any unifying features if they exist. By contrast, a family resemblance conception of 4E constrained by ties of historical influence satisfies these desiderata. The 4E approach should be understood as a set of family resemblances of overlapping ideas, hypotheses, theories and conceptual frameworks about the mind, as well as methods for its study. The family resemblances span several dimensions, including but not limited to a rejection of dualism, non-representational explanations of the mind, phenomenological methods, the importance of embodiment, a dynamical systems perspective, and an evolutionary perspective on the mind. This family resemblance conception is complemented by a historical dimension: the concept of 4E cognitive science emerged in the period following the cognitive revolution of the 1950s. The advantages of this explanation are that it abides by the established use of ‘4E’, explains 4E’s apparent diversity, and warns against reductive explanations. (shrink)

Understanding the origins of human social cognition is a central challenge in contemporary science. In recent decades, the idea of a ‘Theory of Mind’ (ToM) has emerged as the most popular way of explaining unique features of human social cognition. This default view has been progressively undermined by research on ‘implicit’ ToM, which suggests that relevant precursor abilities may already be present in preverbal human infants and great apes. However, this area of research suffers from conceptual difficulties and empirical limitations, (...) including explanatory circularity, over-intellectualisation, and inconsistent empirical replication. Our article breaks new ground by adapting ‘script theory’ for application to both linguistic and non-linguistic agents. It thereby provides a new theoretical framework able to resolve the aforementioned issues, generate novel predictions, and provide a plausible account of how individuals make sense of the behaviour of others. Script theory is based on the premise that pre-verbal infants and great apes are capable of basic forms of agency-detection and non-mentalistic goal understanding, allowing individuals to form event-schemata that are then used to make sense of the behaviour of others. We show how script theory circumvents fundamental problems created by ToM-based frameworks, explains patterns of inconsistent replication, and offers important novel predictions regarding how humans and other animals understand and predict the behaviour of others. (shrink)

The venerable cities of the past, such as Venice or Amsterdam, convey a feeling of wholeness, an organic unity that surfaces in every detail, large and small, in restaurants, shops, public gardens, even in balconies and ornaments. But this sense of wholeness is lacking in modern urban design, with architects absorbed in problems of individual structures, and city planners preoccupied with local ordinances, it is almost impossible to achieve. In this groundbreaking volume, architect and planner Christopher Alexander presents a new (...) theory of urban design which attempts to recapture the process by which cities develop organically. To discover the kinds of laws needed to create a growing whole in a city, Alexander proposes here a preliminary set of seven rules which embody the process at a practical level and which are consistent with the day-to-day demands of urban development. He then puts these rules to the test, setting out with a number of his graduate students to simulate the urban redesign of a high-density part of San Francisco, initiating a project that encompassed some ninety different design problems, including warehouses, hotels, fishing piers, a music hall, and a public square. This extensive experiment is documented project by project, with detailed discussion of how each project satisfied the seven rules, accompanied by floorplans, elevations, street grids, axonometric diagrams and photographs of the scaled-down model which clearly illustrate the discussion. A New Theory of Urban Design provides an entirely new theoretical framework for the discussion of urban problems, one that goes far to remedy the defects which cities have today. (shrink)

On the morning of the third day on Circe's island of Aeaea, Odysseus takes sword and spear in hand and leaves his demoralized and exhausted crew to seek out some sign of habitation. Eventually, from the height of a rocky point, he spies smoke rising in the distance. After debating with himself whether or not to investigate immediately, he determines first to return to his ships, in order to see about his comrades' dinner. Returning to the beach, he encounters an (...) enormous stag, which he takes to have been sent by a god who pitied him and which he kills, binds with a makeshift rope of brushwood and willow branches, and drags back to the camp. (shrink)

The architectural photographer Julius Shulman (1910-2009) is one of the few image makers to have documented, as well as witnessed, nearly an entire century of Los Angeles history.