Introduction
The word information has been used so much that it has come to dominate discourse (Day, 2001). […] Vagueness and inconsistency are advantageous for slogans and using “chameleon words” that assume differing colors in different contexts allows flexibility for readers to perceive what they wish. Buckland, M. (2012). What kind of science can information science be? Journal of the American Society for Information Science and Technology.
I should like to draw an analytical distinction between the notions of “information society” and “informational society,” with similar implications for information/informational economy. The term “information society” emphasizes the role of information in society. But I argue that information, in its broadest sense, e.g. as communication of knowledge, has been critical in all societies […]. In contrast, the term “informational” indicates the attribute of a specific form of social organization in which information generation, processing, and transmission become the fundamental sources of productivity and power because of new technological conditions emerging in this historical period. My terminology tries to establish a parallel with the distinction between industry and industrial. An industrial society (a usual notion in the sociological tradition) is not just a society where there is industry, but a society where the social and technological forms of industrial organization permeate all spheres of activity, starting with the dominant activities, located in the economic system and in military technology, and reaching the objects and habits of everyday life. My use of the terms “informational society” and “informational economy” attempts a more precise characterization of current transformations beyond the common-sense observation that information and knowledge are important to our societies. Castells, M. (2009). The Rise of the Network Society, With a New Preface : the Information Age: Economy, Society, and Culture Volume I (2nd ed)
I look very briefly at our senses of information itself as it is the context for the discussion about informational disciplines to follow. It is certainly a ‘chameleon’ word. It has become so widely used as to become drained of specificity unless explicitly qualified in particular circumstances.
Raymond Williams’ Keywords does not include an entry for ‘Information,’ which is telling. This influential work gave rise to multiple subsequent collections which aim to update it or adapt it to particular domains. I quote from the entries on ‘information’ in three of these here:
Information as keyword — digital or otherwise — did not exist before the twentieth century. […] Then, unexpectedly, in the 1920s this formerly unmarked and unremarkable concept became a focal point of widespread scientific and mathematical investigation. ‘Information’ by Bernard Geoghegan in Peters, B. (Ed.). (2016). Digital Keywords.
Toward the end of the C20 “information” became a popular prefix to a range of concepts that claimed to identify essential features of an emerging new sort of society. The information explosion, information age, information economy, information revolution, and especially information society became commonplace descriptions (Castells, 1996-8; Webster, 2002). These covered, and tried to conceive, disparate phenomena, perhaps unwarrantedly. The concepts appeared superficially to capture similar phenomena, yet on closer inspection centered often on quite different things. For example, their concern ranged over a general increase in symbols and signs that accelerated from the 1960s (the information explosion); the development of information and communications technologies, especially the Internet (the information superhighway, reputedly coined by US vice-president Al Gore), the increased prominence of information in employment (information scientists, information labor, information professions); the growing significance of tradable information (information economy); and concerns for new forms of inequality (the information divide, the information rich/poor).‘Information’ by Frank Webster in Bennett (2005) New Keywords: a Revised Vocabulary of Culture and Society.
As burgeoning use of information in preference to related terms encroaches on the word’s surrounding lexical field, questions arise as to how everything from the human genome to celebrity gossip can so readily be referred to as information. ‘Information’ in MacCabe, C., & Yanacek, H. (2018). Keywords for today : a 21st century vocabulary : the keywords project.
Geoghegan notes the relatively recent general use of the word, and is primarily interested in the information-theoretic work of Shannon and others in the mid-20th century which was important for the development of telegraphy and in subsequent years of computing, cryptography, genetics, network theory and other areas (although was ultimately not very influential in Library Studies or related fields, as Hjørland (2018) points out). The other two excerpts emphasise growing use of the word throughout the latter half of the 20th Century. Webster (a sociologist who has specialized in information-related topics and has written about public libraries) notes the way in which it became attached to various generally descriptive labels, notably of course in the ‘Information society’ and for our purposes ‘Information science.’
The quotes above underline the strong emergence of information-related issues as a topic of investigation, and as an explanatory framework in different contexts. Given this widespread use, any account of information is also going to be provisional and contextual.
As background here, I sketch a very schematic overview of information history which departs from the W. Boyd Rayward (2014) account which influenced it.
This may seem a little hubristic, but I am prompted to insert it here by the common assertion in iSchool materials – accompanied by such phrases as the ‘knowledge economy’ or the ‘information age’ – that the importance of information in our world elevates the work of the iSchool.
However, as Castells notes, this “common-sense observation that information and knowledge are important to our societies,” is not in itself very revealing.
Information has also gone beyond the bounds of any one subject. The chemist or the cultural geographer or the sociologist has an informational perspective. In this context, it seems to me, the promise of the iSchool is not that it has specialist or unique expertise, but that it can bring together a historical perspective and a multidisciplinary focus.
For reasons that should be clear, I do not attempt to define information here. See Bates (2010) for an exhaustive review of definitions, which, more than anything, suggests why a singular view is unlikely.
For convenience, I reference Michael Buckland’s (2017) pragmatic and functional account.
- Information as thing. Informational artifacts – books, passports, menus. Broadly synonymous with Buckland’s inclusive view of ‘document.’
- Information as process. Becoming informed.
- Information as knowledge. Intangible. Fixed in ‘things.’ Imparted in the informing process.
In their reflective overview of information definitions in IS, Dinneen and Bauner (2017) note that “Buckland was aware that the overall account was likely to disappoint the pickiest of theorists.” While a more conceptual characterization—such as Bateson’s "a difference that makes a difference" or Bates’ "the pattern of organization of matter and energy"—might offer additional nuance or insight, it is less well-suited to my purposes. Dinneen and Bauner (2017) favor the recently influential work of Floridi, which also, incidentally, is highlighted in the important information science textbook, Bawden and Robinson (2022).
I reference it mostly because its somewhat technocratic emphasis is convenient. Much of the emphasis of library studies or information science is indeed on the recorded information that can provide a part of the material base for some of the more abstract or general uses above. And also partly because Buckland is such an interesting and historically influential figure in this discussion (librarian, leading Information Science theorist and practitioner, central player in the iSchool movement at Berkeley (Buckland, 2024)).
I take the pivotal mid- to late mid-twentieth Century as a starting point. As noted above, information was foregrounded in a variety of ways at this time, and terms such as ‘information science’ and ‘information society’ emerged. I refer to this as the short age of documents, a reference to the discourse around the challenge of managing recorded information.
Then, we have the long period before this, in which recorded information was manifest in print or manuscript forms. I refer to this as the long age of literacy and print.
And third, we have the short period after this, which we are now living through. While we can characterize this as a digital or network age, the more interesting point here is that an informational perspective becomes more common, extending to social and political contexts. Modern institutional constructs -- markets or bureaucracies for example -- can be seen in informational terms. I refer to this as the informational age (influenced by Castells’ characterization of the current period).
Furthermore, it is now common to reinterpret the past through an informational lens. Perhaps the most interesting recent example of this is Youvel Noah Harari's ambitious The Nexus, which was published recently to mixed reception. He tends to see any ‘intersubjective reality’ as informational. He takes a long historical view, discussing developments as stages in the emergence of information networks. For example, he discusses the difference between democratic and totalitarian regimes as a difference between self-correcting and closed information networks. He talks about civilizations as combinations of a managerial and operational bureaucracy and a legitimating or imposed mythology, each again very much an informational apparatus. He reinterprets the past in terms of an informational present; here he is, for example, talking about the impact of printing: “print allowed the rapid spread not only of scientific facts but also of religious fantasies, fake news, and conspiracy theories.”
In this period also there is a strong emphasis on information critique, and with the advent of AI, perhaps, as I will suggest, we are seeing the apotheosis of the document.
The long age of literacy and print
The emergence of the document marked an important early transition.
Writing allowed thought to be externalized, fixing information in a medium that could persist across space and time. In an oral culture, knowledge had to be remembered. Mnemonic techniques and repetition aided memory. It was retained through song, story and ritual.
As Walter Ong and others have argued, the external documentary accumulation of knowledge co-evolved with a profound shift in the structure of thought, knowledge and communication, evident in the development of more abstract and systematic forms of thinking, the emergence of formal learning and scholarship, the development of laws, the codification of expertise, and so on. This intensified after the invention of movable type.
Libraries are a strongly institutionalized response to the print distribution model, where information or cultural resources were expensive or available in limited ways.
From then until the Second World War, say, information exchange was dominated by the production and exchange of print. Infrastructures and institutions emerged which helped create, manage and preserve documents, including libraries, archives, scholarly societies and publishing, publishers, commercial distribution mechanisms, and so on.
Libraries are a strongly institutionalized response to the print distribution model, where information or cultural resources were expensive or available in limited ways. Efficient access required the physical proximity of collections to their potential readers, and libraries built local collections to make them conveniently available within their communities. In this way, collection size and transaction volumes became a signifier of quality. Those interested in technical, scholarly, cultural or other documents built their workflow around the library.
So, while at a material or mechanical level, we see the progressive intensification and amplification of the production and exchange of documents, a greater variety of ways of processing information, and the massive accumulation of recorded knowledge, the more interesting story involves the mutual interaction between this and social and cultural life.
The progressive connectedness and complexity of social contexts entails progressively more communication across time and space, and a corresponding expansion of the external shared documentary accumulation of expertise and knowledge. Libraries are a part of the apparatus for retaining and sharing that documentary record.
The short age of documents
In the mid-20th Century, the production, circulation, and institutionalization of information expanded significantly, shaped by the ongoing interplay between social structures, technological developments, and organizational demands. The Second World War itself accelerated this, intensifying the need for systematic coordination across scientific, governmental, and industrial domains. There was growth in the scientific and technical literature, while governments and institutions expanded their administrative records, accompanying new forms of bureaucratic surveillance and tracking. Businesses became increasingly reliant on structured data, for decision-making, to optimize workflows, to comply with regulatory frameworks, and so on.
In this way, the volume and variety of documents (again, broadly understood) continued to grow, as did the processes by which they were shared. It is in this period, as noted above, that Information Science emerged, as a response to the challenges of managing this abundance of information in various ways.
Through the sixties and later, there was a focus on the structural changes brought about by “knowledge-intensive production and a post-industrial array of goods and services.” (Lash) This work was informed by the empirical work of Porat and Machlup and others and was given influential synthetic expression as the ‘information society’ or the ‘post-industrial society’ in the work of Daniel Bell. Peter Drucker also popularized the concepts of ‘knowledge work’ and ‘knowledge worker’ during this period.
In successive decades, digital information systems emerged – chemical information, health and legal systems, early knowledge management, and so on.
In this period also there was some modest social institutionalization of Information Science and related areas. The American Documentation Institute (ADI) was founded in 1937, becoming the American Society for Information Science (ASIS) in 1968. It acquired a final ‘t’ for technology in 2000, and finally became the Association for Information Science and Technology in 2013. The Institute of Information Scientists was formed in the UK in 1968, and merged with the Library Association in 2002 to form CILIP. IFIP (International Federation for Information Processing) was founded in 1960.
The evolving informational age
Then it looks at how the logic of information flows reterritorialize into new formations of the brand, the platform, the standard, intellectual property and the network. […] The primary qualities of information are flow, disembeddedness, spatial compression, temporal compression, real-time relations. It is not exclusively, but mainly, in this sense that we live in an information age. Lash, Scott (2002). Critique of information.
In his highly influential three volume The Information Age, Castells talks of a network society and an informational society. I will use this to frame an introduction to the current period, dating from the late 20th Century.
Contributory factors in this period are the accelerated evolution of communications and computational capacity which has provided the material base for a range of other developments. These include the restructuring and intensification of capitalism brought about by deregulation, privatization, global extension and geopolitical changes; the network flows of money, data and people which have changed how we think about the boundaries of organizations, nations and personal relations; and the ongoing transformation of the media, personal communications, and the means of shaping public opinion.
This environment rests on complex network systems, aggregations of data, and applications which communicate via protocols and APIs. This material base has co-evolved with social organization. For Castells (as for Lash and Harari), a key feature is the organizing power of networks, throughout all aspects of what we do.
For example, network effects have led to several dominant platforms that articulate much of our social, cultural and business activities (the Amazoogle phenomenon). Retail, music and entertainment were transformed. The flow of materials is monitored by tracking systems, and is articulated in complex just-in-time supply chains; mobile communications and mapping services have changed our sense of mobility and delivery; distribution chains, the disposition of goods around retail floors, investment decisions, variable real-time pricing, and many other taken for granted aspects of what we do are driven by the collection, exchange and analysis of data.
From a functional point of view, varieties of 'Informationalisation' are visible at all levels in everyday life: doors open automatically, material money is disappearing; advanced instrumentation for observation and analysis is common in the sciences. Increasingly, our activities yield up data which influences what products are offered to us, the news we see, and so on.
Something as apparently simple as the selfie has interacted with behaviors to affect mental wellness, the travel industry and communication.
It is this sense of qualitative change that prompts Manuel Castells to pose the distinction between the ‘information society’ and the ‘informational society.’ We are now seeing an intensification of some of the informational trends he observes as AI becomes more common.
Castells discusses how the network has facilitated broad coordination of interests, in social movements, popular uprisings, or organized crime, for example. It extends to a global scale, where a network of megacities channels power and innovation. He suggests that there is at once a global integration facilitated by the network, but, at the same time, a growing fragmentation between those connected to the network circuits of power and prosperity and those not connected.
More recently, we also see a counter force to global integration. Rather than frictionless global information flow, we are seeing regimes forming around power blocs, with different policy, control and regulatory regimes. Think of the US, EU, Russia and China. There has also been some argument that unequal participation in the ‘knowledge economy’ is a factor in emerging political polarization.
Given this general importance, there has also been an interesting and unsurprising informational development in modern theory. This has come into our field most clearly perhaps in Jurgen Habermas’ concept of the public sphere, but think also, for example, of Anthony Giddens’ concept of ‘reflexive modernity’ or Ulich Beck’s ‘risk society.'
For Giddens' ‘reflexive modernity’ entails the “... the reflexive ordering and reordering of social relations in the light of continual inputs of knowledge affecting the actions of individuals and groups” so that “production of systematic knowledge about social life becomes integral to system reproduction.” Modern life rests on the dynamic reassessment of available information and expertise (in construction, engineering, medicine, technology, …), which builds on the accumulated record of science and technology.
I have chosen to reference the social sciences here, as informationalization and social and cultural change are intricately linked. It is also notable how little reference there is to classical information science in this discussion.
The generative turn: the apotheosis of the document
Instead, these AI systems are what we might call cultural technologies, like writing, print, libraries, internet search engines or even language itself. They are new techniques for passing on information from one group of people to another. Asking whether GPT-3 or LaMDA is intelligent or knows about the world is like asking whether the University of California’s library is intelligent or whether a Google search “knows” the answer to your questions. But cultural technologies can be extremely powerful—for good or ill. Alison Gopnik, 2022.
The current form of generative AI emerged as late as 2023. I find Alison Gopnik’s characterization of it as a cultural technology helpful. She places it in a historical context as the latest technique for passing information from one group of people to another, again considering information very broadly. I was interested to see her place libraries in this frame as well.
Effectively, large language models are statistical models derived from vast accumulations of documentary representations of knowledge. In the context of the narrative presented here, the volume and variety of documents is now so great that they are treated as a proxy for knowledge. Proponents of intelligent AI argue that the models, working with both the broad accumulated representation of knowledge in the training collections, and with massive compute, can find a way to not only summarize and generalize from the content of those documents, but also to replicate the minds that created them.
I tend to Gopnik’s skepticism on this question (see Yiu, E., Kosoy, E., & Gopnik, A. (2024) for an extended argument).
Nevertheless, the processing powers of the models make them very effective for some purposes, and the agentic and applications infrastructure being built on top of them promise to make them more so. We do not know yet whether and where developments will plateau, or how adoption varies by tolerance for hallucinations,[1] or where the impact will be most felt.
However, given the key role of documents (information) in managing complex organizations and interactions, some see the reach of AI as extensive. In this way, the informational, reflexive, networked nature of social life is potentially further intensified.
This extensive informationalization is why Harari, for example, is concerned about the potential reach and impact of AI, as the systemic processability of the connective informational tissue of organizations and systems, he argues, renders them vulnerable to manipulation.
Of course, the ramifications of AI for libraries and for iSchools are accordingly significant. It intensifies some of the trends we have observed, and -- as with other activities at scale -- has both constructive and problematic elements (to use a phrase of Barrett and Orlikowski’s).
If we think of the informational disciplines having a special interest in recorded information, some immediate issues arise.
- Cultural synthesizers. Synthesized content and context add a new dimension and challenge.
- Iterative and chained interaction. We will interact differently with information objects or bodies of knowledge. Think of how larger publishers or aggregators will provide access to the scholarly literature, for example.
- Social confidence and trust. Our sense of authenticity, identity, authorship will all be redefined, creating issues of trust and verification.
- Policy, law and practice will all evolve unevenly in concert.
Information critique
Our simplest actions or interactions now entail complex informational networks and platforms. Think of what is involved in just texting, sending an email or writing in cloud-based Office 365, whatever about group document preparation, remote experiments, or mapping activity.
Day to day behaviors yield up data which is aggregated at scale and used in various ways to monitor, sell or advertise. Large companies have built vast consolidated infrastructure - we are used to thinking of information as immaterial, however, AI has also emphasized how the cloud has boots of concrete. These companies also wield great cultural and economic power - Spotify does our listening for us, Amazon holds sway over merchants on its site.
These social and cultural ramifications mean that undesirable effects are visible and urgent. Addressing these has become an ongoing research, education and advocacy role for the informational disciplines, among others. There is also greater historical sensitivity, an alertness to the ways in which experiences, memories and knowledges may have been suppressed, distorted or invisible (see Benedict Anderson’s classic discussion of museums, maps and other resources in the emergence of nationalism, for example).
Here is a non-exhaustive list of information issues.
- Inequity. Given the centrality of the network and digital resources, differential access creates inequities.
- Surveillance. There is an increase in direct surveillance and also increased collection of data which drives other aspects of our environment. We are generating data shadows which are operationalized in various ways to influence or inform.
- Market concentration. The winner takes all dynamic of network services has resulted in the dominance of several platforms who wield great economic power and influence.
- Dominant or partial perspectives. Perspectives which are historically dominant, or politically motivated, or which reflect imbalances of power and influence may be over-represented in any resource. The plurality of experiences, memories and knowledges is under-represented in the record.
- Dis- and misinformation/’degraded democratic publics’. Our reliance on flows of information has led to concerted attempts to distort, mislead or defraud. Henry Farrell has argued that there is a more fundamental problem, which is “not that social media misinforms individuals about what is true or untrue but that it creates publics with malformed collective understandings.” (Farrell)
- Geopolitical fragmentation. Rather than a global information flow, as noted above, we are seeing regimes forming around power blocs, with different policy, control and regulatory regimes. Think of the US, EU, Russia and China.
Information today – systems of information
Information, then, is at once fugitive and everywhere, chameleon-like.
I noted above how libraries, Information Science, and the iSchool emerged in different phases of the information evolution. And in some ways, they reflect elements of when they emerged.
The library is different from the other two, in that I talk about the library itself as an organization, rather than as a body of knowledge or techniques. Libraries emerged in the first phase described above. Historically, the collection was the core of library identity, as an organized response to accessible distribution in a print world, and to the preservation of knowledge. The library continues as an organized response by cities, universities and others to learning, research and equity of access to the means of creative production. In this way the scope has moved beyond the collection in various ways, as discussed further below. As an organization, the library benefits from education and research in information management topics, but also across a range of other topics (public policy, for example).
Information science emerged, in the second phase, in the mid-20th Century. In this narrative, its origin story splits from the library as information production grows, requiring new methods of organization and access, and it anticipates elements of today’s information environment. As discussed further below, common to definitions of Information Science is a concern with documents (or recorded information, literatures, and similar).
Although, it doesn’t make sense to lean on it too heavily, one might say that Information Science largely retains an information view of the world, concerned with access to and management of information as a thing (in Buckland’s terms).
In the third phase, information is not only seen as something to be managed or discovered, but as an organizing element of social structure and interaction. It has become an object of study across many disciplines and in social and cultural analysis.
The iSchool has emerged in this third phase and it typically embraces a broad set of informational interests. In some ways it subsumes Information Science interests in a very broad view of information in the modern world.
A large part of typical iSchool portfolio is information systems oriented, at undergraduate and graduate levels, meeting needs for workers with technology, business and social skills. It may be a more applied alternative to computer science. It may also encompass expertise in other informational fields (policy, philosophical/social/cultural, data science, HCI, digital humanities, and so on). The broad disciplinary spread also potentially encompasses social and philosophical perspectives as well as very often a strong information critique emphasis.
Borrowing a suggestive phrase from an article by Black and Schiller (2014), one could say that the iSchool is often interested in both information systems and systems of information, ‘systems that create information through social means.’
Of course, the iSchool is not a discipline – it is an evolving academic structure, although, as I have noted, it may be associated with a broad view of Information Science (understood generically not in the classical sense) or Information Sciences. Informatics, a term which emerged in the 1960s (often associated with another term, as in health or social informatics) may also feature.
The focus and disciplinary spread varies across schools.
References
Anderson, B. (2016). Imagined communities: reflections on the origin and spread of nationalism (revised edition). Verso.
Barrett, M., & Orlikowski, W. (2021). Scale matters: doing practice-based studies of contemporary digital phenomena. MIS Quarterly, 45(1). https://doi.org/10.25300/misq/2021/15434.1.3
Bates, M. J. (2010). Information. In M. J. Bates & M. N. Mack (Eds.), Encyclopedia of Library and Information Sciences, 3rd Ed. (Vol. 3). CRC Press. https://pages.gseis.ucla.edu/faculty/bates/articles/information.html
Bawden, D., & Robinson, L. (2022). Introduction to Information Science (Second). Routledge. https://www.routledge.com/Introduction-to-Information-Science/Bawden-Robinson/p/book/9781783304950
Beck, U., & Ritter, M. (1992). Risk society : towards a new modernity. Sage Publications.
Bell, D. (1976). The coming of post-industrial society : a venture in social forecasting. Basic Books.
Bennett, T., Grossberg, L., & Morris, M. (Eds.). (2005). New Keywords: A Revised Vocabulary of Culture and Society. Wiley-Blackwell.
Black, A., & Schiller, D. (2014). Systems of information: The long view. Library Trends, 63(3), 628–662. https://hdl.handle.net/2142/89724
Buckland, M. (2012). What kind of science can information science be? Journal of the American Society for Information Science and Technology, 63(1), 1–7. https://doi.org/10.1002/asi.21656
Buckland, M. (2017). Information and society. The MIT Press.
Buckland, M. (2024). The Berkeley School of Information: A Memoir. https://escholarship.org/uc/item/79v080z7
Castells, M. (2009). The Rise of the Network Society, With a New Preface : the Information Age: Economy, Society, and Culture Volume I (2nd ed). Hoboken, N.J.: John Wiley & Sons, Ltd.
Dinneen, J. D., & Bauner, C. (2017). Information-not-thing: further problems with and alternatives to the belief that information is physical. CAIS/ACSI `17: Proceedings of the Annual Conference of the Canadian Association for Information Science / l’Association Canadienne Des Sciences de l’Information. https://philarchive.org/archive/DINIFP
Farrell, H. (2025). We’re getting the social media crisis wrong. Programmable Mutter. https://www.programmablemutter.com/p/were-getting-the-social-media-crisis
Floridi, L. (2011). The Philosophy of Information. Oxford University Press.
Giddens, A. (1990). The Consequences of Modernity. Polity Press.
Gopnik, A. (2022, July 22). What AI still doesn’t know what to do. Alison Gopnik The Wall Street Journal Columns. http://alisongopnik.com/Alison_Gopnik_WSJcolumns.htm#22Jul22
Harari, Y. N. (2024). Nexus : a brief history of information networks from the Stone Age to AI. Random House, an imprint and division of Penguin Random House LLC.
Hjørland, B. (2018). Library and Information Science (LIS), Part 1. Knowledge organization, 45(3), 232–254. https://doi.org/10.5771/0943-7444-2018-3-232
Lash, S. (2002). Critique of Information. SAGE Publications Ltd. https://doi.org/10.4135/9781446217283
MacCabe, C., & Yanacek, H. (2018). Keywords for today : a 21st century vocabulary : the keywords project. Oxford University Press.
Ong, W. J. (1982). Orality and literacy: the technologizing of the word. Routledge.
Peters, B. (Ed.). (2016). Digital Keywords. Princeton University Press. https://press.princeton.edu/books/paperback/9780691167343/digital-keywords
Rayward, W. Boyd. (2014). Information Revolutions, the Information Society, and the Future of the History of Information Science. Library Trends, 62(3), 681–713. https://doi.org/10.1353/lib.2014.0001
Webster, F. (2014). Theories of the Information Society (4th ed.). Taylor and Francis Group. https://doi.org/10.4324/9781315867854
Williams, R. (2014). Keywords (New Edition). Oxford University Press.
Yiu, E., Kosoy, E., & Gopnik, A. (2024). Transmission Versus Truth, Imitation Versus Innovation: What Children Can Do That Large Language and Language-and-Vision Models Cannot (Yet). Perspectives on Psychological Science, 19(5), 874–883. https://doi.org/10.1177/17456916231201401
Note: I took the feature image in Cambridge.
[1] It is pity that ‘hallucination’ has become the term used here, as it gives a misleading sense of how the LLMs work.
