Collective memory and antisocial media

Regarding the erasure of cultural collective memory, much work – discussed in what follows – has been done within the field of memory studies to explore the mechanics of a shift from a world of power cuts in Coronation Street to a world where no two people see the same content at the same time. Collective or cultural memory – often used interchangeably – has a long history as concepts. The term is usually associated with the sociologist Maurice Halbwachs, who used it to conceptualise what a group can remember and use about its shared past. For Halbwachs, collective memory is ‘what still lives or is capable of living in the consciousness of the group, keeping the memory alive’ (Halbwachs Reference Halbwachs1980). Astril Erll, a leading scholar of cultural memory studies today, illustrates how the terms often function as ‘umbrella terms’ for shared memories of various kinds precisely because the terms seek to draw together often disparate individual memories (Erll et al. Reference Erll, Nünning and Young2008, 2). Rather than assuming shared experiences, when we look to collective or cultural memory, we are looking for points of connection that attach us to each other even as we so often live as disparate and isolated individuals.

This has become harder since the ubiquity of digital culture. While mass media technology often implies shared experiences of the world unified in watching (Gitlin Reference Gitlin1980), Andrew Hoskins has explored how ‘Generative and Agentic AI’s extracting, remixing and replaying of interactions, shards you and your identities anew, rendering it difficult to imagine a group, experience or event, around which a collective memory could cohere’ (Hoskins Reference Hoskins2025). In other words, the increased personal curation of newsfeeds, social networks, audio, and visual media (which reaches its peak with chatbots and LLMs) gives each single individual a bespoke and personalised experience of their world. These increasingly isolating feedback loops keep users in a cycle of interacting with their unique machines and thereby inhibit or prevent shared cultural memory from being developed in a coherent way between individuals. For Danny Pilkington, this individualisation of memory that is encouraged by our technologies has its cause in both capitalist and neoliberal principles of individualism. He argues that the growing role of AI and algorithmic technologies is ‘shaping our engagement with society along […] lines of individualism. Ultimately, it is his conclusion that ‘the scope of memory is being reduced and constrained within the prism of capitalism’ (Pilkington Reference Pilkington2024). In short, it has been convincingly illustrated that some of our common practices online today have the effect of reshaping our experience of memory in the service of individualistic capitalist and neoliberal models of society.

Similarly, seeing the shift in memory as reflecting an increasing individualism, Hoskins calls this shift in cultural memory ‘the connective turn’, which he describes as ‘a paradigmatic shift in the treatment and comprehension of memory and its functions and dysfunctions’ that is ushered in by new digital technologies (Hoskins Reference Hoskins2011). After such a turn (and despite the language of social media), we ‘share’ much less with each other than we did before, as we become more reliant on a digital and individual rather than an analogue and collective memory. Focusing on social media, Jacobsen and Beer discuss how media platforms that hold vast amounts of biographical data about our lives ‘repackage our past content as ‘memories’ and deliver them back to us (Jacobsen and Beer Reference Jacobsen and Beer2011, 117). This amounts to something like an ‘outsourcing’ of memory to our devices, when previously memory would more likely be outsourced to communities, friends, and families with whom these memories are ‘shared’.

As such, the increasing reliance on digital platforms – even when they present themselves as social and connective – serves to disconnect and dissocialise our memories. The term ‘digital memory’ has been used to explore this often-contradictory shift that has occurred in practices of memory since the ubiquity of digital media. Introducing the idea in Digital Memory Studies, Hoskins notes that what is at stake with the idea of ‘digital memory’ is the contradiction between a contemporary media that claims to be ‘participatory’ and the fact that our engagement with this media produces cycles of compulsive behaviour that only symbolically involve ‘sharing’ (Hoskins Reference Hoskins and Hoskins2017, 2). In a sustained study of Hoskins – whose work is principal to the development of the field of digital memory studies – Robin Ekelund summarises what he sees as the key aspect of Hoskins’ position:

From this perspective, there is something uniquely antisocial about social media and the LLM chatbots and AI companions that dominate today. While social networks keep us connected in some ways, when it comes to memory, they serve the function of disconnection.

In a contribution critical for conceptualising the changing nature of memory, Linden, Hachem, and Kondyli link the changes in collective memory today to ideas of ‘play’ as they were developed by theorists such as Roger Callois, Johan Huizinga, and others. Building on the idea prominent since the 90s that memory is collective, rather than individual (Halbwachs and Coser Reference Halbwachs, Coser and Coser1992), Linden, Hachem, and Kondyli borrow the concept of children growing up surrounded by an ‘attachment village’ (Neufeld and Maté Reference Neufeld and Maté2013). These are groups of relatives, friends, and neighbours – perhaps even including institutions and organisations like schools, churches, and community centres – that individuals learn from and rely on as they develop from childhood. These institutions or communities serve as collective memory playgrounds where young (or old) individuals can play with ideas of identity, stories of personal, and collective history and belonging, and the idea of collectively living and remembering. In other words, in an attachment village, the feedback loops between an individual and their communities are responsible for building their sense of being in relation to others.

Linden, Hachem, and Kondyli point out that ‘in modern nuclear families, youngsters are surrounded by screens that provide information and, more recently, by smart devices to interact’ with. This change from an attachment village to a machinic agent is framed as a change in the nature of play. Proposing the idea of ‘prompting as play’, they conceive of a key change in the way young individuals test out and play with ideas and curiosities. While the children of Huizinga’s time used playgrounds, family meals, and community activities as a ‘testing ground’ to play with knowledge and interaction, the children of LLMs and agentic AI use the input boxes of ChatGPT and Gemini as their playground. As a result, rather than developing identities in relation to others in their attachment village, these users develop identities in relation to their devices. Rather than feedback loops reaching out into and returning from a community of others, these are feedback loops between the individual and their personalised algorithms that cut out the role of other human agents in the process of playing and remembering.

In Antisocial Media, Siva Vaidhyanathan discusses this personalisation of content in terms of politics. Vaidhyanathan notes many ways in which the structure of Facebook’s curational tools serves to move us away from being a society that can collectively develop shared rules and habits. He writes that ‘when we combine the phenomena of propaganda with filter bubbles, we can see the toxic mixture that emerges: Facebook users are incapable of engaging with each other upon a shared body of accepted truths’ (Vaidhyanathan Reference Vaidhyanathan2021, 7). In light of the above, we might draw a parallel with the role of play, which can be conceived as critical in the social collective rule-making of children. Vaidhyanathan focuses on how this lack of a social body of shared rules can be manipulated for political purposes, while Merrin and Hoskins point out that Facebook in particular had jumped keenly on these technologies but after the rise of the far-Right on these platforms, from 2016 the ‘backlash’ caused Facebook to make a U-turn and ‘burned by the issue of politics, re-oriented its news-feed algorithm towards groups and non-political content’ (Merrin and Hoskins, 86). As Merrin and Hoskins point out, the depoliticisation of Facebook far from solves the problem: whether the content of our media is political or not, its form – developed in our political and economic context – that is antisocial. An effect of this antisocial technology is that users perceive their world as internally rational and consistent to themselves without reality-testing this with others. The problem, then, is not so much that we have a chaos of disinformation and misinformation but that ‘each sharded individual exists in a personally-formed and encompassed informational bubble that does afford them order and intelligibility’ (Merrin and Hoskins, 12). This personal bubble of intelligible media might, as we shall see, lead into psychosis.

These are some of the key contributions in memory studies and adjacent fields that have begun the work of identifying the impact of technology (and its economic drivers) on memory, of which there is much more work already done. Taken together, they sketch out a picture of modern life in which it is clear that our habitual patterns of cognition and recognition are subtly and importantly changing. In general, we can conceptualise this change as one favouring increasing personalisation and individualisation over social and connective collective memory-making habits. The next section turns to the question of how we can theorise the effects of this shift. If, for Danny Pilkington, the cause of the changes that we are witnessing in practices of memory is capitalism and neoliberalism, I argue in what follows that the effects of the changes that we are witnessing in practices of memory include psychosis.

Introducing psychosis and foreclosure

Psychosis is a clinical category that is used in the diagnosis of just 1.5–3.5% of Americans and approximately 0.7% of Europeans, but the use of the term is going up sharply. It is applied in a range of contexts to describe complicated and overlapping symptoms, but it is usually associated with (a) hearing voices, (b) experiencing intense fantasies that are difficult or impossible to escape, and (c) having a deep feeling of loneliness and severe isolation. While we all have intrusive thoughts and irrational fears and dreams, in a psychotic experience, the subject cannot distinguish from – or even doubt – the reality of these fantasies. They appear as indisputable truths to which the subject feels deeply connected. The psychotic feels immensely frustrated and isolated by the sense that others cannot share their experience of reality.

The symptoms of psychosis – isolated fantasies, strong convictions, and experiences of intense loneliness – might also describe the experience of those with extensive reliance on social and digital media. As mentioned in the introduction, in 2025, OpenAI hired a forensic psychiatrist to combat the number of users who had reported sliding into mental health crises after using the software. In the same year, PsychologyToday reported on the first death from ‘ChatGPT-induced psychosis’, and the most popular thread on Reddit was a sharing circle of those whose husbands and wives had been driven to psychotic episodes by newly available LLMs and AI (see Marchiano Reference Marchiano2025).

Formally speaking, psychosis means something more specific than these general and broad symptoms, and many of these recent reports are anecdotal and based on conjecture. Stijn Vanheule’s book Diagnosis and the DSM (Vanheule Reference Vanheule2014) shows how many versions of definitions for psychosis there are – even in official medical language – and how, over time, it has become increasingly diffuse and broadly applied.

The theory of psychosis in psychoanalysis was developed by Sigmund Freud in the early 20th century, most notably in his famous text Psycho-Analytic Notes on an Autobiographical Account of a Case of Paranoia, written in 1925. His concept of psychosis was then advanced most significantly by Jacques Lacan, perhaps the most influential follower of Freud, in his important seminar on psychosis given between 1955 and 1956. It was then further extended in the 1990s by Jaques-Alain Miller, founder of the École de la Cause Freudienne, which continues to be one of the biggest psychoanalytic institutions today, whose work culminated in his 2008 essay ‘Ordinary Psychosis Revisited’. Since then, it has been treated by many psychoanalysts and theorists whose ideas are used in this article. Despite their differences, for most of these writers and practitioners, psychosis has a clear and particular meaning. Psychosis is a term that denotes one of just three ‘subject positions’, along with neurosis and perversion, into which almost all of us can or could be categorised. Recently, Leon Brenner has argued for a fourth category: that of the autist (see Brenner, Reference Brenner2021). For those in psychoanalysis, compared with neurotic and perverse subjects, psychotic subjects have a particular psychic structure.

To put it simply, psychosis, in this formal psychoanalytic sense, involves an externalisation of internal trauma. This causes the psychotic subject to perceive a gap in reality and subsequently begin to produce fantasies that promise – often with more and more farfetched possibilities – to close this gap. Neurotics, by comparison to psychotics, tend to internalise and locate a lack, gap, or inadequacy in themselves. While the language of psychoanalysis can often seem technical and specific, in one sense, it is more general than other scientific uses of the term. Psychosis, in this sense, is a way of describing patterns or structures that we see in the world and how we position ourselves in relation to others. It is this sense of psychosis as a subjective position that we can occupy.

The psychoanalyst Jacques Lacan developed a new concept to explain how the psychotic relates to their external world in this fantastical way: the concept of foreclosure. Foreclosure describes a type of repression that is unique to the psychotic individual, which involves burying their sense of failure and inadequacy not internally within their own unconscious (as others might) but projecting it out into the world. In other words, they perceive a gap in the world rather than in themselves. For Lacan, the father often stands in for the law or for prohibition because he provides the initial ‘No’ – the ‘No-of-the-Father’ – which insists that the young subject separate from the Mother. We can think of this straightforwardly as the first time the father separates the child from the mother’s breast: the father often operates as our first encounter with a social world of more than two people (the baby and its mother).

But the father does not have to stay by his child through their entire life, preventing them from trying to reconnect with the mother at each step. Instead, in the normal development of the child, it learns to displace and sublimate its original desire for the mother from which it has been separated into other forms of desire that are acceptable to social norms, codes, and laws. In doing so, we start to learn the codes and rules of social life. We may come to desire a close friendship with a girl or a boy in our class, or we may come to passionately pursue a sport, or artistic project, or academic success. In each unique case, we learn to work within social rules towards a pleasurable goal. This could be comparable to Gabor Maté’s concept of the attachment village discussed in the previous section: these are the rules and agreements of a community that the child learns to operate within.

These norms are referred to as the ‘Name-of-the-Father’ because they replace the rule-setting role the father initially represented. They allow us to channel our desires into the social fabric throughout our adult life – circling their desire around various stand-in objects that operate as replacements for the original lost desire of the child before its socialisation into language and society. While the father does not have to stand by to say no at each point, the Name-of-the-Father persists to psychically remind the subject of its rules and duties.

To understand psychosis in Lacanian terms, it is necessary to see the difference between repression and foreclosure. Through the process of repression, an element of signification is cast into the unconscious, where it continues to threaten the conscious subject. On the other hand, through foreclosure, something is rejected from the symbolic entirely with no hope of return. Then, the subject indulges in more and more deeply held fantasies which appear to speak to this irreparable hole not in the subject themselves but in external reality. Lacan writes that ‘in psychosis, reality itself initially contains a hole that the world of fantasy will subsequently fill’ (Lacan Reference Lacan1997, 45). In foreclosure, something is banished not to the unconscious but from reality itself, leaving a hole in the world. Psychosis, as we have seen, involves an externalisation of internal trauma, so that the subject perceives a gap in reality itself and subsequently begins to produce fantasies that promise to close this gap. To properly consider the structure of the psychotic and the effect that foreclosure and a lack of the Name-of-the-Father have on the subject, it is helpful to look at the origins of the concept and the first case of psychosis in psychoanalysis.

The original psychotic

Freud first treated the topic of psychosis in 1895, though his theory was only fully formulated when he discussed it in depth in his 1909 essay ‘Notes upon a Case of Obsessional Neurosis’. In this foundational text, he studied the case of Judge Daniel Schreber, a highly intelligent and cultured man who produced and published a vivid account of his nervous illness dominated by the desire to become a woman, terrifying delusions about his doctor, and a belief in his own special relationship with God. These were published as his memoirs under the title Memoirs of My Nervous Illness. Freud would see Schreber as a schizophrenic, and he would use him to develop his conceptual understanding of psychosis.

Freud’s analysis of Schreber was grounded in the identification of two key components of Schreber’s psychotic delusion: his desire to become a woman and his desire to be subsequently impregnated by God. Schreber believed that he would be impregnated by God and give birth to a new race of men that would solve all the inconsistencies in reality. In his memoir, he reports that he woke up one morning with the thought that ‘it really must be rather pleasant to be a woman succumbing to intercourse’ and wondered whose idea that was or whose voice or desire he had intercepted (Schreber Reference Schreber2000, 63). Freud, fascinated by the expression of desire to be impregnated as a woman, writes:

For Freud, the narcissistic fantasy of being the godlike figure of the Redeemer is the primary drive of the delusion, with the desire to be a woman a secondary desire presented as an inevitable means to the end of his ultimate desire – to be impregnated by God.

This conclusion is supported in Schreber’s own memoir, in which he writes that he sees his emasculation as a necessary and inevitable step towards his appointed fate as the carrier of God’s child. The impregnation as a woman is a ‘means to an end’ vis-à-vis his ultimate fantasy of restoring the order of society in the role of Redeemer:

It is worth noting here that Schreber uses the phrase ‘Order of Things’ because it shows that his fantasy about himself is related to an entire worldview. It is supported by a conviction that the world is wrong and that it must be put right, and by his own, particularly appointed role in the achievement of that correction to the Order of Things. Schreber sees himself as tasked with playing a particular role in the restoration of world order. This is a prominent feature of psychosis – even in the terms used in clinical discourse today– and a parallel can be drawn between this and the sycophancy of LLMs and its ability to reinforce a narcissistic impulse (see Clegg Reference Clegg2025) and perhaps to the increase in conspiracy theory on online platforms, whose thinking is often based on a fantasy of a restoration of the order of things (through revealing a secret truth hidden from others) and has been connected to psychosis in recent medical studies (see Veling et al. Reference Veling, Sizoo, Van Buuren, Van Den Berg, Sewbalak, Pijnenborg, Boonstra, Castelein and Van Der Meer2022).

The attachment to bigger things is a strong characteristic of psychosis, which often involves delusions of grandeur, Napoleon complexes, and Redeemer fantasies. On the contrary, neurotic subjects tend to focus their mental energy on smaller objects – the obsessive love of another person, a penchant for books or flowers, or an attachment to a teddy bear, a photograph, artwork, or a family heirloom. In 2009, Thomas McGlashan revisited the Schreber case, arguing that psychosis is the result of reduced ‘cathectic’ capacity (McGlashan 2009). Freud uses the term cathexis to refer to the investment of mental or emotional energy in a person, object, or idea, and for him, the process is normal and necessary for the subject to live with other people and objects in the world. If we develop a love of collecting stamps or vinyl records or an attachment to Pokémon, this indicates that we have the neurotic ability of cathexis that allows us to relate to objects and to other people. The psychotic, on the other hand, struggling to cathect to objects, people, or ideas, relates to the bigger picture, focusing on their own individual relationship to whole worldviews and ideas of the greater order of things. In a digital social world that encourages individual thinking in lieu of collective social life, we perhaps see less ability to relate to others and more tendency to think of ourselves in relation to larger worldviews.

If and when the world is corrected, the psychotic subject imagines that they would experience an ultimate form of pleasure. One of the most important parts of Freud’s analysis of Schreber is when he notices the connection between the religious delusion and the gender delusion in his fantasy. He does so via the concept of bliss.

Here, Schreber’s fantastical solution to the problems in the world is to see himself as especially appointed as God’s wife. He believes that a restoration of the order of things and of his place within the system will ultimately make it possible to experience a future bliss – a bliss that promises to close the gap at the heart of identity – but only by closing the gap in the disordered world in the resumption of the ‘Order of Things’. Like a sleeping beauty dreaming of the magic kiss from a prince, this anticipated bliss would represent a point of desire located in the future of his fantasy – a point that will never come but to which whole lives can be (tragically) dedicated.

Ordinary psychosis

This situation no doubt appears extreme: we are not all psychotics in the way that Judge Schreber might have been. In 1998, Jacques-Alain Miller first proposed the concept of ‘ordinary psychosis’. For Miller, psychosis becomes less associated with deviations from norms and realities and more with a normal, ‘ordinary’, position that we tend to occupy as subjects today. But there is a difference between Miller’s ordinary psychotic and the original psychotics discussed above. While the traditional psychotic subject has no Name-of-the-Father, in ordinary psychosis, Miller writes, ‘we have no Name-of-the-Father, but something is there, a supplementary device (Miller Reference Miller and Natalie2008, 161), which allows us to function. ‘You always have something that enables the subject to get away or to continue to survive’, writes Miller‘, and in some way the true Name-of-the-Father is no better than this well-fitting make-believe’ (Miller Reference Miller and Natalie2008, 161). For Miller, we are all now living as psychotics, without access to the Name-of-the-Father (without agreed social rules and habits), but with a supplementary device which keeps us just on the right side of the line and, at least to an extent, in touch with reality. If full psychosis involves isolation and the absence of society, ‘ordinary psychosis’ combines this isolation and antisocial experience of the world with the illusion of or impulse towards a collective social experience.

The supplementary device that replaces the Name-of-the-Father is a ‘make believe’, but it allows us to continue without descending into full psychosis. Perhaps the contemporary embodiment of this illusory make-believe is the idea of ‘social’ media or the implication that our engagement with AI is collective or collaborative. These antisocial technologies, which appear to connect us, supplement their isolating and polarising effect on us by promising us connection, which operates in lieu of a social world or community, but keeps us just on the right side of psychosis. Due to the move from collective to individualised memory, we lose a certain function of society (in psychoanalytic terms, the Name-of-the-Father), but we replace it with a make-believe illusion of a social or connected world. It might not be a coincidence that Miller writes of the normalisation of psychosis at the same time that the first iPhone was released (2007) and in the early boom years of social media, marking a significant point in the omnipresence of smartphones and social profiles that both promise connection and collectivity and – as we have seen – produce antisocial isolation.

We see this pattern in conspiracy theory networks like those of QAnon or of The Flat Earth Society, or in those whose addictions to generative AI tools like ChatGPT or Replika result in serious romantic relationships with algorithms. Conspiratorial thinking online involves isolation and fantasy (even comparable to the original psychotics), but it also involves the attempt at social connection through forums and group-making and the impulse to convince and include others in the fantasy. Similarly, users who develop relationships with simulators or chatbots tend to do so while seeking or desiring a connection with other humans. In late 2025, the Chinese government imposed a ban on romantic AI companions for minors, leading to a backlash from those in such relationships, with many claiming their genuine social connections had been ‘deleted’ (see Chia Reference Chia2025). In this sense, when we engage with the personalised feedback loops of search engines, curated newsfeeds, companion simulators and generative AI tools, we are asked to occupy the position of the psychotic alone with their fantasies, even though at the same time we are often seeking social connections when we use these tools. The ordinary psychosis of today, then, involves being caught between the tendency towards isolated fantasy and the desire to hang on to shared reality.

Psychosis and AI

The way that AI is trained upon and organises data is critical to the effects it has on the memory habits of its users.Footnote ¹ Data archives (user data, behavioural logs, sensor data), knowledge bases (AI model weights, code libraries, organisational records), and infrastructural memories (cloud servers, documentation, metadata systems) inform how a program such as ChatGPT personalises itself to each user so that ultimately each person is excluded from every other person’s search results and responses. Ultimately, then, while the model itself is built on the collective synthesised data of all user histories that we can think of as a collective memory, at the user-facing end, the program ‘shards’ each user’s experience from all others to ensure that engagement with the tool encourages a personal individual experience of memory.

In fact, since at least 2024, OpenAI has been working directly on the role of ‘memory’ in the development of ChatGPT. It has justified its practices through the offer of transparency and control to users: in 2025, they wrote to users that ‘you are in control of ChatGPT’s memory’ and that to ‘see what ChatGPT remembers about you, you can also ask it’ (OpenAI 2024). Yet, they actively pursue the increasing personalisation of memory histories for each user: ‘the more you use ChatGPT, the more useful it becomes. New conversations build upon what it already knows about you to make smoother, more tailored interactions over time’ (OpenAI 2024). The issue here is that control and transparency do not solve the problem, nor do they even make it particularly visible. Since the merits of personalisation have been effectively sold to users, they often willingly contribute to the development of personal tailoring strategies for their own convenience, often without attention to the effects of this on cultural and collective memory.

In the introduction, I suggested a type of memory which isolates rather than connects and which lends itself to the type of fantasy thinking that is prominent in psychotic subjects. We have now seen how the type of computation deployed in the development of AI tools lends itself to the encouragement of this form of memory rather than to the shared collective memory, which it might be more likely to encourage in an environment not driven by the accumulation of capital. Alongside this, we have speculated that the psychoanalytic concept of psychosis can help us theorise the effects of this isolation and separation, which lead us into less social and more individualised networks, resulting in loneliness and often in personal fantasy and delusion.

The discourse around the impact of the internet on its users has been dominated in many cases by the idea of attention span and distraction. Within academic research, ‘attention studies’ have developed quickly into a significant field. Such approaches generally focus on the effects of the user’s experience on their own individual ability to consume, react, and actively or passively engage with the content they consume (Castro and Pham Reference Castro and Pham2020, 2, see also Wu Reference Wu2022). Yet the effects of these technologies cannot be understood by analysing the feedback loops between the individual users and their machines alone. What a memory studies approach adds is attention to the relationship between users and their wider community. By framing the question of the impact of technology on memory as a change in the relationship between the individual and the collective, these approaches highlight how the effects of these technologies are not found only in the feedback loops between the user and their machines but in the changes in the fabric of their social relations. In other words, the question is often one of what is not happening when such technologies become ubiquitous as much as it is about what is.

In psychoanalysis, psychosis is ultimately caused by a failure to effectively integrate into the social order. As the English psychoanalyst of the 1970s, Donald Winnicott, points out, psychosis should be seen less as evidence that something has gone wrong and more as evidence that something has not gone right. For Winnicott, ‘it is wrong to think of psychotic illness as a breakdown, it is a defence organisation relative to a primitive agony, and it is usually successful’ on the part of the subject when they develop in a ‘deficient environment’ to avoid psychosis (Winnicott Reference Winnicott1974, 103). In the context of this article, we might suggest that a world in which cognition and recognition are individualised and personalised is a deficient environment, one without sufficient collective and social memory, which may push those people living in it towards increasingly psychotic thinking.

Though it does not approach the topic from a clinical perspective, this article proposes that in light of the changes in the nature of memory that have been discussed in memory studies over recent decades and in light of the technical structure of online computing today that is dominated by personalisation and curation, the concept of psychosis can help us conceptualise the effects of these changes on a generation of individuals who depend increasingly on these tools. If there is any validity to the idea that these technologies encourage psychosis, this should serve as a warning that the loss of collective and social memory could have a serious impact on a generation whose socialisation involves more interaction with algorithms than with their communities.