AI Slop and Pseudo-History on YouTube
AI Slop and Pseudo-History on YouTube: When History Becomes Sensationalist Content
AI Slop and Pseudo-History on YouTube: When History Becomes Sensationalist Content
DESCRIPTION: AI history videos on YouTube market sex, violence and secret knowledge as education, and turn every gap in the sources into a conspiracy. An analysis drawing on Lacan, Haug, Baudrillard and research into illusory truth.
History as a commodity
AI drivel, pseudo-history and the allure of secret knowledge on YouTube
On YouTube, a form of pseudo-history is emerging, fuelled by AI scripts, synthetic voices and generated images. It sorts the past according to its sensational value, peddles sex, violence and secret knowledge under the guise of education, and turns every gap in the historical record into a conspiracy. Here, this format is analysed through the lens of Lacan, Haug and Baudrillard, alongside research into conspiracy thinking and illusory truth.
‘The Romans’ secret sexual practices’. ‘What historians are keeping quiet about the Vikings’. ‘The most cruel punishments of the Middle Ages’. ‘The truth about the Templars’. Anyone who clicks on such titles does not initially know who they come from, what the claims are based on, or whether they contain any verifiable assertions at all. The thumbnail might show a woman in an artificially antique setting, a barbaric-looking warrior, a blood-stained dungeon or a parchment page bearing a seal. Beneath the image is a title that combines knowledge and taboo. The video promises both enlightenment and titillation.
The voice begins with a claim. It sounds calm, knowledgeable and free from doubt. An emperor is said to have done this, a people is said to have performed that ritual, a court is said to have been characterised by sexual debauchery, a church is said to have removed certain facts from its archives. Whilst the voice speaks, a sequence of generated images flashes by. They show people who never lived, in rooms that never existed, engaged in actions that no source describes. The images are not consistently absurd. They are too polished for that. They combine familiar elements of the past with a visual style familiar from TV series, computer games, adverts and television documentaries. It is precisely for this reason that the impression arises that the voice has made something visible.
AI Slop on YouTube: Pseudo-history as a series
This pseudo-history is recognisable by its consistent approach. Real names, dates and historical events provide the material. A language model fills in the context, a synthetic voice transforms it into a narrative, and images provide scenes that seem to race ahead of the sentences. The viewer sees a Roman meal whilst hearing about Roman sexuality. They see a torture chamber whilst medieval criminal law is being discussed. They see an artificially aged face whilst a video explains a ruler’s secret motives. The scene brings the claim to life, even though it provides no evidence.
Inaccurate and sensationalist historical narratives existed long before the videos. Early mythology explained the origins of peoples and cities through gods, giants and monsters. Herodotus reports on ants in India, larger than foxes, which dig gold-bearing sand from the earth. Plutarch himself stated that he wrote biographies, and embellished them with omens, dreams and anecdotes. Flavius Josephus has 960 besieged people at Masada die together and puts two long speeches into the mouth of their leader, which no one could have written down. The biblical authors recounted the history of Israel as a story of salvation, complete with miracles, plagues and a sun that stood still over Gibeon. The *Physiologus*, thought to have been written in Alexandria between the 2nd and 4th centuries, transformed zoology into a doctrine of faith: the pelican revives its dead young with its own blood; the lion sleeps with its eyes open. Jacobus de Voragine’s *Legenda Aurea*, compiled around 1260, collected the lives of saints full of miracles and became one of the most widely read books of the late Middle Ages. In the 16th century, the ‘Neue Zeitungen’ (New Newspapers) reported on leaflets featuring woodcuts, birth defects, blood rain and celestial phenomena as signs of the approaching Day of Judgement. Pseudoscience and conspiracy literature are just as old.
The only things that are new now are the sheer volume and the imagery. Suetonius, who wrote about Tiberius on Capri, and Procopius, who slandered Empress Theodora in his *Secret History*, were already familiar with sexualised narratives. AI turns these into historical pornography. The term is modelled on the English ‘disaster porn’ and ‘trauma porn’, which refer to the sensationalist display of suffering, and here also carries the literal, sexual connotation. For every sentence about Roman sexuality, the AI provides the appropriate body; for every torture chamber, the victim; and the channel thumbnails are designed specifically to feature these images. What the images fail to deliver, the conspiracy narrative makes up for. Generative AI also allows scripts, voices and images to be produced in series. A channel can apply the same template to Rome, Egypt, the Viking Age, medieval courts, the Second World War or ‘lost civilisations’ without having to undertake a lengthy study of any particular era. The narrative simply needs a theme that offers enough well-known names and enough scope for mysteries.
In July 2025, YouTube renamed its rule against ‘repetitious content’ to ‘inauthentic content’. It targets mass-produced videos created solely to generate views. History channels remain unaffected by this. Every video has a new theme, a new name, a new century. The only thing that is repeated is the process, and the rule does not cover that.
The past, sorted by sensational value
The historical significance of a topic counts for little in this process. What matters is whether it can be used to create a title that triggers curiosity, revulsion, outrage, fear or the desire for exclusive knowledge. The past is sorted by its sensational value. It appears as a storehouse of scandals, secrets, violence and sexuality. An era becomes a ‘dark’ century, a dynasty an incestuous drama, a religion a secret cult, an administrative issue a conspiracy, and a gap in the historical record suppressed knowledge.
This selection has consequences. Anyone who continually consumes history in this form encounters a past in which everything revolves around crime, decadence, hidden practices and secret powers. Historical societies appear as a playground for extreme scenes. The social order, legal distinctions, economic dependencies, concepts and self-interpretations of the time disappear behind the spectacle. What remains is the cruel emperor, the erotic slave, the fanatical priest, the barbaric warrior, the corrupt court and the victim in the dungeon.
Such images organise the past according to the expectations of the present-day gaze. Antiquity is supposed to be luxurious, sexual and cruel. The Middle Ages are supposed to be filthy, superstitious and violent. The ‘Orient’ is supposed to be mysterious, erotic and despotic. The Viking Age is supposed to deliver masculinity, snow, blood and raw freedom. These visual patterns stem from a long history of novels, paintings, adventure films, colonial fantasies, historical series, computer games and advertising. AI uses these to generate new variations of a past that has already been shaped by the media.
The clips and videos therefore often look like short trailers for a series that never existed. Every change of scene offers a new promise: something momentous, forbidden or revealing is about to be revealed. The historical subject matter functions as a brand and requires no accuracy. ‘Templars’ stand for secrets, conspiracy and hidden treasures. The video can rely on these cultural shorthands because they are already present in the audience’s minds.
Education as a licence for sex and violence
Clicking on ‘The Secret Sexual Practices of the Romans’ is regarded as education. This is a crucial part of the format. The title sells a sexualised scene, but the word ‘history’ sanitises the consumption. Anyone viewing erotic images is learning about antiquity. Anyone clicking on torture is learning about medieval legal practice. The historical framework lends the allure a culturally acceptable form.
People open such videos out of curiosity, because of a specific historical interest, out of boredom, or because the platform presents them with these videos. The point here concerns the logic of the format. It selects its topics in such a way that education and titillation overlap. The viewer can be shocked, outraged or fascinated, whilst at the same time telling themselves they have learnt something.
Historical distance reinforces this licence. One can view past violence without encountering a history of present-day violence. The sexual exploitation of enslaved women can become a visual motif because the woman depicted appears as part of a backdrop called Rome. The torture of a prisoner in the Middle Ages becomes a detail of a grim world that can be consumed. The suffering remains real. The distance of the centuries makes it easier to transform that suffering into visual material.
The pattern is an old one. The Roman arena celebrated the emperor’s generosity and the order of the empire; the execution grounds of the early modern period celebrated justice and deterrence. Both presented suffering as a spectacle and provided the very reason why one was permitted to look on. The AI video digitally revives this arrangement. Its justification is called education, and it caters to the same sadistic voyeurism. In ‘Drives and Drive Fates’, Freud described sadism and voyeurism as parallel pairs of drives, with their passive counterparts, masochism and exhibitionism. The historical framework provides both with a stage on which they appear as a thirst for knowledge.
Three differences reveal what the format has changed. The arena was a collective ritual in which the crowd watched itself and reaffirmed its order. The video is watched alone, in bed or on the train, and no one sees the viewer. Those being tortured have been dead for centuries and the images are synthetic, so no one suffers for the sake of the scene any longer. This makes the pleasure cheap and arbitrarily repeatable. Finally, the video reverses a movement that Norbert Elias described as a process of civilisation: cruelty was relegated to behind the scenes over the course of centuries. Elias recalls the burning of cats at the Feast of St John in 16th-century Paris. Foucault begins *Discipline and Punish* with the torture of Damien in 1757 and shows how the spectacle of punishment subsequently disappeared. The Slop brings torture back onto the screen. History itself serves as its backdrop.
Here, the critique touches on a problem that goes beyond the question of individual errors. AI-Slop transforms historical figures into visual archetypes for contemporary needs. The cruel ruler reinforces the belief that earlier times were morally inferior. The exoticised woman permits a sexualised gaze upon a past rendered alien. The barbaric ritual turns a culture into a theatre of revulsion and superiority. The fanatic confirms the present’s self-portrayal as rational and enlightened. History becomes material through which viewers can simultaneously derive excitement and achieve moral purification.
The gaze in Lacan: witness to a forbidden world
Lacan’s theory of the gaze offers a concept for this. For Lacan, the gaze originates from the image: the picture plane determines what the viewer is permitted to see and how they relate to the visible. The AI video presents violence, sexuality and otherness in such a way that the viewer can appear as a witness to a forbidden world. They see what they are supposedly not meant to see, and they see it under the protection of the claim that they are merely imagining it. The image binds them to the visible whilst simultaneously allowing for distance: that was a long time ago, those are other people, that is history.
The concept of the ‘gaze’ would only be meaningful here if it described the specific pictorial order. AI images place the viewer in a specific position vis-à-vis the past. They stand before the scene as someone who is both in the know and morally superior. They now know how cruel people were back then, how sexually uninhibited an imperial court is said to have been, or how secretive the rituals of a foreign culture were. The past becomes the object of a mixture of revulsion, lust and self-assurance.
‘What Historians Conceal’: Secret Knowledge as Status
This mixture also explains the power of the phrase ‘What historians keep quiet about’. Beyond the information itself, it offers a position within a social hierarchy. The viewer gains access to knowledge that others supposedly do not possess. Historians, schools, museums, archives and the media appear as institutions that omit something or deliberately conceal it.
This gesture begins even before the first sentence of the video. The title assigns the roles. Here stands the courageous whistleblower. There stand the institutions that remain silent. The viewer is addressed as someone who is no longer willing to be fobbed off with the official history. Their ignorance is attributed to a culprit. What would normally be the consequence of the fact that nobody can know everything becomes a sign that others are keeping the truth under lock and key.
Research into conspiracy theories describes motives that may align with such content. People sometimes seek clear-cut explanations, a sense of control, and a social position that sets them apart from others. Douglas, Sutton and Cichocka summarise these motives as epistemic, existential and social needs. This research says nothing about the individual viewer of a video on the Templars or ancient Rome, nor does it address the impact of a specific title. However, it helps us understand why the notion of secret knowledge can be so appealing. It promises clarity, control and a small boost to one’s status.
How a gap in the sources becomes a conspiracy
A video claiming that historians are concealing something alters the nature of criticism. The question of sources – otherwise a self-evident part of historical work – can then be seen as a defensive manoeuvre by those who are allegedly covering things up. Anyone who is restrictive comes across as cowardly. Anyone who cites multiple interpretations comes across as indecisive. Anyone who says that the evidence does not allow for a clear conclusion confirms, in the eyes of some viewers, precisely the suspicion that a truth must not be spoken.
In this way, a gap in the historical record easily becomes a conspiracy. Where a document is missing, the video can create a scene. Where a chronicler’s account is contradictory, it can provide the ‘true’ explanation. Where historians argue over different causes, it can proclaim a secret cause. Yet the work of history begins precisely where one resists the temptation to fill every gap.
What historical sources achieve
Historical sources are not complete. Laws do not tell us how people obeyed them in everyday life. Chronicles describe events from a particular perspective. Images can glorify those in power. Court records speak under duress, within specific procedures and with an inherent interest in maintaining order. Even a large volume of material does not exempt one from the task of examining a source’s origin, purpose and limitations. The historian who states that something remains open is identifying the limits of what can be deduced from the material.
For the clickbait video, this limit is unsatisfactory. It needs a conclusion. It needs a cause that can be condensed into a single sentence, a motive that makes a character understandable, a revelation that comes at the end of the narrative. ‘We don’t know’ is not a title. ‘The sources contradict one another’ does not make for a thumbnail. ‘Researchers are debating several explanations’ brings the series to a close.
The synthetic voice knows no doubt
AI makes it easier to fill the gap immediately. It can generate a plausible scene from just a few keywords, a psychological biography from a historical name, and a linear plot from a controversial question. This technical smoothness is part of the persuasive power. A language model formulates transitions, motives and causes with the same composure with which it writes a shopping list or a summary. The synthetic voice conveys this composure in its tone. It shows no hesitation in finding the right words, no interruptions, no visible uncertainty. It sounds as though it must have been that way.
The fantasy of knowledge without residue predates AI. Lacan offers a concise formula for this. The videos promise a past in which every mystery has a solution and every solution was merely hidden until now. One may find this promise psychologically relieving. It frees one from chance, contradiction and ignorance. At the same time, it accustoms one to a poor form of understanding, in which every ambiguity must have a secret cause.
Commodity aesthetics according to Haug: entertainment in the guise of education
Wolfgang Fritz Haug used the concept of ‘commodity aesthetics’ to describe how commodities offer their utility as an appearance. Form, packaging and advertising demonstrate, before the purchase, what the commodity is supposedly capable of delivering. They are intended to generate desire and facilitate access. In the case of the AI history video, the promised benefit is education. The video aims to create the impression that, after clicking, the viewer will know more about Rome, the Middle Ages or a past culture.
The form of the explanation takes on this task early on. The voice sounds matter-of-fact. The map organises the space. The year organises the time. The image gives the event a face. The artificially aged portrait seems to gaze at the historical figure from afar. A page of parchment, a seal or a ruin creates a sense of proximity to the archive. The viewer is presented with the trappings of historical representation without learning who produced them or what they are based on.
Haug helps us to recognise the moral character of the packaging here. The video sells entertainment in the guise of education, and the product combines appeal with self-respect.
This combination is what makes the videos so useful for a platform that needs to measure and channel attention. An academic lecture on Roman family law, the uncertainty surrounding ancient population figures or the problems of archaeological dating is difficult to transform into a sequence of images that generates fresh excitement every few seconds. A video about ‘the secret sexual practices of the Romans’ solves this problem. It can sprinkle in historical names and terms without having to sacrifice the subject’s intrigue. The pretence of education protects the allure. The allure funds the pretence of education.
The technical production reinforces this connection. A single historian cannot publish a carefully researched documentary every day. An AI channel can produce new variations of the same basic format every day. The dramatic arc remains the same. A mystery is introduced, the official explanation is dismissed as inadequate, a scene provides the emotional climax, and the revelation comes at the end. After that, the next episode begins with a new name and the same promise.
Baudrillard and the visual aspect of historicity
Baudrillard offers a useful distinction regarding the visual aspect of this production. The video rarely cites sources. It cites the images through which viewers have learnt to recognise sources. An artificially aged face is reminiscent of a portrait from the archives. A mould-stained parchment is reminiscent of a document. A map with red arrows is reminiscent of a historical explanation. A slow-motion image of a computer-generated city evokes a reconstruction. The videos reproduce the visual appearance of historicity.
The generated images depict the past as the present imagines it. The Roman banquet is reminiscent of a lavish streaming series or a perfume advert featuring columns and gold. The medieval dungeon is reminiscent of fantasy and computer games. The Viking is reminiscent of a fitness campaign featuring fur, a beard and snow. The ‘Orient’ appears as a mixture of harem, palace, veils and secret passageways. The platform replicates these stereotypes in great numbers.
It is precisely here that the danger of the seemingly neutral illustration lies. The image makes the claim sensually plausible. An image of a Roman banquet can provide the voice with a scene in which the alleged sexual debauchery already seems visible. An image of chains, torches and dark walls can foreshadow the claim about medieval punishments. No one needs to explicitly state that the image is a source. Its position in the video is sufficient. It is there whilst the voice speaks, giving the statement a visible world.
The visual language spares the viewer the question of where the claim comes from. It gives them the feeling of having already seen something. The form of presentation renders source criticism superfluous, because it creates the impression that the matter has already become vividly clear. The AI produces an impression of the past that can be used for the next sentence, the next clip and the next revelation.
Illusory truth: the effect of repetition
Research into illusory truth begins much earlier. In experiments, repeated statements are, on average, more likely to be believed to be true than new ones. Dechêne, Stahl, Hansen and Wänke summarised 51 studies and reported moderate effects. However, the studies deal with short statements, trivia questions and headlines. They do not show that AI videos rewrite collective memory, nor do they show that viewers accept every repeated claim.
The experiments do, however, explain why familiarity has consequences. Repetition can skew judgement, even when people fundamentally possess correct knowledge. Fazio and colleagues found that existing knowledge does not reliably prevent repeated false statements from appearing more credible. Henderson, Simons and Barr demonstrated that the effect diminishes without further exposure. For the platform, therefore, a continuous supply of content is key.
AI-Slop primarily repeats the same ‘teaching format’ about history, even when the claims change. One episode explains that Rome was sexually uninhibited. The next explains that the Middle Ages were characterised by particularly cruel punishments. The next claims that an elite concealed knowledge about the Templars. The subjects change, but the structure remains the same. The past is presented as material that has been institutionally sanitised, morally censored or deliberately concealed. The channel appears to be a place where this sanitisation is reversed.
Whether such repetition generates false memories or enduring convictions has not been investigated. The question of what attitude towards history such videos instil nevertheless remains. They instil suspicion. They instil a view of violence and sexuality as a source of ‘ ’ insight. They instil the expectation that every gap is a cover-up, every nuance an evasion, and every historical explanation a weaker version of the grand revelation.
What AI Slop sells
The real deluge, therefore, consists of a continuous organisation of attention. A title promises secret knowledge. An image conveys an atmosphere of truth. A voice lends the whole thing the tone of an explanation. A scandal keeps the scene in motion. The next episode repeats the formula with different names. AI makes this series cheap. YouTube makes it visible. History provides the images and figures that can be slotted into the cycle.
Anyone who criticises these videos must therefore be more precise than the videos themselves. Not every AI illustration is pseudo-history. Not every popular history piece is clickbait. Not every interest in violence, sexuality or secrets is morally suspect. The issue concerns a specific mode of production. It turns the past into a backdrop for stimuli that masquerade as education. It bestows upon curiosity about the forbidden the dignity of knowledge. It transforms historical uncertainty into the assertion of a suppressed truth.
The harm lies in a particular way of seeing and hearing. The past appears there as something that is only worthwhile if it shocks, outrages or initiates. It appears as a world in which historians say too little, sources reveal too little, and a synthetic voice finally articulates what no one else dared to say. The viewer becomes accustomed to the idea that history feels good when it reveals a secret scene to them whilst simultaneously assuring them that they are wiser than the rest.
That is the commodity that AI-Slop sells. It consists of images of violence, sex and otherness, of the gesture of secret knowledge, and of the reassuring voice that fills in all the gaps. The historical claim merely provides the pretext.
The key points in brief
• AI history videos on YouTube follow a set formula: real names and dates, a script generated by the language model, a synthetic voice, and generated images.
• What is new is the sheer volume and the imagery: AI turns history into historical pornography, supplemented by conspiracy theories. YouTube’s rule against ‘inauthentic content’ does not apply here, because the topics change whilst only the method remains the same.
• The past is sorted by its sensational value. The historical context makes the consumption of sex, violence and decadence appear to be educational. The arena and the place of execution justified the same sadistic voyeurism in the name of order and justice.
• The phrase ‘What historians keep quiet’ promises secret knowledge and a small boost to one’s status. Every gap in the historical record becomes a conspiracy.
• The images generated evoke the look of archives, parchment and maps, making claims seem plausibly credible without substantiating them.
• Repetition makes statements in the experiment more credible. Whether AI-generated content creates lasting convictions has not been investigated.
• The videos instil a certain attitude: suspicion towards sources and historians, and violence and sexuality as sources of knowledge.
Sources
· YouTube Help: YouTube channel monetisation policies, section ‘Inauthentic content’. https://support.google.com/youtube/answer/1311392
· Social Media Today (2025): YouTube Clarifies Monetisation Update Around Inauthentic and Repeated Content. https://www.socialmediatoday.com/news/youtube-clarifies-monetization-update-inauthentic-repeated-content/752892/
· Lacan, J. (1973): The Seminar, Book XI. The Four Fundamental Concepts of Psychoanalysis. Paris: Seuil. German edition: The Four Fundamental Concepts of Psychoanalysis. Olten 1978.
· Douglas, K. M., Sutton, R. M., Cichocka, A. (2017): The psychology of conspiracy theories. Current Directions in Psychological Science 26 (6), 538–542. https://doi.org/10.1177/0963721417718261
· Haug, W. F. (1971): A Critique of Commodity Aesthetics. Frankfurt am Main: Suhrkamp.
· Baudrillard, J. (1981): Simulacra and Simulation. Paris: Galilée.
· Dechêne, A., Stahl, C., Hansen, J., Wänke, M. (2010): The truth about the truth: A meta-analytic review of the truth effect. Personality and Social Psychology Review 14 (2), 238–257. https://doi.org/10.1177/1088868309352251
· Fazio, L. K., Brashier, N. M., Payne, B. K., Marsh, E. J. (2015): Knowledge does not protect against illusory truth. Journal of Experimental Psychology: General 144 (5), 993–1002. https://doi.org/10.1037/xge0000098
· Henderson, E. L., Simons, D. J., Barr, D. J. (2021): The trajectory of truth: A longitudinal study of the illusory truth effect. Journal of Cognition 4 (1), 29. https://doi.org/10.5334/joc.161
· Freud, S. (1915): Instincts and Their Destinies. International Journal of Medical Psychoanalysis 3, 84–100. Collected Works, Vol. X.
· Elias, N. (1939): On the Civilising Process. Basel: Haus zum Falken.
· Foucault, M. (1975): Surveiller et punir. Naissance de la prison. Paris: Gallimard. German: Überwachen und Strafen. Frankfurt am Main 1976.
Related
AI Slop: Artificial intelligence, digital videos and AI rubbish on the internet and on TikTok & Co.
The Aesthetics of Commodities According to Wolfgang Fritz Haug – What MrBeast Has to Do with It
The secret symbols of the Middle Ages: what medieval illuminators wanted to tell us
Rage-baiting on social media: recognising and understanding it
DESCRIPTION: AI history videos on YouTube market sex, violence and secret knowledge as education, and turn every gap in the sources into a conspiracy. An analysis drawing on Lacan, Haug, Baudrillard and research into illusory truth.
History as a commodity
AI drivel, pseudo-history and the allure of secret knowledge on YouTube
On YouTube, a form of pseudo-history is emerging, fuelled by AI scripts, synthetic voices and generated images. It sorts the past according to its sensational value, peddles sex, violence and secret knowledge under the guise of education, and turns every gap in the historical record into a conspiracy. Here, this format is analysed through the lens of Lacan, Haug and Baudrillard, alongside research into conspiracy thinking and illusory truth.
‘The Romans’ secret sexual practices’. ‘What historians are keeping quiet about the Vikings’. ‘The most cruel punishments of the Middle Ages’. ‘The truth about the Templars’. Anyone who clicks on such titles does not initially know who they come from, what the claims are based on, or whether they contain any verifiable assertions at all. The thumbnail might show a woman in an artificially antique setting, a barbaric-looking warrior, a blood-stained dungeon or a parchment page bearing a seal. Beneath the image is a title that combines knowledge and taboo. The video promises both enlightenment and titillation.
The voice begins with a claim. It sounds calm, knowledgeable and free from doubt. An emperor is said to have done this, a people is said to have performed that ritual, a court is said to have been characterised by sexual debauchery, a church is said to have removed certain facts from its archives. Whilst the voice speaks, a sequence of generated images flashes by. They show people who never lived, in rooms that never existed, engaged in actions that no source describes. The images are not consistently absurd. They are too polished for that. They combine familiar elements of the past with a visual style familiar from TV series, computer games, adverts and television documentaries. It is precisely for this reason that the impression arises that the voice has made something visible.
AI Slop on YouTube: Pseudo-history as a series
This pseudo-history is recognisable by its consistent approach. Real names, dates and historical events provide the material. A language model fills in the context, a synthetic voice transforms it into a narrative, and images provide scenes that seem to race ahead of the sentences. The viewer sees a Roman meal whilst hearing about Roman sexuality. They see a torture chamber whilst medieval criminal law is being discussed. They see an artificially aged face whilst a video explains a ruler’s secret motives. The scene brings the claim to life, even though it provides no evidence.
Inaccurate and sensationalist historical narratives existed long before the videos. Early mythology explained the origins of peoples and cities through gods, giants and monsters. Herodotus reports on ants in India, larger than foxes, which dig gold-bearing sand from the earth. Plutarch himself stated that he wrote biographies, and embellished them with omens, dreams and anecdotes. Flavius Josephus has 960 besieged people at Masada die together and puts two long speeches into the mouth of their leader, which no one could have written down. The biblical authors recounted the history of Israel as a story of salvation, complete with miracles, plagues and a sun that stood still over Gibeon. The *Physiologus*, thought to have been written in Alexandria between the 2nd and 4th centuries, transformed zoology into a doctrine of faith: the pelican revives its dead young with its own blood; the lion sleeps with its eyes open. Jacobus de Voragine’s *Legenda Aurea*, compiled around 1260, collected the lives of saints full of miracles and became one of the most widely read books of the late Middle Ages. In the 16th century, the ‘Neue Zeitungen’ (New Newspapers) reported on leaflets featuring woodcuts, birth defects, blood rain and celestial phenomena as signs of the approaching Day of Judgement. Pseudoscience and conspiracy literature are just as old.
The only things that are new now are the sheer volume and the imagery. Suetonius, who wrote about Tiberius on Capri, and Procopius, who slandered Empress Theodora in his *Secret History*, were already familiar with sexualised narratives. AI turns these into historical pornography. The term is modelled on the English ‘disaster porn’ and ‘trauma porn’, which refer to the sensationalist display of suffering, and here also carries the literal, sexual connotation. For every sentence about Roman sexuality, the AI provides the appropriate body; for every torture chamber, the victim; and the channel thumbnails are designed specifically to feature these images. What the images fail to deliver, the conspiracy narrative makes up for. Generative AI also allows scripts, voices and images to be produced in series. A channel can apply the same template to Rome, Egypt, the Viking Age, medieval courts, the Second World War or ‘lost civilisations’ without having to undertake a lengthy study of any particular era. The narrative simply needs a theme that offers enough well-known names and enough scope for mysteries.
In July 2025, YouTube renamed its rule against ‘repetitious content’ to ‘inauthentic content’. It targets mass-produced videos created solely to generate views. History channels remain unaffected by this. Every video has a new theme, a new name, a new century. The only thing that is repeated is the process, and the rule does not cover that.
The past, sorted by sensational value
The historical significance of a topic counts for little in this process. What matters is whether it can be used to create a title that triggers curiosity, revulsion, outrage, fear or the desire for exclusive knowledge. The past is sorted by its sensational value. It appears as a storehouse of scandals, secrets, violence and sexuality. An era becomes a ‘dark’ century, a dynasty an incestuous drama, a religion a secret cult, an administrative issue a conspiracy, and a gap in the historical record suppressed knowledge.
This selection has consequences. Anyone who continually consumes history in this form encounters a past in which everything revolves around crime, decadence, hidden practices and secret powers. Historical societies appear as a playground for extreme scenes. The social order, legal distinctions, economic dependencies, concepts and self-interpretations of the time disappear behind the spectacle. What remains is the cruel emperor, the erotic slave, the fanatical priest, the barbaric warrior, the corrupt court and the victim in the dungeon.
Such images organise the past according to the expectations of the present-day gaze. Antiquity is supposed to be luxurious, sexual and cruel. The Middle Ages are supposed to be filthy, superstitious and violent. The ‘Orient’ is supposed to be mysterious, erotic and despotic. The Viking Age is supposed to deliver masculinity, snow, blood and raw freedom. These visual patterns stem from a long history of novels, paintings, adventure films, colonial fantasies, historical series, computer games and advertising. AI uses these to generate new variations of a past that has already been shaped by the media.
The clips and videos therefore often look like short trailers for a series that never existed. Every change of scene offers a new promise: something momentous, forbidden or revealing is about to be revealed. The historical subject matter functions as a brand and requires no accuracy. ‘Templars’ stand for secrets, conspiracy and hidden treasures. The video can rely on these cultural shorthands because they are already present in the audience’s minds.
Education as a licence for sex and violence
Clicking on ‘The Secret Sexual Practices of the Romans’ is regarded as education. This is a crucial part of the format. The title sells a sexualised scene, but the word ‘history’ sanitises the consumption. Anyone viewing erotic images is learning about antiquity. Anyone clicking on torture is learning about medieval legal practice. The historical framework lends the allure a culturally acceptable form.
People open such videos out of curiosity, because of a specific historical interest, out of boredom, or because the platform presents them with these videos. The point here concerns the logic of the format. It selects its topics in such a way that education and titillation overlap. The viewer can be shocked, outraged or fascinated, whilst at the same time telling themselves they have learnt something.
Historical distance reinforces this licence. One can view past violence without encountering a history of present-day violence. The sexual exploitation of enslaved women can become a visual motif because the woman depicted appears as part of a backdrop called Rome. The torture of a prisoner in the Middle Ages becomes a detail of a grim world that can be consumed. The suffering remains real. The distance of the centuries makes it easier to transform that suffering into visual material.
The pattern is an old one. The Roman arena celebrated the emperor’s generosity and the order of the empire; the execution grounds of the early modern period celebrated justice and deterrence. Both presented suffering as a spectacle and provided the very reason why one was permitted to look on. The AI video digitally revives this arrangement. Its justification is called education, and it caters to the same sadistic voyeurism. In ‘Drives and Drive Fates’, Freud described sadism and voyeurism as parallel pairs of drives, with their passive counterparts, masochism and exhibitionism. The historical framework provides both with a stage on which they appear as a thirst for knowledge.
Three differences reveal what the format has changed. The arena was a collective ritual in which the crowd watched itself and reaffirmed its order. The video is watched alone, in bed or on the train, and no one sees the viewer. Those being tortured have been dead for centuries and the images are synthetic, so no one suffers for the sake of the scene any longer. This makes the pleasure cheap and arbitrarily repeatable. Finally, the video reverses a movement that Norbert Elias described as a process of civilisation: cruelty was relegated to behind the scenes over the course of centuries. Elias recalls the burning of cats at the Feast of St John in 16th-century Paris. Foucault begins *Discipline and Punish* with the torture of Damien in 1757 and shows how the spectacle of punishment subsequently disappeared. The Slop brings torture back onto the screen. History itself serves as its backdrop.
Here, the critique touches on a problem that goes beyond the question of individual errors. AI-Slop transforms historical figures into visual archetypes for contemporary needs. The cruel ruler reinforces the belief that earlier times were morally inferior. The exoticised woman permits a sexualised gaze upon a past rendered alien. The barbaric ritual turns a culture into a theatre of revulsion and superiority. The fanatic confirms the present’s self-portrayal as rational and enlightened. History becomes material through which viewers can simultaneously derive excitement and achieve moral purification.
The gaze in Lacan: witness to a forbidden world
Lacan’s theory of the gaze offers a concept for this. For Lacan, the gaze originates from the image: the picture plane determines what the viewer is permitted to see and how they relate to the visible. The AI video presents violence, sexuality and otherness in such a way that the viewer can appear as a witness to a forbidden world. They see what they are supposedly not meant to see, and they see it under the protection of the claim that they are merely imagining it. The image binds them to the visible whilst simultaneously allowing for distance: that was a long time ago, those are other people, that is history.
The concept of the ‘gaze’ would only be meaningful here if it described the specific pictorial order. AI images place the viewer in a specific position vis-à-vis the past. They stand before the scene as someone who is both in the know and morally superior. They now know how cruel people were back then, how sexually uninhibited an imperial court is said to have been, or how secretive the rituals of a foreign culture were. The past becomes the object of a mixture of revulsion, lust and self-assurance.
‘What Historians Conceal’: Secret Knowledge as Status
This mixture also explains the power of the phrase ‘What historians keep quiet about’. Beyond the information itself, it offers a position within a social hierarchy. The viewer gains access to knowledge that others supposedly do not possess. Historians, schools, museums, archives and the media appear as institutions that omit something or deliberately conceal it.
This gesture begins even before the first sentence of the video. The title assigns the roles. Here stands the courageous whistleblower. There stand the institutions that remain silent. The viewer is addressed as someone who is no longer willing to be fobbed off with the official history. Their ignorance is attributed to a culprit. What would normally be the consequence of the fact that nobody can know everything becomes a sign that others are keeping the truth under lock and key.
Research into conspiracy theories describes motives that may align with such content. People sometimes seek clear-cut explanations, a sense of control, and a social position that sets them apart from others. Douglas, Sutton and Cichocka summarise these motives as epistemic, existential and social needs. This research says nothing about the individual viewer of a video on the Templars or ancient Rome, nor does it address the impact of a specific title. However, it helps us understand why the notion of secret knowledge can be so appealing. It promises clarity, control and a small boost to one’s status.
How a gap in the sources becomes a conspiracy
A video claiming that historians are concealing something alters the nature of criticism. The question of sources – otherwise a self-evident part of historical work – can then be seen as a defensive manoeuvre by those who are allegedly covering things up. Anyone who is restrictive comes across as cowardly. Anyone who cites multiple interpretations comes across as indecisive. Anyone who says that the evidence does not allow for a clear conclusion confirms, in the eyes of some viewers, precisely the suspicion that a truth must not be spoken.
In this way, a gap in the historical record easily becomes a conspiracy. Where a document is missing, the video can create a scene. Where a chronicler’s account is contradictory, it can provide the ‘true’ explanation. Where historians argue over different causes, it can proclaim a secret cause. Yet the work of history begins precisely where one resists the temptation to fill every gap.
What historical sources achieve
Historical sources are not complete. Laws do not tell us how people obeyed them in everyday life. Chronicles describe events from a particular perspective. Images can glorify those in power. Court records speak under duress, within specific procedures and with an inherent interest in maintaining order. Even a large volume of material does not exempt one from the task of examining a source’s origin, purpose and limitations. The historian who states that something remains open is identifying the limits of what can be deduced from the material.
For the clickbait video, this limit is unsatisfactory. It needs a conclusion. It needs a cause that can be condensed into a single sentence, a motive that makes a character understandable, a revelation that comes at the end of the narrative. ‘We don’t know’ is not a title. ‘The sources contradict one another’ does not make for a thumbnail. ‘Researchers are debating several explanations’ brings the series to a close.
The synthetic voice knows no doubt
AI makes it easier to fill the gap immediately. It can generate a plausible scene from just a few keywords, a psychological biography from a historical name, and a linear plot from a controversial question. This technical smoothness is part of the persuasive power. A language model formulates transitions, motives and causes with the same composure with which it writes a shopping list or a summary. The synthetic voice conveys this composure in its tone. It shows no hesitation in finding the right words, no interruptions, no visible uncertainty. It sounds as though it must have been that way.
The fantasy of knowledge without residue predates AI. Lacan offers a concise formula for this. The videos promise a past in which every mystery has a solution and every solution was merely hidden until now. One may find this promise psychologically relieving. It frees one from chance, contradiction and ignorance. At the same time, it accustoms one to a poor form of understanding, in which every ambiguity must have a secret cause.
Commodity aesthetics according to Haug: entertainment in the guise of education
Wolfgang Fritz Haug used the concept of ‘commodity aesthetics’ to describe how commodities offer their utility as an appearance. Form, packaging and advertising demonstrate, before the purchase, what the commodity is supposedly capable of delivering. They are intended to generate desire and facilitate access. In the case of the AI history video, the promised benefit is education. The video aims to create the impression that, after clicking, the viewer will know more about Rome, the Middle Ages or a past culture.
The form of the explanation takes on this task early on. The voice sounds matter-of-fact. The map organises the space. The year organises the time. The image gives the event a face. The artificially aged portrait seems to gaze at the historical figure from afar. A page of parchment, a seal or a ruin creates a sense of proximity to the archive. The viewer is presented with the trappings of historical representation without learning who produced them or what they are based on.
Haug helps us to recognise the moral character of the packaging here. The video sells entertainment in the guise of education, and the product combines appeal with self-respect.
This combination is what makes the videos so useful for a platform that needs to measure and channel attention. An academic lecture on Roman family law, the uncertainty surrounding ancient population figures or the problems of archaeological dating is difficult to transform into a sequence of images that generates fresh excitement every few seconds. A video about ‘the secret sexual practices of the Romans’ solves this problem. It can sprinkle in historical names and terms without having to sacrifice the subject’s intrigue. The pretence of education protects the allure. The allure funds the pretence of education.
The technical production reinforces this connection. A single historian cannot publish a carefully researched documentary every day. An AI channel can produce new variations of the same basic format every day. The dramatic arc remains the same. A mystery is introduced, the official explanation is dismissed as inadequate, a scene provides the emotional climax, and the revelation comes at the end. After that, the next episode begins with a new name and the same promise.
Baudrillard and the visual aspect of historicity
Baudrillard offers a useful distinction regarding the visual aspect of this production. The video rarely cites sources. It cites the images through which viewers have learnt to recognise sources. An artificially aged face is reminiscent of a portrait from the archives. A mould-stained parchment is reminiscent of a document. A map with red arrows is reminiscent of a historical explanation. A slow-motion image of a computer-generated city evokes a reconstruction. The videos reproduce the visual appearance of historicity.
The generated images depict the past as the present imagines it. The Roman banquet is reminiscent of a lavish streaming series or a perfume advert featuring columns and gold. The medieval dungeon is reminiscent of fantasy and computer games. The Viking is reminiscent of a fitness campaign featuring fur, a beard and snow. The ‘Orient’ appears as a mixture of harem, palace, veils and secret passageways. The platform replicates these stereotypes in great numbers.
It is precisely here that the danger of the seemingly neutral illustration lies. The image makes the claim sensually plausible. An image of a Roman banquet can provide the voice with a scene in which the alleged sexual debauchery already seems visible. An image of chains, torches and dark walls can foreshadow the claim about medieval punishments. No one needs to explicitly state that the image is a source. Its position in the video is sufficient. It is there whilst the voice speaks, giving the statement a visible world.
The visual language spares the viewer the question of where the claim comes from. It gives them the feeling of having already seen something. The form of presentation renders source criticism superfluous, because it creates the impression that the matter has already become vividly clear. The AI produces an impression of the past that can be used for the next sentence, the next clip and the next revelation.
Illusory truth: the effect of repetition
Research into illusory truth begins much earlier. In experiments, repeated statements are, on average, more likely to be believed to be true than new ones. Dechêne, Stahl, Hansen and Wänke summarised 51 studies and reported moderate effects. However, the studies deal with short statements, trivia questions and headlines. They do not show that AI videos rewrite collective memory, nor do they show that viewers accept every repeated claim.
The experiments do, however, explain why familiarity has consequences. Repetition can skew judgement, even when people fundamentally possess correct knowledge. Fazio and colleagues found that existing knowledge does not reliably prevent repeated false statements from appearing more credible. Henderson, Simons and Barr demonstrated that the effect diminishes without further exposure. For the platform, therefore, a continuous supply of content is key.
AI-Slop primarily repeats the same ‘teaching format’ about history, even when the claims change. One episode explains that Rome was sexually uninhibited. The next explains that the Middle Ages were characterised by particularly cruel punishments. The next claims that an elite concealed knowledge about the Templars. The subjects change, but the structure remains the same. The past is presented as material that has been institutionally sanitised, morally censored or deliberately concealed. The channel appears to be a place where this sanitisation is reversed.
Whether such repetition generates false memories or enduring convictions has not been investigated. The question of what attitude towards history such videos instil nevertheless remains. They instil suspicion. They instil a view of violence and sexuality as a source of ‘ ’ insight. They instil the expectation that every gap is a cover-up, every nuance an evasion, and every historical explanation a weaker version of the grand revelation.
What AI Slop sells
The real deluge, therefore, consists of a continuous organisation of attention. A title promises secret knowledge. An image conveys an atmosphere of truth. A voice lends the whole thing the tone of an explanation. A scandal keeps the scene in motion. The next episode repeats the formula with different names. AI makes this series cheap. YouTube makes it visible. History provides the images and figures that can be slotted into the cycle.
Anyone who criticises these videos must therefore be more precise than the videos themselves. Not every AI illustration is pseudo-history. Not every popular history piece is clickbait. Not every interest in violence, sexuality or secrets is morally suspect. The issue concerns a specific mode of production. It turns the past into a backdrop for stimuli that masquerade as education. It bestows upon curiosity about the forbidden the dignity of knowledge. It transforms historical uncertainty into the assertion of a suppressed truth.
The harm lies in a particular way of seeing and hearing. The past appears there as something that is only worthwhile if it shocks, outrages or initiates. It appears as a world in which historians say too little, sources reveal too little, and a synthetic voice finally articulates what no one else dared to say. The viewer becomes accustomed to the idea that history feels good when it reveals a secret scene to them whilst simultaneously assuring them that they are wiser than the rest.
That is the commodity that AI-Slop sells. It consists of images of violence, sex and otherness, of the gesture of secret knowledge, and of the reassuring voice that fills in all the gaps. The historical claim merely provides the pretext.
The key points in brief
• AI history videos on YouTube follow a set formula: real names and dates, a script generated by the language model, a synthetic voice, and generated images.
• What is new is the sheer volume and the imagery: AI turns history into historical pornography, supplemented by conspiracy theories. YouTube’s rule against ‘inauthentic content’ does not apply here, because the topics change whilst only the method remains the same.
• The past is sorted by its sensational value. The historical context makes the consumption of sex, violence and decadence appear to be educational. The arena and the place of execution justified the same sadistic voyeurism in the name of order and justice.
• The phrase ‘What historians keep quiet’ promises secret knowledge and a small boost to one’s status. Every gap in the historical record becomes a conspiracy.
• The images generated evoke the look of archives, parchment and maps, making claims seem plausibly credible without substantiating them.
• Repetition makes statements in the experiment more credible. Whether AI-generated content creates lasting convictions has not been investigated.
• The videos instil a certain attitude: suspicion towards sources and historians, and violence and sexuality as sources of knowledge.
Sources
· YouTube Help: YouTube channel monetisation policies, section ‘Inauthentic content’. https://support.google.com/youtube/answer/1311392
· Social Media Today (2025): YouTube Clarifies Monetisation Update Around Inauthentic and Repeated Content. https://www.socialmediatoday.com/news/youtube-clarifies-monetization-update-inauthentic-repeated-content/752892/
· Lacan, J. (1973): The Seminar, Book XI. The Four Fundamental Concepts of Psychoanalysis. Paris: Seuil. German edition: The Four Fundamental Concepts of Psychoanalysis. Olten 1978.
· Douglas, K. M., Sutton, R. M., Cichocka, A. (2017): The psychology of conspiracy theories. Current Directions in Psychological Science 26 (6), 538–542. https://doi.org/10.1177/0963721417718261
· Haug, W. F. (1971): A Critique of Commodity Aesthetics. Frankfurt am Main: Suhrkamp.
· Baudrillard, J. (1981): Simulacra and Simulation. Paris: Galilée.
· Dechêne, A., Stahl, C., Hansen, J., Wänke, M. (2010): The truth about the truth: A meta-analytic review of the truth effect. Personality and Social Psychology Review 14 (2), 238–257. https://doi.org/10.1177/1088868309352251
· Fazio, L. K., Brashier, N. M., Payne, B. K., Marsh, E. J. (2015): Knowledge does not protect against illusory truth. Journal of Experimental Psychology: General 144 (5), 993–1002. https://doi.org/10.1037/xge0000098
· Henderson, E. L., Simons, D. J., Barr, D. J. (2021): The trajectory of truth: A longitudinal study of the illusory truth effect. Journal of Cognition 4 (1), 29. https://doi.org/10.5334/joc.161
· Freud, S. (1915): Instincts and Their Destinies. International Journal of Medical Psychoanalysis 3, 84–100. Collected Works, Vol. X.
· Elias, N. (1939): On the Civilising Process. Basel: Haus zum Falken.
· Foucault, M. (1975): Surveiller et punir. Naissance de la prison. Paris: Gallimard. German: Überwachen und Strafen. Frankfurt am Main 1976.
Related
AI Slop: Artificial intelligence, digital videos and AI rubbish on the internet and on TikTok & Co.
The Aesthetics of Commodities According to Wolfgang Fritz Haug – What MrBeast Has to Do with It
The secret symbols of the Middle Ages: what medieval illuminators wanted to tell us
Rage-baiting on social media: recognising and understanding it
Newsletter
Möchten Sie informiert werden, wenn ich einen neuen Blogbeitrag veröffentliche? Tragen Sie sich gern hier ein.
Klicken Sie auf „Kostenlos abonnieren“, wenn Sie meine Datenschutzerklärung gelesen und akzeptiert haben. Ihre Einwilligung können Sie jederzeit widerrufen.
Newsletter
Möchten Sie informiert werden, wenn ich einen neuen Blogbeitrag veröffentliche? Tragen Sie sich gern hier ein.
Klicken Sie auf „Kostenlos abonnieren“, wenn Sie meine Datenschutzerklärung gelesen und akzeptiert haben. Ihre Einwilligung können Sie jederzeit widerrufen.