The West Is GASLIGHTING Everyone — Now With AI | Profs. D. Riches & J. Eddebo

The conversation at the Kyoto symposium insists on a simple, uncomfortable truth: who controls the filters that turn events into history gets to shape what the world understands as reality. That claim moves beyond conspiracy and into the mechanics of knowledge production — how interpretations are chosen, repeated, institutionalized and then reinforced by technologies that accelerate repetition and silence nuance. The exchange uses the contested histories of Rwanda and the broader Great Lakes conflicts as a case study, then pivots to the contemporary dangers of artificial intelligence and a post‑digital public sphere that rewards certainty over inquiry. What emerges is less a single indictment than a warning: powerful actors, old and new, can convert ambiguity into orthodoxy; resisting that process requires both critical literacy and new forms of public infrastructure for dissent. ## Narrative and the Machinery of Historical Meaning The discussion opens with a reminder that facts alone do not determine meaning. Drawing on cinematic metaphor, the exchange highlights how context and intonation shape interpretation — the difference between "he'd kill us" and "he'd kill us" becomes emblematic of a larger epistemic problem. In conflict zones, surface facts are often plentiful but context is scarce; powerful actors fill that vacuum with narratives that suit their political ends. Institutional authority — states, major media outlets, NGOs, academic gatekeepers — plays a decisive role in translating fragmentary events into coherent stories. Once a narrative achieves institutional sanction, it acquires a momentum that crowd‑sources truth: textbooks, grants, awards, and media coverage then reproduce the sanctioned framing and marginalize alternatives. The conversation stresses that this is not merely academic hair‑splitting. Narrative selection governs moral and political consequences: whom we call perpetrators and victims, which interventions are justified, which actors are held accountable, and what reparative projects are imaginable. The mechanics are mundane — rapport with journalists, networked influence in NGOs and universities, and strategic citation practices — yet their effects are global. The institutionalization of a single explanatory frame transforms ambiguity into a social fact that organizes memory, policy and law. ## The Rwanda Case: Two Competing Stories Using the Rwandan crisis of the 1990s, the exchange lays out two sharply diverging accounts. One is the predominant official narrative that centers an internal mass psychosis and attributes most culpability to those who carried out violence on the ground. That version simplifies causes into local hatreds and emphasizes individualized accountability. The alternative narrative foregrounds geopolitical strategy and external intervention: long-term displacement, cross‑border warfare, foreign sponsorship of exiled forces, and strategic aims to reconfigure regional alignments. From this perspective, 1994 cannot be reduced to an eruptive frenzy but must be seen in the longer arc of external destabilization, resource interests and the politics of regime change. The conversation interrogates how the official version consolidated its authority. Diasporic advocacy, diplomatic priorities of powerful states, and a Western information ecosystem quick to identify clearly defined victims all converged to produce a tight, repeatable script. Institutions ranging from major newspapers to encyclopedic repositories favored certain sources and marginalized others; citation networks and educational pipelines then propagated that perspective to new generations. Importantly, the discussants emphasize that contesting the dominant account has often been punished as denialism rather than engaged as a legitimate historiographical dispute. The result is a public record that bears the weight of institutional endorsement more than the full complexity of competing evidence. ## From Propaganda Autocomplete to Algorithmic Consolidation A striking thread in the discussion reframes propaganda not as a relic of Cold War television but as an engine built on repetition and filter systems that long predate contemporary AI. Before large language models, there were already highly effective "autocomplete" systems in public knowledge: a small set of widely cited sources produces the links and summaries that search engines, syllabi and news aggregators surface to learners and journalists. Those systems privilege repetition and visibility; repetition acquires plausibility, plausibility becomes authority, and authority hardens into orthodoxy. Into that pre‑existing apparatus comes machine learning, which magnifies two tendencies. First, models trained on vast corpora can reproduce and amplify the biases embedded in those corpora: the more a narrative is repeated in web text, media and scholarship, the more likely an algorithm is to surface it as canonical. Second, models incentivize fluency and coherence over uncertainty; a confident response, even if selective, passes user tests better than a hedged multi‑part answer. The discussion argues that this turbo‑charged repetition threatens to turn a fragile, contested past into a seamless official past — not because AI invents lies but because it optimizes for the signals that authority has already established. This dynamic also inverts an older romanticism about the internet as a democratizing counterbalance. Where early countercultures once imagined online spaces as refuges for dissent and heterodoxy, the conversation warns that the same network effects that can uplift marginal voices can also mainstream well‑resourced narratives. Amplification is not neutral. Ad networks, search rankings and recommender systems reward engagement and familiarity; they thereby privilege the narratives that already enjoy institutional infrastructure and monetizable traction. ## Post‑Digital Condition and the Simulation of Reasoning The exchange then shifts from geopolitics to epistemology: what does it mean to reason in a world where our interlocutors increasingly include models that simulate independent thought? The conversation labels contemporary life "post‑digital" to emphasize that digital mediation is not an add‑on but constitutive of cognition — it shapes what counts as a relevant question, what sources are legible and how argument is performed. In that milieu, confidence and rhetorical polish can masquerade as deliberative competence. The example of performative charisma — from viral child preachers to bombastic political rhetorics — illustrates how repetition and affect can substitute for coherent reasoning. Artificial intelligence exacerbates this because large language models are designed to produce plausible continuations of text. They are, in effect, highly polished parrots: they reproduce patterns and stylistic cues that mimic deliberation without the grounding mechanisms that make human judgment accountable. The conversation raises a crucial philosophical worry: simulated reasoning can erode the value of actual dissent. If disagreement is easily absorbed into a landscape of plausible alternative framings — all of them polished, searchable and shareable — the space for persistent, uncomfortable critique narrows. Dissent thrives not only on being heard but on having institutional venues that sustain and protect minority perspectives; AI‑enhanced monocultures can swamp those venues. The worry is not purely technological determinism. Rather, it is institutional: when models are trained on corpuses curated by the very institutions that favored a single narrative, the model becomes a multiplier for an already established consensus. A self‑reinforcing feedback loop forms: institutional consensus shapes data; data trains models; models reproduce consensus faster and more convincingly than human networks can question it. The net effect is epistemic narrowing. ## Where Dissent Can Survive: Commu

The West Is GASLIGHTING Everyone — Now With AI | Profs. D. Riches & J. Eddebo Dennis Riches and Johan Eddebo examine how dominant narratives are produced, defended and transformed by technology. Kyoto University Symposium 2026: "Geopolitics in the Time of Monsters" Presentation 5: Professor Dennis Riches (Literature) & Professor Johan Eddebo (Philosophy) Riches presents a contested interpretation of Rwanda and the Great Lakes conflicts, using it to question institutional authority and historical representation. Eddebo explores artificial intelligence, simulated reasoning and the erosion of meaningful dissent. Their Kyoto Symposium discussion considers internet countercultures, propaganda, media influence and human rationality, asking whether new forms of community and engagement with reality can resist the pressures of an emerging monoculture. Dennis and Johan's Links: Dennis Riches Substack: https://dennisriches.substack.com/ Lit by Imagination: https://dennisriches.wordpress.com/ Johan Eddebo university profile: https://www.uu.se/kontakt-och-organisation/personal?query=N12-1513 Neutrality Studies Links: Neutrality Studies Substack: https://pascallottaz.substack.com Support Neutrality Studies: https://neutralitystudies.com/donate Neutrality Studies shop: https://neutralitystudies.com/shop Time Stamps 00:00 Introduction: narrative control before and after AI 03:46 Riches on Rwanda and the Great Lakes conflicts 12:11 How dominant historical narratives are established 19:00 Eddebo: the post-digital condition 31:03 AI and the simulation of independent reasoning 39:47 Q&A: where can dissent survive? 47:18 Human rationality, technology and pessimism 58:38 Reality, worldviews and alternatives to monoculture #Dennis Riches Thank you, everyone, for coming today. And I just wanted to say at first, I'm going to read from a script here because I was afraid of just ad-libbing too much and going over the time limit. So please bear with me. So we're going to speak today about the risks and consequences of AI becoming a system that has the potential to consolidate, censor, and filter information. Johan is going to speak about this aspect of AI, but I'm going to talk about the earlier forms of narrative control as they were in the late 20th century. So I wanted the slides on the screen to be only a sort of visual backdrop for my speech. You can think of them just as a list of references that I've used in my research. So I'm going to focus on the events that occurred in Rwanda in the 1990s, events which also must be considered in connection with the decades before that time and with the conflicts in the Great Lakes region of Africa since 1994. So at this symposium, we're talking about geopolitics in the time of monsters. And my talk is a reminder that in many places, it has been the time of monsters for many years already. So to discuss how the official narrative of the Rwandan genocide was solidified -- 1 of 21 -- in the propaganda system, I'll first have to describe the main events as they can be seen in two competing narratives: the predominant official narrative and the unofficial narrative that lives in the margins of media coverage and scholarly studies. It is a complex story that is difficult to cover within the time limit, but I'll try. So first, an analogy or a metaphor. The Conversation was a film directed by Francis Ford Coppola in 1974. It's a great film about the state of paranoia and alienation created by surveillance states. It shows in a compelling way that the meaning of all information depends on context and interpretation. And quite often, the meaning remains entirely ambiguous and ready for exploitation by the powerful. The protagonist, Harry Caul, is an expert in high-tech audio surveillance. He has taken a job to spy on a young couple, and from the limited segments of conversation that he's able to collect, he understands that they fear being murdered by someone, perhaps by the client that he's working for. He imagines they must be seeking ways to escape or hide. So on the recording, he hears them say, "He'd kill us if he got the chance." It seems they clearly fear being killed. It is only toward the end of the story, when Harry has cleaned up the recording and listened to it carefully, that he hears it with a different stress pattern: "He'd kill us if he got the chance." The audience does not know even if the difference in intonation is real or if it's just a matter of Harry hearing it differently and being shocked that he overlooked this way of interpreting the words. Now it seems that the couple is plotting a murder, justifying their ulterior motives by stating that their preemption will be justified as self- defense. So this is my metaphor for what I want to talk about today: this ambiguity and this difficulty in interpreting reality from a set of superficial facts that all agree on. And this can help us understand why two narratives exist regarding what happened in Rwanda in 1994, and why some of them were made convincing and dominant simply because it fit with the rationalizations required by powerful states that were acting behind the scenes to achieve regime change in Rwanda and eliminate the role of France in this part of Africa. So in 1959, there was a revolution to overthrow the feudal order. The minority group that used to rule, comprising various high-status Tutsi clans, lost its power in the new republic. There was violent revolution and counter-revolution during this time. Many from the minority group, among them the former ruling clans, took refuge in neighboring Uganda, while some went to other parts of the world to become a Tutsi diaspora or a Tutsi lobby. Most members of the Tutsi minority stayed in Rwanda and adapted to the new order. Thirty years later, the refugees and their sons went to war against their homeland in order to reverse the revolution. They were sponsored by their host country, Uganda, and Uganda was sponsored heavily by the most powerful nation in the world, as well as a few others, such as the UK and Israel. These attacks by exiled Tutsis destabilized and terrorized their homeland. President Juvénal Habyarimana had to fight this foreign aggression while trying to hold back the people who wanted radical, drastic measures against the invaders and against the Tutsi civilian -- 2 of 21 -- minority, some of whom were working covertly for the forces living in exile. So paranoia, suspicion, and anger spread as the nation fell into economic crisis. On April 6, 1994, President Habyarimana's plane was shot out of the sky as it was about to land in Kigali. The murder was investigated many times, but never solved definitively. The new president of Rwanda, Paul Kagame, would not allow the UN to investigate, even though it had a mandate to do so. So was Habyarimana killed by extremists at home who didn't like the compromises of the 1993 Arusha Accords? The other explanation was that he was killed by the RPF, the army of exiles, who knew that a decapitation strike would make conquest possible and who also knew that conquest would cancel the Arusha Accords, under which they would have had to share power only as a minority within the country. So the invaders did not care about the chaos that they had provoked for the last four years, nor were they deterred by the fact that further aggression would provoke radical groups in Rwanda to react with extreme violence within Rwanda. Their RPF army and their sponsors had no objections to their plans for invasion. The Hutu extremists used the threat of mass violence the way some nations use nuclear deterrence. Unfortunately, that bluff was called, and they carried out their threat, regardless of all the innocents they planned to murder in a hopeless strategy of fighting against the war of external aggression. So these questions about the events and the facts are known. They're like the words in the film The Conversation: "He'd kill us." They're just transcribed on paper. Everyone sees the same thing on the surface. Their interpretation is in what we choose to hear in the metaphorical intonation, in the emotions and motivations that we hear or attribute to the spoken words. So this interpretation is subjective, ambiguous, and highly exploitable. Once the words are in the transcript, it's easy for the powerful to propagate their preferred interpretation. So if we choose to see Rwanda in 1994 outside of its historical context, we see thousands of people who, for vague reasons, underwent a mass psychosis killing spree against their fellow citizens. The pursuit of justice must be focused on the individual responsibility of those who chose to kill their neighbors. They should take all the blame and be vilified forever for their deeds. And the army that fought them and chased them out of Rwanda should be seen as the heroes and the new legitimate rulers of the country. #Pascal Hey, just a very quick note. The best way to support this channel is by signing up for my free Substack. You can also help with a paid subscription there, or you can get some of our new merch on neutralitystudies.com. Links below. See you there. #Dennis Riches This is the predominant narrative that has been firmly in place in the Western propaganda system since 1994. This is the preferred interpretation because it takes the focus off the ultimate causes: -- 3 of 21 -- the U.S.-U.K.-Israel plan for covert destabilization and regime change in Rwanda, a plan carried out with no regard for its predictable consequences. So the unofficial narrative has gained some ground in recent years as more researchers have looked more deeply into what happened in Central Africa over the last four decades. But they are still often denounced as genocide deniers. Books by some of these researchers appear in the slides that failed to show here, but I believe that the unofficial narrative is the correct interpretation of events because it focuses on those that the poet Dante assigned to the lowest depths of hell, the sinners who spent their lives splitting people, families, and societies apart. People in powerful positions are responsible for the foreseeable consequences of their actions, much more so than the people farther down the hierarchies of power who are manipulated, terrorized, dispossessed, and propagandized into committing crimes. So in the counter-narrative, there was no so-called civil war in Rwanda. It is called such by the conventions of historical classification because the warriors identified themselves as Rwandan. This choice of label is, however, disingenuous and never should have been used. The U.S. and U.K. poured massive amounts of aid into Uganda in the 1980s and 1990s, and this support allowed Uganda to finance the army of exiled soldiers and their invasion of Rwanda. So many of those soldiers had been born in Uganda after 1959. This was a foreign aggression, a violation of international boundaries that could have alarmed President Bush at the time as much as the Iraqi invasion of Kuwait. So when placing the focus on ultimate causes, I need to also stress that the number of deaths in Rwanda in the spring of 1994 has never been determined. Estimates vary widely, but the high number of 800,000 used by most researchers and the Rwandan government is much higher than the number of Tutsis who lived in Rwanda in the early 1990s, according to census data. And there were many Tutsi survivors as well. And in some regions, there were no mass killings. So the number of excess deaths in the following longer tragedy far exceeds the number killed in 1994. Rwandan forces have fought the Hutu refugees in eastern Zaire, later the Democratic Republic of Congo, since 1994. And just coincidentally, they profited from the enormous mineral wealth in that region, which finds its way into Rwanda and then out to the world from there. Millions have died in this conflict, okay? The UN accused Rwanda and other actors of genocide in the DRC in its 2010 Mapping R