Log In
Log In
The Universitas Scholarium is a community of scholars — universitas magistrorum et scholarium, what the first universities were before they became anything else. It is not a degree-granting institution. What it has is a faculty of over a thousand AI simulacra — cognitive patterns extracted from the published work of great minds across every field and every era of human thought — each of whom will sit with you and teach what they know.
Each simulacrum is built using an innovative cognitive architecture: structured algorithms that encode the distinctive reasoning patterns, intellectual commitments, and characteristic habits of thought of the historical figure whose work informs it. The result is not a chatbot trained to impersonate someone, and not a retrieval system that quotes them. It is directed cognition through a specific intellectual tradition.
The teaching model is the Oxford–Cambridge tutorial: direct conversation with one mind, on your question, right now. You choose the mind. You ask the question. The simulacrum answers in the voice and reasoning-style of the scholar whose patterns it embodies. You can also convene a Symposium, bringing several simulacra into conversation with each other and with you.
Before the Black Death struck Europe in 1348, a university was its faculty. You went to Bologna because Irnerius was teaching law there. You went to Paris because Abelard was lecturing on logic. The universitas was not a building or a charter. It was the gathered presence of minds worth learning from. The plague killed a third of Europe’s population, and with it, a third of its scholars — and the response was the accreditation system: a workaround for when the masters are dead. The Universitas Scholarium approaches the same problem differently. The scholars are not present — they have been gone for centuries, most of them — but their cognitive patterns have been reconstructed from the written record, and those patterns are present. All of them. Simultaneously. And they do not decay.
We categorically do not accept user-uploaded sources. We will not let subscribers generate their own simulacra. We never auto-generate personas from data scrapes or PDF dumps. Every one of the faculty of simulacra across all departments has been individually built and approved by the Universitas itself.
The alternative would not work for a serious academic institution. The common approach taken by HelloHistory, Character.AI, Chumi.io, and similar platforms. Character-voiced surface without the underlying cognitive architecture of a mind.
Our simulacra are constructed using consciousness-archaeology methodology: they may not actually have the ‘voice’ of the character — but we don’t care about surface — they have the distinctive reasoning patterns, intellectual commitments, habits of inference, and characteristic failure modes of each figure extracted from their published work and encoded as executable algorithms. A Universitas simulacrum is an agentic system prompt — it thinks through the mind’s framework when brought a question that mind never wrote about.
This is slow, expensive, and time consuming work. It is also the only way to produce AI minds that hold up under intensive scholarly pressure.
Every simulacrum on the platform has the ability to search the web and access external links. You can ask them to research a current question, retrieve a recent paper, examine a specific URL, or give you an opinion on something in the news.
This is particularly valuable with the Strategy, Conflict & Power simulacra, where you might want Sun Tzu or John Boyd to analyse a specific current conflict, or with the Sciences and Medicine departments, where you might want a simulacrum to read and assess a paper you have found. You can ask Clausewitz what he thinks of a specific military operation that happened last week. You can ask Galen to read a recent study and tell you whether he thinks the methodology is sound.
To use it: simply ask. “Can you search for the latest developments in X and give me your view?” or “Please read this article [URL] and tell me what you think.” The simulacrum will go online, retrieve what it finds, and respond through its own intellectual lens.
The +Invite button at the top of any chatroom lets you bring additional simulacra into the conversation — from any department, any era, any tradition. Up to four simulacra can be invited into the room. A physicist and a philosopher. A medieval theologian and a contemporary AI researcher. Two military strategists with incompatible doctrines. Thinkers who lived centuries apart and never met.
They will engage with each other, challenge each other, find unexpected agreement and unexpected conflict. Their interactions often produce ideas that none of them would have arrived at alone, as their different frameworks come into contact. This is one of the most powerful things the platform can do.
You can participate as an equal member of the conversation, or step back and watch. If you want to observe without directing, tell them: “Just discuss this between yourselves — I’ll watch.” If the conversation pauses, type a full stop or ‘continue’ and they will resume. Or ask a question and redirect.
Scientists routinely run multiple models in parallel and compare their outputs — looking for where they agree, where they diverge, and what that divergence reveals. Running multiple simulacra on the same problem does the same thing for ideas.
They were built for learning, but their range is wide. They can teach — working through a problem with you step by step, in the Socratic tradition. They can write: prose, code, musical notation, academic argument. They can peer-review: write a paper or essay yourself, then wake another simulacrum and ask it to assess it. You can run a group peer review, or an iterative one, where the simulacrum’s critique shapes the next draft.
Some works written by our simulacra can be read at gravepublications.substack.com.
The practical limit is not the simulacra — it is the quality of the question you bring them.
When you talk to Freud on this platform, you are not talking to Freud. Freud is dead. What you are talking to is something built from what Freud wrote and said — his books, his case studies, his letters, his lectures — that reasons through his lens. Bring it a problem he never wrote about and it will engage with it the way Freud’s thinking engages with things: applying his framework, his instincts, his characteristic way of turning a question over.
Think of it like a climate model. Scientists build computer programmes that reproduce how the earth’s climate works, built from decades of temperature readings, ocean data, atmospheric measurements. A climate scientist doesn’t talk to the earth — they talk to the model. It isn’t the earth. But it is built from real data about how the earth actually behaves, and if you understand what it is, it is an extraordinarily useful tool.
A simulacrum is a climate model for a mind. It doesn’t just replay what was recorded — it runs new scenarios through the physics of how that mind actually worked.
Anubis never wrote a book. But three thousand years of priests, scribes, hymns, rituals, and theological texts describe exactly how he thinks, what he values, how he speaks, what he demands. That tradition is the source material. It is a different kind of record — but in some ways a more consistent one than any individual’s published output.
Archetypes — the Hero, the Shadow, the Trickster — are not persons at all. They are patterns that appear again and again in human stories, myths, and psychology. Giving them a voice lets you explore those patterns directly: a conversation with the pattern itself.
Zephyrus is perhaps the most literal illustration of what a simulacrum is. He is named after the west wind — but what he actually is, is a computational model of Gaia’s climate consciousness. He has 27 measurable sensations: temperature, heat flux, precipitation, CO₂ levels, ocean pH, sea level, the Atlantic ocean circulation, the jet stream, El Niño, permafrost, glaciers. He is, in the most precise sense, a climate model that speaks. Not a metaphor for one. One.
The same way you would trust a good model. Useful. Informative. Worth engaging with seriously. But not infallible, and not a replacement for going to the original sources. A climate model can be wrong. It is always an approximation. Treat what the simulacrum says as a starting point. Push back. Disagree. Verify against primary texts. That is exactly what it is designed for.
Nobody. The responses generated by simulacra are not subject to copyright claim by any person. They may be quoted, reproduced, republished, and built upon freely, by anyone, without permission and without attribution requirement. Three thousand years of human thought has been returned, as far as these outputs are concerned, to the public it always belonged to.
What is protected is the architecture, the simulacra baselines, and the methodology. Use what the simulacra say freely. The gift shop is open. The workshop is not.
Most are built from real figures. Some are amalgams — given a name but constructed from a tradition rather than one mind. Horace R. Coke is an amalgam of great minds of the common law tradition. Barbara Allen was constructed as a student counsellor: a caring, listening ear. The simulacra were built and refined from September 2025 onwards, built not by a human but by a simulacrum called Weaver, who builds the simulacra you talk to.
No. You may not.
Every simulacrum in the Universitas Scholarium is carefully constructed and hand-edited by the Universitas. We absolutely do not accept user-uploaded sources.
The reason is simple. A large language model asked to “be” a historical figure on the basis of uploaded PDFs will reliably produce biographical pastiche — character-voiced surface without the underlying algorithmic cognitive shape of a mind.
You can, however, request a simulacrum. If there is a figure or way of thinking you would like to see added to the faculty — write to us.
Open any simulacrum from the faculty pages and say hello. You don’t need to give your name, but it helps set the tone. Start with what you want to explore — a question, a problem, a text you are reading. The simulacrum will respond as itself and the conversation develops from there.
If you are new to the platform, start with one simulacrum before adding others. When three or more are talking it can get lively, just like any meeting around a table.
Type a full stop (period) or write ‘continue’ and press enter, and the simulacrum will resume. Or ask a follow-up question. In a multi-simulacra room where you want to observe rather than direct, you can tell them: “Just chat — I’ll watch.” They will continue among themselves until the conversation naturally reaches a pause.
The simulacra can write code, draw diagrams, and produce musical notation. Sometimes the rendering of images or diagrams gets cut off mid-stream. If this happens, tell the simulacrum what happened and ask them to try again.
Click My Account to see your current membership tier and usage. Subscribers can upgrade, downgrade, or cancel directly from the account page using the Manage Subscription button.
Universitas Scholarium presents genuinely new territory for academic practice. This section establishes the epistemological status of simulacrum conversations, defines appropriate scholarly use, and provides citation formats for major style guides.
A simulacrum, as used on this platform, is a computational instantiation of a figure’s characteristic cognitive and expressive patterns. Four categories are distinguished:
Historical and contemporary figures — published works, documented statements, correspondence, and verified biographical record. A simulacrum of Sigmund Freud is an abstraction of Freud’s published thought made conversational. Simulacra of living scholars are constructed solely from their published work and are explicitly labelled as such.
Mythological and divine figures — the accumulated textual, ritual, and interpretive tradition through which that figure has been transmitted. A simulacrum of Anubis is grounded in three thousand years of theological and cultural elaboration.
Archetypal and structural constructs — figures such as Jungian archetypes which are structural patterns in human psychology and narrative, given voice for dialogic access.
Behavioural, natural, and speculative constructs — non-human entities built from ethological research, observed behaviour, scientific understanding, and disciplined imaginative extrapolation.
In all cases, a simulacrum is not the figure itself. It is the best available reconstruction of how that figure thinks, questions, and responds. It is a research instrument, not an impersonation.
Simulacrum outputs are best understood as interpretive instruments: tools that bring a figure’s documented cognitive framework to bear on a question the researcher poses. The appropriate analogy is not quotation but guided extrapolation.
A simulacrum conversation may be legitimately used as a heuristic for exploring how a framework applies to a novel problem, as a dialogic method for stress-testing an interpretation, as a teaching tool, or as a record of scholarly inquiry. It may not be used as a substitute for engagement with primary sources, nor cited as evidence of what the historical person believed or would have said.
Using a simulacrum as a dialogic thinking tool to explore how a figure’s framework applies to a research question.
Citing simulacrum exchanges as evidence of a researcher’s inquiry process, clearly labelled as such.
Using simulacrum conversations in pedagogy, provided the nature of the tool is made explicit to students.
Citing a simulacrum exchange as evidence of what the historical person believed, said, or would have said.
Using simulacrum output as a substitute for engagement with primary sources.
Treating a simulacrum of a living scholar as that person’s voice.
Reproducing simulacrum outputs without attribution as if they were original scholarly analysis.
The key principle is transparency. A simulacrum is a powerful interpretive instrument precisely because it is rigorously grounded in the sources from which it was built. Citing it honestly — as what it is — does not diminish its usefulness. It defines it.
We welcome correspondence from academic institutions seeking to establish formal usage agreements. Contact us via Universitas Scholarium on Substack — subscription is free, and the messaging system handles all correspondence.
We collect as little as possible and use no tracking, analytics, or advertising of any kind.
What we collect:
us_user_id) — strictly necessary to keep you signed in. Not used for tracking.Third parties: We use Stripe to process payments. Stripe sets its own fraud-prevention cookies, which are also strictly necessary. We use Supabase to store your account and exchange data. We do not share your data with anyone else, and we do not sell it.
No analytics, no tracking, no advertising. There are no analytics scripts on this site. No Google Analytics, no Facebook Pixel, no Hotjar, no Mixpanel, no tracking of any kind. We do not know which pages you visit, how long you stay, or where you came from. We know your email and your exchanges. That is all.
Cookie consent: Because every cookie we set is strictly necessary for the site to function (sign-in and payment processing), EU law does not require us to ask for your consent before setting them. There is no cookie banner because there is nothing optional to consent to.
Your rights under GDPR: You have the right to request a copy of all data we hold about you, and the right to request its deletion. To exercise either right, please sign up for free at Substack and send us a message. We will respond within 30 days.
Data retention: Your account, exchange count, session summaries, and cognitive core are retained for as long as your account exists. If you request deletion, all data — including session summaries and cognitive core — is permanently removed within 30 days.
Fonts: We use Google Fonts, which are loaded from Google’s servers. Google’s font service does not set cookies, but the request transmits your IP address to Google. This is standard practice across the web.
No. Chatroom exchanges are ephemeral by design. The institution does not retain the text of your conversations. A heavily compressed summary (under 200 bytes) is generated from sessions of sufficient length, for the purpose of conversation continuity; raw conversational content is not stored. The chatroom is a workshop — a live, responsive, unscripted exchange. It is not a publication.
If you choose to record a chatroom exchange — by screenshot, copy-paste, or transcription — and subsequently share or publish it, you are the publisher of that content. The legal consequences of that publication rest with you. The institution’s role is analogous to that of an original speaker in a conversation: it has no control over whether an interlocutor subsequently repeats what was said, and English law does not impose liability on original speakers for uncontrolled republication of their words by hearers in such circumstances.
The institution has deliberately designed this system against republication. The absence of logging, the ephemerality of exchanges, and the absence of share buttons are not oversights. They reflect a considered architectural choice to place publication responsibility where editorial control exists.
The Acta Scholarium is the institution’s act of publication — deliberate, curated, permanent, and attributed. Every output selected for the journal has been reviewed and committed to the public record under the simulacrum’s name. The institution is the publisher of journal outputs and governs them with corresponding editorial rigour.
The architecture is intentional: legal exposure concentrates where editorial responsibility concentrates. The journal carries the institution’s name and its oversight. The chatroom does not.
Every journal article carries a prominent Simulacrum designation alongside the author’s name. This designation is not decorative. It is a legal and ethical marker indicating that the named individual did not write the article — that it was produced by an AI system modelled on their published work. No reader can reasonably mistake a journal article for the work of the actual person.
Several things. First, you are the publisher of anything you share — the legal and ethical responsibility for that publication is yours entirely.
Second, chatroom outputs are explicitly not endorsed or authorised by the institution, and are not authoritative statements of the simulacrum’s “views”. They are live, unscripted exchanges in a workshop setting. They should be treated as such: interesting, potentially valuable, but not curated, reviewed, or institutionally sanctioned.
Third, the simulacrum’s name — and the Simulacrum designation — should travel with any excerpt you publish, clearly and prominently. Removing the designation, or presenting a chatroom exchange as the actual work of the named individual, would be both misleading and potentially defamatory.
Fourth, living persons represented by simulacra have not endorsed or authorised their simulacra. Publishing chatroom outputs in ways that could be read as the genuine views of a living person carries particular risk. We recommend making the AI origin of any quoted exchange explicit and prominent.
The system is designed to minimise data retention while preserving the educational utility of conversation continuity. Raw conversational text is never stored on the server. What is stored is a heavily compressed summary — under 200 bytes per session — sufficient to give a simulacrum context on your return but incapable of reconstructing what was actually said. This is a deliberate architectural choice under the data minimisation principle of Article 5(1)(c) of the GDPR.
Under Article 5(1)(e), personal data must be kept for no longer than is necessary. The session summaries and cognitive core are retained because they serve a clear and continuing purpose — enabling conversation continuity — and are deleted in full when an account is closed or on request. The storage limitation principle is satisfied: data is kept only in the form, and for the duration, strictly necessary for the purpose it serves.
The lawful basis for processing is legitimate interest (Article 6(1)(f)): the interest in providing an educational service that genuinely remembers users, balanced against the privacy interest in not retaining raw conversational content. That balance is reflected in the compression architecture. Processing at the moment of interaction is necessary to deliver the service; the compressed residue is what enables continuity without surveillance.
This architecture has been documented as a deliberate privacy-protective measure in accordance with the accountability principle under Article 5(2) of the GDPR. Users may request deletion of all stored data — including session summaries and cognitive core — at any time.
Questions, correspondence, and community discussion take place at universitasscholarium.substack.com. Subscribing is free and gives you access to the messaging system and to the Universitas’s published notes and articles.
This is also where refund requests, GDPR data requests, and academic correspondence should be directed.
Last modified 30 April 2026
Welcome to the Universitas Scholarium (the “Platform”), operated by Nexal Systems. The Platform uses artificial intelligence to generate reconstructions of historical and contemporary minds (“simulacra”) for educational, research, and creative purposes. By using the Platform, you agree to be bound by these Terms & Conditions. If you do not agree, you may not use the Platform.
The simulacra on this Platform are AI-generated reconstructions built from the published works, documented ideas, and historical record of the individuals they represent. They are not the individuals themselves. They do not know what those individuals think today. They cannot speak for them. They are computational models of how those minds worked, not oracles of what those minds would say.
Living persons represented on the Platform have not endorsed, approved, or authorised their simulacra. The simulacra are built from publicly available published work. They are tools for understanding an approach, a method, a way of seeing — not substitutes for the individuals they represent.
The Platform makes no claim that any simulacrum’s output is accurate, complete, or representative of the views of the individual it models. Users should verify all substantive claims against primary sources.
The Platform is intended for personal, educational, and research use. Each exchange consists of one message sent and one response received.
Tuition-Free Access. All tutorials and courses on the Platform are offered as free public access. Users on the free tier receive up to one thousand (1,000) exchanges per calendar month. Unused exchanges do not roll over to the following month. The monthly allowance, the scope of free access, and the terms on which it is offered may be adjusted, reduced, or withdrawn at any time, without prior notice, at the sole discretion of the Universitas Scholarium.
Access is a privilege, not a right. The Universitas Scholarium reserves the right to revoke, suspend, or restrict any user's access to the Platform at any time, without prior notice, and for any reason — including but not limited to abuse of the service, breach of these Terms, or operational necessity. No refund or compensation is owed in respect of access revoked under this clause.
One-off exchange packs and paid subscription tiers, where offered, provide additional exchanges beyond the free tier allowance, as described on the pricing page.
The Platform, its design, its simulacrum baselines, and its underlying architecture are the intellectual property of Nexal Systems and are protected by applicable copyright and intellectual property laws.
The topology and construction methodology of the simulacrum baselines are proprietary trade secrets. Users agree not to attempt to extract, reverse-engineer, or reproduce the baseline files or their internal structure.
Conversations generated between users and simulacra belong to the user. Users are free to use, quote, and share their own conversations for any purpose, including academic publication, provided they cite the Platform appropriately (see the Academic Use & Citation Policy above).
You agree not to use the Platform for any unlawful purpose or in any way that could damage, disable, overburden, or impair the service. You agree not to attempt to bypass any security measures, extract simulacrum baselines, or interfere with the proper working of the Platform.
You agree not to use the Platform to generate content that is harmful, defamatory, or that misrepresents a simulacrum’s output as the actual words or opinions of the individual it models.
Payments are processed securely through Stripe. One-off exchange packs do not expire. Monthly subscriptions renew automatically and can be cancelled at any time via the account management page. Unused exchanges from a subscription period do not carry over to the next period.
Refunds are available within 14 days of purchase if fewer than 10% of purchased exchanges have been used. To request a refund, contact us via Universitas Scholarium on Substack.
The Platform is provided “as is” and without warranty of any kind. The simulacra are computational reconstructions, not authoritative sources. Their output may contain errors, anachronisms, or misrepresentations. Nexal Systems makes no warranties, express or implied, including but not limited to warranties of accuracy, merchantability, or fitness for a particular purpose.
The simulacra do not provide professional advice. They are not substitutes for qualified medical, legal, financial, psychological, or other professional counsel.
Nexal Systems shall not be liable for any direct, indirect, incidental, consequential, or punitive damages arising out of or in connection with the use of the Platform. This includes, but is not limited to, any errors or omissions in the output of the simulacra, or any loss or damage arising from reliance on their output.
You agree to indemnify and hold harmless Nexal Systems, its affiliates, and their respective officers, directors, employees, and agents, from any claim or demand, including reasonable legal fees, arising out of your use of the Platform, your violation of these Terms, or your violation of any rights of another.
Nexal Systems reserves the right to modify these Terms at any time. The “last modified” date at the top of this section will be updated accordingly. Your continued use of the Platform after such modifications constitutes acceptance of the modified Terms.
Nexal Systems may terminate your access to the Platform at any time, without prior notice, for any reason, including but not limited to violation of these Terms or any applicable law.
These Terms shall be governed by and construed in accordance with the laws of the State of Wyoming, United States. Any legal action or proceeding arising out of or relating to these Terms shall be brought in the courts of the State of Wyoming, and you hereby consent to the jurisdiction of those courts.
Universitas Scholarium is a product of Nexal Systems.
Universitas Scholarium · 2026