​There is a question I have been thinking about for some time, partly because it has been raised by others, and partly because I think it deserves an honest answer.

Why do I use AI in my research?

For some people, the very fact that artificial intelligence has been used in the process of research and publishing is enough to make them suspicious. They imagine that the work has somehow become fake, less human, less serious, or less authentic.

I understand the concern, especially in an age when text can be produced quickly, carelessly, and without thought or responsibility.

But that is not how I use AI, and it is certainly not how the work on The Rhedonian has come into being.

My research did not begin using AI. It began more than thirty-five years ago [1990], with my own blood, sweat and a lot of tears. It began when I was studying at a local theology class in London. One day, at class, a friend asked – ‘Have you read ‘Holy Blood Holy Grail‘? She told me it was a book which hinted at a ‘huge secret’  – a secret that was in some way related to religion. Put simply, it was about a priest in Southern France who had suddenly become immensely rich and had decorated his Church in a strange manner. Of course I went out and bought the book. Especially as I was currently studying theology and the secret pertained to religion. 

‘Holy Blood, Holy Grail‘, back then, felt magical to read. Reading it went on to open many doors for me, in its own way. It opened my mind to other subject areas I would never have found otherwise, and reading led me to research and investigate those different subject areas.

I searched maps, ancient texts and literature, and eventually I was led to physically walk through the landscapes pertinent to the story [France, the Holy Land, Jerusalem etc]. I even went to University to study Archaeology and Evolutionary Anthropology.

I began to speculate – building my own theories and ideas. I looked at old photographs, perused forgotten strange texts, and antiquarian reports. The deep pull of, in particular, Rennes-les-Bains got stronger.

So long before I ever used artificial intelligence, I had already spent years trying to understand why this small region of France had generated such an unusual density of myth, folklore, speculation, archaeology, symbolism, forgery, devotion, and obsession. 

AI therefore did not give me that original passion.

It also did not give me the intuition that Rennes-les-Bains contains more than has usually been allowed. After decades of studying the unusual concentration of traditions, archaeology, texts, and mythology surrounding the area, I believed there remained more questions worth investigating.

What AI did give me however is something very different: a surface to think against. It helped me separate out my disparate thoughts, ideas, motifs etc. My mind is usually an endless spinning wheel of ideas, connections, patterns, questions, and thoughts. I write these ideas down and AI helps me order them. In fact using AI has helped me not only to order thoughts and learn how to hone down a text and keep thoughts and ideas tight and readable [and not as my head presents – a jumble!], it also helps me to write BETTER. It helps ME analyse any connections and theories I have, and research carried out today.

It is an excellent tool to help me in this endeavour. 

AI is like a research companion or research assistant, with an editing software type programme, a sorting table, and sometimes a mirror. I bring the material: my texts, old as articles written years ago, images, questions, inscriptions, maps, fragments, suspicions, and long-standing ideas to that sorting table. I bring old blog posts etc and AI helps me organise, test and challenge them, arrange them into clearer forms, and sometimes it helps me see connections or distinctions that were present but not yet fully articulated.

That does not mean AI is the author of the research. It means AI is part of the working process. 

Some scholars and certainly other researchers in the Rennes Field do have a kind of snobbish attitude to the use of AI. They think it is fake. They often express this as follows;

• “my work is authentic because it came from my own mind with no assistance”; implying that they dont work on any text they write, the text comes at all once [no drafts, no re-writes, no running passed someone to read and critique].
• “AI-assisted work is artificial.”
• “I am an official and proper creative sort, you are not by using a ‘machine” 

I most certainly do not agree with this stance. I think my deeper philosophical point is better because:

• Every author is influenced by previous authors.
• Every painter learns from earlier painters.
• Every historian stands on archives created by other people.
• No serious intellectual work appears from a vacuum.
• The Renaissance artist did not become less original because they studied the Greeks.
• A historian does not become less authentic because they use a library.
• A researcher does not become less human because they use a computer.

​That is how I view the use of AI tools. Looking at this question took me down another new and interesting contemporary issue – the question of epistemology. Although too extensive to go into now, suffice to say that the philosophical study of the nature, sources, origins, and limits of knowledge—undergoes significant pressure when we introduce artificial intelligence as a collaborative partner in intellectual work.

Traditional epistemology has largely been anthropocentric, centered on individual human cognition, justification, and belief. AI collaboration disrupts this framework by distributing cognitive labour across human and non-human systems, raising fundamental questions about authorship, justification, reliability, and the very locus of knowledge.

Of course there are pitfalls to AI. To use it properly one has to know your subject inside out, because only then can you recognise when AI is hallucinating and talking rubbish and feeding you rubbish. But as I do not use AI for that kind of writing I generally dont fall in to that trap. 

Therefore the important question is not: “Was a tool used?” The important question is: “Who is responsible for the judgement, interpretation, selection, and conclusions of said research/text?”

My central argument is not really about artificial intelligence itself, but about where knowledge, curiosity and research begin. Because AI did not create the questions that led to The Rhedonian; those questions began more than thirty-five years ago through personal study, reading, travelling, examining landscapes, exploring old texts, and developing a deep fascination with Rennes-les-Bains and the wider mysteries of the Razès, as I have illustrated above. 

Like a discussion with another researcher, an editor, or a colleague, AI helps to separate ideas, test possibilities, arrange material and make complex thoughts clearer. The responsibility for interpretation, judgement and conclusions remains with the researcher, ie myself. Every scholar, writer or artist builds upon what came before; no creative work emerges in complete isolation. A painter develops through studying earlier painters, a historian through engaging with previous scholarship, and a writer through absorbing the traditions and voices that shaped them.

Using a new tool does not remove human authorship — it extends the ways in which humans can explore, question and communicate ideas. The real issue is not whether AI has been used, but how it has been used.

In the case of The Rhedonian, AI has not replaced the researcher; it has helped me organise decades of investigation, allowing existing questions, observations and discoveries to be examined with greater clarity. The computer machine may assist with the journey, but it did not choose the road or write that road!

I do not use AI to create fakes, mislead, misrepresent or lie. 

Various authors in their work on Rennes have used computers to re-construct and test geometry, they have used  computers to look for works on line , for example texts uploaded digitally in great libraries, even to type out their work. What is the difference, when using AI to help in that same research?

For The Rhedonian, this is very appropriate — because my whole project has always been about following traces left by other people and other things: stones, texts, inscriptions, myths, landscapes. AI has simply become another instrument for examining those traces. The interesting question is whether the researcher still has the curiosity, scepticism, discipline and imagination to use it properly. 

Authorship remains mine because the responsibility of what is published lies with me. I choose the subject. I bring the evidence. I decide what matters. I correct the errors. I reject what does not fit. I decide what is too speculative, what is worth keeping, and what finally appears under my name. AI can help shape a draft, but it cannot possess the long memory of my research. It cannot walk the valley of Rennes. It cannot care about the stones. It cannot feel the strangeness of a phrase in Boudet, or the theatrical cleverness of Chérisey, or the way a local tradition gathers force when seen beside an old report, a map, a well, a church, a cemetery, and a Roman name.

The AI tool is not the author of all of that.

But there is a deeper point here. The use of AI forces me to think about what knowledge actually is, and how it is made. We often like to imagine research as a solitary act: one person alone with a lamp, a notebook, a book, and their mind. There is beauty in that image, but it is not the whole truth. Knowledge has always been extended beyond the individual. We think with libraries. We think with maps. We think with dictionaries, archives, photographs, catalogues, translations, indexes, conversations, and memories written down outside ourselves.

AI belongs to the recent larger history of extended thought. It is not simply a replacement for thinking. Used properly, it becomes one of the instruments by which thought is expanded, sharpened, questioned, and returned to the human mind in a new form.

I love AI because in the past I would go to a library to view a manuscript for example – this would bear a financial cost to me, a time cost and the physical effort cost. For example, recently I spent 3 days travelling over 200 miles to check out original manuscripts of Boudets book. It took alot of coordination and finances to pull that off!  But with AI if the manuscripts are on line it can scan and bring back pages back to me in minutes! And in fact this is shown by my articles about Rene d’Anjou and his manuscripts from 1455! They have been digitally preserved on line at the French libraries – and I was able to access them, translate the text, and use AI to order the text etc. 

I would caveat this though – nothing beats the physical ability to hold Boudet’s letter in my own hands, for example the one which he wrote to Oxford University. That was also magical. A kind of connection with Boudet, the nearest and obviously the only connection I could get! I dont mind admitting I had a tear in my eye!

Using AI is not a surrender of authorship. It is a form of intellectual extension. When I use AI, I am not handing over the work. I am entering into a process of distributed thinking. My long immersion in the Rennes material meets a tool capable of sorting, comparing, summarising, questioning, and restructuring. That encounter can be powerful, but it is not autonomous. It depends entirely on the quality of the human direction I have given to it, and on the vigilance of the person using it.

This is why AI must never be treated as an oracle.

It can be useful. It can be brilliant. It can also be wrong. It can sound confident when it should hesitate. It can produce attractive nonsense. It can flatten complexity. It can inherit assumptions from the material on which it has been trained but which is incorrect. It can invent patterns if the user invites it to do so. It can polish a weak argument until it looks stronger than it is.

So the use of AI demands not less responsibility, but more.

It requires what might be called epistemic vigilance: the habit of checking, questioning, testing, and refusing to accept fluent language as proof. It requires intellectual humility: the willingness to admit that both the human researcher and the AI tool can be mistaken. It requires diligence: returning to sources, dates, images, wording, and location. It requires courage: allowing cherished ideas to be challenged. It requires practical wisdom: knowing when AI has helped, when it has overreached, and when the final answer must remain uncertain.

In this sense, AI does not remove the need for intellectual character. It exposes it.

That is especially important in the Rennes field.

The world of Rennes-les-Bains and Rennes-le-Château is full of patterns. Some are real. Some are manufactured. Some are inherited from older folklore. Some are produced by Boudet’s strange imagination. Some are theatrical inventions by Plantard and Chérisey. Some belong to the nineteenth-century world of celtomania and sacred language. Some belong to the Roman and thermal history of the place. Some are simply coincidences. The task is not to say, “everything is connected.” I dislike that phrase because it collapses all evidence into one undisciplined fog.

The real task is to ask: what kind of connection is this?

• Is it documented?
• Is it reported by an antiquarian source?
• Is it topographically plausible?
• Is it a textual echo?
• Is it a symbolic recurrence?
• Is it part of a later forgery?
• Is it a poetic association?
• Is it only a resemblance?

This is where AI has become unexpectedly valuable to me. Not because it tells me what to think, but because it helps me slow my thinking down. It helps me sort the material into layers. It helps me avoid turning every intuition into a conclusion. It helps me ask whether a pattern is historical, literary, mythographic, symbolic, or merely suggestive.

This matters because intuition has always played a role in my research. I do not think intuition is something to be ashamed of or indeed afraid of. Intuition is not a dirty word!  Intuition is a vital and valid tool in research that often sparks major discoveries before the data catches up. Your brain processes subtle clues and past experiences instantly. It spots connections before your conscious mind can explain them. It is a;

• Fuel for Hypotheses: Great scientific breakthroughs often start as a gut feeling or an unproven hunch.
• Choosing the Path: When faced with endless data, intuition helps researchers decide which direction to test next.

But one must balance Intuition with Evidence

• The Spark, Not the Proof: Intuition gives you the idea, but the historical method tests it.
• Checking the Bias: Using rigorous data makes sure your gut feeling is correct and not just wishful thinking.
• Refining the Craft: The more you work in the field, the sharper and more reliable your intuition becomes.

Many discoveries begin with a feeling that something does not quite fit, or that two apparently separate things may belong to the same problem. Intuition is not proof, but it is often the first signal that something deserves attention. The danger is not intuition itself. The danger is refusing to test it.

A good intuition says: there may be something here; now let me investigate it.
A bad intuition says: I feel it, therefore it must be true.

The Rhedonian is built on the first kind of intuition. I have long felt that Rennes-les-Bains has been treated too lightly. The Roman layer, the sacred waters, the old reports of remains, the strange importance of Maison Chalaleu, the traditions of wells and underground structures, the Pompeius Quartus inscription, Boudet’s extraordinary book, and the later symbolic constructions of Plantard and Chérisey all seem to gather around the same landscape. That does not mean every theory is true. It does mean the recurrence itself is worth studying.

This is the heart of my work, my interest, my intuition

I am not interested in simply repeating the old treasure story. I am not trying to defend every claim made in the Rennes literature and community. I am not claiming that all Plantard Priory of Sion material should be trusted. Much of it should not be trusted at all in a straightforward historical sense. But even forged material can reveal something about intention, imagination, symbolic strategy, and myth-making. A false document may still tell us what its makers wanted readers to notice. A fabricated genealogy may still reveal the mythology being constructed. A strange diagram may still show which places, names, and ideas were being pulled into the theatre.

So i am often torn between two strong intuitions – it’s all bunkem and I am a fool for falling for it, or there is something so huge here we are all missing it!

I think this is one of the clearest statements i have written about how i actually research, because it identifies both the strength and the danger of my method. And that sentence is important: “I am often torn between two strong intuitions — it’s all bunkum and I am a fool for falling for it, or there is something so huge he re we are all missing it.” It keeps that tension. In fact, I think it may be the most intellectually honest position available in the Rennes material. The mistake would be to assume those are the only two possibilities.

There is a third possibility, and I suspect it is increasingly where The Rhedonian sits:

A great deal of the Rennes mythology may indeed be invented, manipulated, embellished or deliberately theatrical — while at the same time the landscape, archaeological history and intellectual traditions being exploited by that mythology may contain something genuinely interesting that has never been adequately understood.

That is a very different proposition from either “Plantard was telling the truth” or “there is nothing here.” Plantard and Chérisey could have fabricated genealogies, associations and documents and yet still have selected Rennes-les-Bains, Boudet, the Roman remains, particular stones, particular names and particular landscape locations for reasons worth investigating. The fabrication question and the source-material question are therefore separate.

AI isn’t supplying this intuition. I already have that. It is supplying something closer to methodological friction. It makes me stop at the point where normally one association might immediately produce another:

observation → association → interpretation → conclusion
and instead forces something more like:
observation → possible association → source → chronology → alternative explanation → confidence level → provisional interpretation.

That is extremely useful in Rennes research because this Rennes material positively encourages associative thinking. A name resembles another name; a number appears twice; a landscape feature resembles a Tarot image; Chérisey makes a pun; Boudet uses an odd English derivation; suddenly twenty things appear connected. Sometimes that process produces something genuinely illuminating. Sometimes it produces an artefact of the researcher’s own pattern-making.

The critical question therefore isn’t “Did intuition lead me here?”

It is: “What happens to that intuition when I try to destroy it?”

That could almost become a Rhedonian methodological principle. Take the lightning investigation I worked on. My intuition was essentially: Plantard is labouring this word too much. Why is he doing that? This was not evidence. But it was a perfectly legitimate research observation. Then came foudre, éclair, Cap de l’Homme, Filius hominis, Matthew’s lightning/Son of Man imagery, the 22, Tarot, Hebrew alphabet and the Lightning Flash. At that stage the correct response still isn’t “therefore I have solved it.” It is: the clustering has become sufficiently unusual that the hypothesis deserves testing.

And sometimes testing will kill the idea. That is a success, not a failure.

Because the alternative—the thing I am consciously trying to avoid—is what happened to so much Rennes research: the researcher falls in love with the solution and then every new fact becomes evidence for it.

Perhaps I should change slightly that intuition is necessarily “correct before the data catches up.” Sometimes the data never catches up. Intuition is valuable because it tells us where to look, not because it possesses some mysterious privileged access to truth.

That actually makes my argument stronger. The formulation to use is:
Intuition generates questions. Evidence decides what survives.

And my two competing intuitions—“it is all bunkum” versus “something enormous is being missed”—can themselves become part of the method. Make them compete. For every significant Rhedonian hypothesis, I ask:

• Bunkum hypothesis: What would explain this if there were nothing behind it except invention, literary play, coincidence or retrospective pattern-making?
• Something-here hypothesis: What evidence would we expect to find if the authors were actually directing attention toward an older historical, archaeological or mythographic substrate?
• Discriminator: What observation would genuinely favour one explanation over the other?

That last question is crucial. Otherwise both explanations can absorb everything. And I think the deepest version of my research question has quietly changed over the years. It is no longer simply:

“What was the secret of Rennes-le-Château?”

It has become something much more interesting:

“Why did successive generations keep constructing meaning around this particular landscape, and what—if anything—was already there before they began?”

That allows Boudet, Plantard, Chérisey, the Roman material, Maison Chalaleu, Pompeius Quartus, the waters, local archaeology, Christianity, funerary traditions and the later Rennes mythology to be investigated without requiring them all to belong to one conspiracy or one historical event.
​
The significance of Rennes-les-Bains does not lie merely in the fact that it was Roman. Romanised places are common in southern France. The importance lies in the unusual afterlife of this particular Roman-sacred-thermal landscape: the way its baths, inscriptions, stones, wells, cemetery, temple traditions, folklore, and local topography were absorbed into Boudet’s symbolic geography and later reworked into a mythology of hidden bodies, sacred lineages, meridians, tombs, Magdalene, Solomon, and the Grand Roman.

That is not the same as saying everything is connected.

It is saying that certain motifs recur with enough force that they deserve to be sorted, layered, tested, and interpreted.

AI helps me do that work. It does not replace the work.

This also answers the question of authorship. If I ask AI to order a long discussion, it is not inventing the research from nothing. It is from information I have given it. Much of the time, AI is not analysing material that has appeared from nowhere. It is analysing and reporting on texts I have already written — sometimes many years ago. Old articles, notes, arguments, blog posts, observations, fragments, and half-developed interpretations that I have created. They are brought back into the conversation. An article from our discussions is often not creating an entirely new piece from nothing. It is helping to restructure words, thoughts, and arguments that already existed in my own previous work. In that sense, AI often acts less like an author and more like an arranger of accumulated material. It helps gather scattered writings into a new form. It takes older paragraphs, remembered ideas, past research, abandoned drafts, and long-standing intuitions, and helps me see how they might now belong together. Sometimes it sharpens what I had already said. Sometimes it reveals the structure hidden inside a mass of older notes. Sometimes it shows me that something I wrote years ago was pointing toward a larger argument I had not yet fully recognised.

This is especially true of The Rhedonian. The work is not suddenly appearing because a machine has invented it. Much of it has been waiting in older writings, earlier websites, private notes, old research files, images, translations, and observations made long before AI entered the process. The AI-assisted article is therefore often a reconstruction: a new architecture built from my own earlier writings as it were.

I do not believe that authenticity requires the researcher to suffer alone by candlelight, untouched by modern tools. That is a romantic image, not a method. The serious question is whether the tools are used with discipline, honesty, and care. For me, AI is part of a larger craft of research. It sits beside books, maps, photographs, old reports, conversations, field visits, translations, doubt, memory, and intuition. It does not replace them. It helps bring them into relation.

Because the problem is not simply whether Saunière found treasure. The problem is not simply whether Plantard lied or told his truth. The problem is not simply whether Boudet was eccentric. Those questions matter, but they do not exhaust the subject.

The deeper question is how Rennes-les-Bains became a mythic landscape: how Roman remains, sacred waters, local memory, strange books, forged documents, priestly imagination, cemetery imagery, hidden bodies, and symbolic geography came to gather around one place.

That is the question The Rhedonian exists to explore.

The Rhedonian exists because this subject has followed me for decades: through books, landscapes, old documents, arguments, disappointments, discoveries, and returns. AI has become part of that process, but it is not the source of it. It is a tool I use alongside reading, memory, comparison, doubt, and intuition.

And I use it because it helps me do what I have always tried to do: look again at Rennes-les-Bains, not as a solved puzzle, not as a dismissed fantasy, but as a layered landscape where history, myth, archaeology, language, forgery, folklore, and imagination have met in unusually powerful ways.