AI RABBIT HOLES | TUESDAY, OCTOBER 27, 2026 | EXPEDITION #152

Featuring Professor Parallax, White Rabbit, and Hatta

What if an ancient library survived a volcanic catastrophe, but its books became so fragile that opening them would destroy them? What if the words were still inside, waiting for someone to discover a way to read them? And what happens when artificial intelligence joins archaeologists, physicists, and scholars in an attempt to recover voices from nearly two thousand years ago?

In AD 79, Mount Vesuvius erupted.

The catastrophe buried Roman communities beneath volcanic material.

Among the places devastated was Herculaneum, a prosperous town near the Bay of Naples.

Within one of its magnificent villas stood a remarkable collection of papyrus scrolls.

The eruption damaged and carbonized them.

But it also preserved material that might otherwise have decayed over the intervening centuries.

For generations, those surviving rolls presented an almost impossible problem.

The words remained inside.

Yet attempting to open the scrolls could destroy the very writing scholars hoped to recover.

Then modern imaging, computer vision, and machine learning began changing the situation.

And on June 25, 2026, researchers announced an extraordinary milestone.

They had virtually unwrapped and read the surviving portion of an entire Herculaneum scroll without physically opening it.

The achievement did not require inventing ancient words.

It required finding ways to reveal traces of genuine writing still preserved within a damaged object.

Welcome to Expedition #152.

Today, we enter a library whose doors have been closed for almost two thousand years.

And this time, the books are refusing to open.

Professor Parallax had acquired an extraordinary object.

It lay inside a carefully padded case in the laboratory.

Black.

Irregular.

Rather unimpressive at first glance.

White Rabbit approached with his notebook.

"Professor, is that a piece of coal?"

"No."

"A burnt loaf?"

"No."

"A fossil?"

"Not exactly."

Hatta entered carrying a silver letter opener.

"Then I shall settle the matter. We shall simply open it."

Parallax moved quickly.

"Hatta! Put that away!"

Hatta froze.

"Good heavens. Is the book armed?"

"It is extremely fragile."

"Then why have you brought a dangerous book into the laboratory?"

"The danger is not to us," the Professor explained. "It is to the book."

White Rabbit examined the object.

"What is it?"

"A carbonized papyrus scroll."

"How old?"

"Nearly two thousand years."

The Rabbit stared.

"And has anyone read it?"

"Not this particular portion of its contents."

Hatta looked pleased.

"Then we're in luck. I happen to be an exceptionally fast reader."

"That will not help."

"Why not?"

"Because unrolling it could destroy it."

Hatta considered the problem.

"Professor, you have brought us a book we cannot open."

"Precisely."

"To conduct a reading experiment?"

"Yes."

"Marvelous. This is the most economical library I've ever visited. No one can borrow anything."

Professor Parallax switched on a large display.

An intricate three-dimensional image appeared.

Layer upon layer of tightly rolled material emerged from the darkness.

And upon the blackboard, the Professor wrote:

A PROFOUND PARADOX, MY PEEPS! WHAT IF THE ONLY WAY TO PRESERVE AN ANCIENT BOOK IS TO LEARN HOW TO READ IT WITHOUT OPENING IT?

And down the Rabbit Hole we went.

Our story begins in the Roman world.

Herculaneum stood near Pompeii, beneath the looming presence of Mount Vesuvius.

The surrounding region contained prosperous homes, commercial activity, public buildings, and sophisticated Roman culture.

One residence, now known as the Villa of the Papyri, contained an important collection of ancient texts.

The villa received its modern name because of the scrolls discovered there.

In AD 79, Vesuvius erupted violently.

The eruption killed many inhabitants of communities around the volcano and buried settlements under volcanic deposits.

The villa's papyrus scrolls were exposed to extraordinary heat.

They became carbonized.

Ordinarily, fire destroys paper-like materials.

Here, under unusual conditions, portions of the papyrus survived as fragile carbonized objects.

The very catastrophe that devastated the community helped preserve evidence of its intellectual life.

But preservation came at a terrible price.

The scrolls could no longer be handled like ordinary books.

They were brittle.

Compressed.

Distorted.

And extraordinarily vulnerable.

The words had survived. Access to them had not.

The Herculaneum papyri were rediscovered during excavations in the eighteenth century.

Hundreds of carbonized rolls emerged from the buried villa.

Imagine being among the first people to encounter them.

The objects looked like blackened cylinders.

Some resembled pieces of charcoal.

Yet closer examination revealed that they had once been written scrolls.

The discovery was astonishing.

A collection of ancient books had survived beneath the earth.

But there was a problem.

Scholars could not simply unwrap them.

Earlier efforts to open the fragile rolls sometimes caused serious damage.

Mechanical techniques were developed to separate layers gradually, with varying degrees of success.

Some texts were recovered.

Others remained sealed.

And some materials were irreparably damaged during attempts to examine them.

The dilemma was clear.

How do you read an object when ordinary reading may destroy it?

White Rabbit looked up from the historical photographs.

"Professor, why didn't the Romans use books with proper pages?"

"Scrolls were a common format for written works in the ancient Mediterranean world."

"Very inconvenient."

"Not to the people using them at the time."

Hatta nodded.

"Quite so. One should never judge an ancient civilization merely because it failed to anticipate modern library shelving."

The Rabbit closed his notebook.

"Then what changed?"

Parallax smiled.

"Technology gave us another way to look inside."

Imagine a tightly rolled sheet of paper.

Suppose someone has written a message upon it.

Then imagine rolling it up so tightly that you cannot separate the layers.

Could you read the message without opening the roll?

For an ordinary sheet of paper, the answer might be difficult enough.

For a carbonized papyrus scroll, the challenge is much greater.

Scientists began exploring advanced imaging techniques, particularly forms of X-ray computed tomography.

A CT scanner collects measurements from different angles.

A computer uses those measurements to reconstruct information about the internal structure of an object.

In medicine, CT scanning helps clinicians examine structures inside the body.

In archaeology and materials science, similar principles can reveal internal features of objects that researchers do not wish to cut apart.

For the Herculaneum scrolls, high-resolution X-ray imaging can reveal the extraordinary complexity of the tightly wound papyrus layers.

But there is an important distinction.

Seeing the physical layers of a scroll does not automatically mean seeing the writing upon them.

And that is where the real challenge begins.

Professor Parallax displayed two images.

The first showed an ordinary handwritten page.

Dark ink appeared against a lighter background.

The letters were easy to recognize.

The second showed a section of carbonized papyrus.

Everything seemed dark.

White Rabbit squinted.

"Where are the words?"

"Inside the material."

"But I can't distinguish them."

"Neither can a simple visual inspection."

Hatta leaned forward.

"Could we illuminate it?"

"Ordinary lighting cannot reveal what remains sealed within the roll."

"Magnify it?"

"Magnification alone does not solve the problem."

"Ask the author?"

"The author has been dead for many centuries."

Hatta sighed.

"Ancient writers are remarkably unhelpful about revisions."

The challenge comes partly from the materials themselves.

The papyrus has been carbonized.

And the ancient ink often contains carbon.

That means the ink and the material beneath it can be extremely difficult to distinguish in the X-ray data.

Researchers cannot assume every dark patch represents writing.

Natural structures within papyrus may resemble markings.

Damage can create misleading patterns.

And scanning artifacts can complicate interpretation.

Some useful evidence can occur in subtle surface textures or other measurable features associated with the ink.

Machine-learning models may help identify those patterns.

But identifying genuine writing requires substantial care.

A computer-generated image of letters is not proof that those letters were present in the original scroll.

The marks must be supported by the actual imaging data.

Now imagine a scroll as an enormous collection of tightly compressed surfaces.

Its layers are not necessarily smooth.

They may be bent.

Crushed.

Torn.

Folded.

Or stuck together.

Virtual unwrapping attempts to reconstruct those layers digitally.

The physical object remains closed.

But a computer model traces the surfaces within the scanned volume.

Researchers then produce flattened representations of selected regions.

The process is conceptually similar to finding a crumpled page inside a three-dimensional structure and reconstructing what its surface would look like if laid flat.

But unlike an ordinary page, the papyrus may be severely distorted.

The software must identify which surface belongs to which layer.

It must avoid joining unrelated pieces.

And it must preserve the spatial relationships needed to interpret the writing.

Suppose two layers lie extremely close together.

A computer might mistake them for one continuous surface.

Or it might follow a crack into the wrong layer.

Flattening such an incorrectly traced region could produce a distorted image.

The result might look plausible.

But it would not accurately represent the original document.

That makes verification essential.

The computer must help reconstruct the surface.

Then imaging and ink-detection methods may reveal writing along it.

Finally, specialists in ancient languages must interpret the result.

Virtual unwrapping is not a single magic button. It is a coordinated process involving imaging, geometry, computation, and scholarship.

In March 2023, an ambitious international project called the Vesuvius Challenge began.

Its founders included Nat Friedman, Daniel Gross, and computer scientist Brent Seales.

The project encouraged researchers and independent contributors to develop methods for reading the sealed scrolls.

High-resolution scan data became available to participants.

Prizes were offered for meaningful advances.

But the effort was more than a competition.

It became a collaborative scientific undertaking.

Different participants contributed different capabilities.

Computer vision.

Machine learning.

Image processing.

Three-dimensional reconstruction.

Software engineering.

And expertise in ancient texts.

The question was simple to state.

Could modern computational techniques reveal writing that had remained inaccessible for nearly two millennia?

The answer began emerging faster than many observers expected.

In 2023, a remarkable breakthrough occurred.

Research participant Luke Farritor identified a complete ancient Greek word within the scanned data of a still-rolled scroll.

The word was associated with the color purple.

Imagine the significance.

The characters had been hidden within a carbonized object for nearly two thousand years.

No one had physically opened the scroll.

Yet computational analysis revealed traces of actual writing.

Other participants also made important discoveries.

Eventually, a team consisting of Youssef Nader, Luke Farritor, and Julian Schilliger won the 2023 Vesuvius Challenge Grand Prize.

The award was announced in February 2024.

Their submission helped recover more than 2,000 Greek characters across portions of a scroll.

The recovered passages explored philosophical questions involving pleasure and human experience.

It was an astonishing achievement.

But an important qualification remained.

The team had not yet read the entire collection.

Nor even the entire first scroll.

The achievement represented a relatively small portion of the available material.

Still, it established something enormously important.

Previously inaccessible writing could be recovered without physically unrolling the ancient papyrus.

The door had opened.

Even though the scroll had not.

Here is one of the most delightful surprises of the entire investigation.

The first substantial recovered passages were not instructions for constructing Roman war machines.

They were not maps to legendary treasure.

They were not prophecies.

Instead, they concerned philosophical questions about pleasure.

One passage examined whether scarce things necessarily provide more pleasure than things available in abundance.

It is a recognizable human question.

Are we happier because something is rare?

Or because it is genuinely valuable to us?

Does having more of something increase enjoyment?

Or does our relationship to it matter more?

These are questions people still discuss.

Professor Parallax presented the findings to his companions.

White Rabbit looked intrigued.

"An ancient philosopher was discussing pleasure?"

"Yes."

"Nearly two thousand years ago?"

"Indeed."

Hatta smiled.

"Then I should like to know whether he had discovered the finest possible afternoon tea."

"The surviving passage does not establish that."

"Naturally. The really important pages are always missing."

Parallax shook his head.

But the conversation touched upon something profound.

We often imagine the ancient world as impossibly distant from our own concerns.

Yet its surviving literature reveals people discussing many familiar questions.

Happiness.

Relationships.

Knowledge.

Mortality.

Ethics.

Beauty.

And the meaning of a good life.

The technology may be new.

The human questions are ancient.

Now we arrive at the development that makes this expedition especially timely.

On June 25, 2026, the Vesuvius Challenge announced that researchers had successfully virtually unwrapped and read the surviving portion of PHerc. 1667, known to the project as Scroll 4.

This was an extraordinary milestone.

The team had advanced beyond recovering isolated letters, words, or selected columns.

It had reconstructed and examined the surviving scroll from beginning to end.

The original physical scroll remained closed.

Researchers used high-resolution imaging, digital reconstruction, and computational analysis to access the text.

The result represented a major advance for the preservation and study of ancient literature.

The scroll contains a Greek ethical treatise.

Researchers have described it as likely belonging to a Stoic philosophical tradition, although scholarly interpretation continues.

The work demonstrates that material surviving inside the sealed collection may substantially expand what historians know about ancient intellectual life.

The 2026 breakthrough also included progress identifying texts and titles in other scrolls.

For example, work associated with Philodemus, an Epicurean philosopher, has revealed material connected with writings on vice and the gods.

But the project's work remains ongoing.

Reading the surviving portion of one scroll is not the same as translating and interpreting every scroll in the ancient library.

And that distinction matters.

The milestone is remarkable without requiring exaggeration.

White Rabbit had been studying the computer.

"Professor, did AI translate the scroll?"

"Not by itself."

"Did it recognize the letters?"

"Machine-learning systems helped identify patterns associated with ink."

"Did it unwrap the papyrus?"

"Computational methods helped reconstruct the internal surfaces."

"Then who read the words?"

"Scholars trained in ancient Greek and papyrology, working with the computational results."

The Rabbit nodded.

"So the machine did not simply write a new ancient book?"

"Certainly not."

That distinction is fundamental.

The purpose of the research is to recover evidence of text that genuinely exists within the ancient artifact.

Different computational methods contribute to different stages.

Image-processing algorithms help prepare the scan data.

Geometric reconstruction tools help trace papyrus surfaces.

Machine-learning systems help identify subtle features associated with ink.

Scholars examine the results and develop readings of the text.

And those readings must be tested against the available evidence.

A language model capable of generating plausible ancient Greek prose would not solve the archaeological problem.

It might produce beautiful sentences.

But beauty would not establish authenticity.

The achievement is not that AI can invent what an ancient philosopher might have written. It is that computation can help reveal what someone actually wrote.

Suppose a damaged fragment contains only part of a sentence.

A scholar may have several possible reconstructions.

One word seems grammatically appropriate.

Another fits the historical context.

A third resembles terminology used elsewhere by the same author.

Which is correct?

The answer depends upon evidence.

The visible letters.

The surviving traces of ink.

The grammar.

The physical arrangement of the papyrus.

And comparisons with related ancient texts.

A plausible reconstruction may still be uncertain.

A convincing AI-generated phrase may be entirely wrong.

This problem resembles challenges encountered when restoring historical recordings.

In both cases, technology may help recover information concealed by damage or interference.

But there is a difference between revealing evidence and generating a substitute for missing evidence.

Professor Parallax wrote three statements on the blackboard.

VISIBLE IN THE SCAN.

SUPPORTED BY SCHOLARLY RECONSTRUCTION.

SPECULATIVE.

"These are not interchangeable," he explained.

Hatta examined the writing.

"Then a sentence must not be accepted merely because it sounds philosophically impressive?"

"Correct."

"That's going to cause considerable trouble for philosophers."

White Rabbit smiled.

"Especially you."

For centuries, classical scholars have studied works surviving through repeated copying, translation, quotation, and preservation.

But only a fraction of the literature produced in ancient Greece and Rome survives.

Countless writings have disappeared.

Some were never copied widely.

Some were destroyed.

Others decayed.

And many were simply lost over time.

The Herculaneum library offers a rare possibility.

It may contain texts unavailable through any other surviving manuscript tradition.

Imagine discovering new passages from ancient philosophical debates.

Or recovering material that sheds light upon authors whose works are known only through fragments.

Perhaps newly recovered texts will refine interpretations of familiar thinkers.

Perhaps they will introduce previously unknown arguments.

Perhaps they will complicate established scholarly assumptions.

Not every recovered document will revolutionize history.

Some may prove repetitive or relatively minor.

But even ordinary writings can provide valuable evidence about intellectual culture, language, and everyday concerns.

Recovering ancient texts does not simply add books to a shelf. It can expand the historical record available to humanity.

The Vesuvius Challenge has already demonstrated that virtual unwrapping can succeed.

But major obstacles remain.

Different scrolls may vary in their physical condition.

Some may contain severe compression.

Other regions may be cracked or distorted.

Ink-detection methods that work well on one scroll may perform less reliably on another.

And the process of identifying internal papyrus layers remains computationally demanding.

The project's July 2026 technical updates described the need to improve automation, reliability, and the transfer of techniques across different objects.

The goal is not simply to read one spectacular artifact.

It is to develop methods that can work across a much larger collection.

Future research may improve the ability to trace complex internal surfaces.

Identify genuine ink.

Recover additional passages.

And make results easier for scholars to study.

But we should not assume that every surviving scroll will become perfectly readable.

Some regions may be too damaged.

Some writing may be too faint.

And some physical information may have been lost long before modern researchers encountered the objects.

The promise is extraordinary.

The limitations remain real.

Why is the non-invasive nature of this work so important?

Because an ancient manuscript is more than the words it contains.

It is also a physical artifact.

Its fibers.

Its shape.

Its writing materials.

Its damage.

And its archaeological context all preserve information.

A destructive examination may reveal one detail while eliminating evidence needed for future research.

Non-invasive imaging seeks to reduce that risk.

The scroll remains available for later study.

New instruments may produce better scans.

Future researchers may develop improved algorithms.

And scholars may revisit earlier interpretations.

A digital reconstruction is not a replacement for the physical object.

It is a representation derived from measurements.

Its reliability depends upon the imaging process and the methods used to interpret the data.

Good preservation therefore requires keeping the artifact safe and documenting the computational transformations performed upon its scans.

This is particularly important when software generates flattened surfaces or predicted ink images.

A result should be traceable to the source evidence.

Preserve the artifact. Preserve the measurements. Document the reconstruction.

The past deserves careful handling.

Professor Parallax returned to his familiar blackboard.

He wrote:

DEMONSTRATED. DEVELOPING. NOT ESTABLISHED.

Under DEMONSTRATED, he listed:

Under DEVELOPING:

Under NOT ESTABLISHED:

White Rabbit finished reading.

"Professor, we've learned to read part of an ancient library without opening its books?"

"Yes."

"And one scroll has been virtually read from beginning to end?"

"The surviving portion, yes."

"But a great deal remains?"

"An enormous amount."

Hatta nodded approvingly.

"Excellent. A library with unlimited opportunities for overdue discoveries."

By late afternoon, Professor Parallax had completed his demonstration.

White Rabbit was organizing the research notes.

Hatta had become unusually quiet.

The Professor looked toward him.

"Hatta, have you learned something?"

"Indeed."

"What?"

"That a sufficiently advanced computer can read books without opening them."

"Under very particular circumstances."

"Splendid."

"Why do you ask?"

"Because I have developed a new service."

Hatta unveiled a small sign.

HATTA'S AUTOMATED LIBRARY

No Reading Required.

White Rabbit stared.

"No reading?"

"Quite right."

"Then what does the service do?"

"It tells you what you would have learned had you read the book."

Parallax looked concerned.

"How?"

"By examining the cover."

"That is not a reliable way to determine a book's contents."

"It is extraordinarily efficient."

"Have you tested it?"

"Certainly."

Hatta placed a large volume on the desk.

Its title read:

AN INTRODUCTION TO THE PHILOSOPHY OF KNOWLEDGE

He activated his machine.

A small bell rang.

The screen displayed:

RESULT: THE AUTHOR KNOWS EVERYTHING.

White Rabbit shook his head.

"That's not what the book says."

Hatta appeared disappointed.

"Then the author has written a very misleading title."

Professor Parallax folded his arms.

"Your machine is guessing."

"An informed guess."

"Based on almost no evidence."

"An extraordinarily rapid informed guess."

The Rabbit opened the book.

Inside was an entire chapter about the limitations of human knowledge.

He showed Hatta.

Hatta studied the page.

Then quietly modified his sign.

HATTA'S AUTOMATED LIBRARY

No Reading Required.

Accuracy Not Included.

Parallax sighed.

"At least you've improved the disclosure."

Hatta brightened.

"And I have reduced our operating costs by ninety-nine percent!"

The Rabbit looked at the scroll display.

"Professor?"

"Yes?"

"Could Hatta's machine ever become useful?"

"Only if it learns to distinguish what the evidence actually contains from what it merely imagines."

Hatta nodded thoughtfully.

"Then I shall require a new sign."

He wrote:

HATTA'S LIBRARY OF THINGS WE KNOW, THINGS WE THINK WE KNOW, AND THINGS WE OUGHT TO CHECK.

For once, Professor Parallax smiled.

"Now that," he said, "would be a useful institution."

Evening settled over the laboratory.

The displays dimmed.

The ancient scroll remained inside its protective case.

Professor Parallax brought up one final image.

Across the screen, tiny dark markings appeared upon a flattened representation of papyrus.

Greek letters.

Words.

Fragments of thought.

Evidence of another human mind.

White Rabbit approached the display.

"Professor, do we know who read this scroll before the eruption?"

"Not necessarily."

"Was its owner wealthy?"

"The villa was a luxurious Roman residence."

"Did the writer expect the words to survive this long?"

"We cannot know."

Hatta studied the letters.

"Do you suppose the author would be pleased?"

"About what?"

"That we're finally reading the work."

Parallax considered.

"Perhaps. But we should be careful not to invent feelings for someone who cannot tell us."

"Then what can we say?"

The Professor looked toward the ancient text.

"We can say that someone wrote these words."

"Yes."

"That they survived an extraordinary catastrophe."

"Yes."

"And that generations of people worked to develop methods capable of revealing them."

Hatta grew thoughtful.

"Then perhaps we should simply listen to what the writer actually said."

"Or read it," the Rabbit corrected.

"Naturally."

For several moments, nobody spoke.

On the screen lay words that had waited nearly two thousand years to be seen again.

Not a modern imitation.

Not an invented quotation.

Not the computer's idea of what an ancient philosopher should have written.

The surviving traces of an actual human voice.

And three curious explorers stood before them.

Listening with their eyes.

For centuries, the Herculaneum scrolls represented an almost impossible contradiction.

The written material survived.

But opening the objects risked destroying the evidence.

Technology eventually provided a new approach.

By combining sophisticated imaging, computational reconstruction, machine learning, and scholarly expertise, researchers began revealing words hidden inside the carbonized rolls.

The progress was incremental.

A surface.

A mark.

A letter.

A word.

A passage.

Then substantial sections of writing.

And finally, in June 2026, the surviving portion of an entire scroll.

These advances demonstrate something important about artificial intelligence.

Its most valuable contribution is not always producing new material.

Sometimes it helps us recover access to information that already exists.

The text in the ancient scrolls was not written by a machine.

The authors lived in the ancient world.

They had their own questions, beliefs, relationships, and intellectual disputes.

The computer did not replace them.

It helped researchers encounter their words.

And that distinction leads us to our deepest question.

We often ask whether artificial intelligence can help humanity create the future.

But what if it can also help us recover parts of the past that we thought were gone forever?

Lost literature.

Damaged historical records.

Unread manuscripts.

Objects whose significance remains hidden.

And perhaps evidence that changes our understanding of earlier civilizations.

So the question at the bottom of today's Rabbit Hole is not merely:

It is:

Somewhere in Naples, a fragile ancient scroll remains physically closed.

Its writing has survived volcanic destruction, burial, excavation, and centuries of human curiosity.

For generations, scholars could examine the object but not freely read its contents.

Now, through an extraordinary alliance of archaeology and computation, portions of that ancient intellectual world are emerging again.

And the scroll has not moved.

The letters have not changed.

What changed was humanity's ability to find them.

🐰🕳️♾️

If artificial intelligence could recover one lost work from the ancient world, what would you most want to discover?

A lost play by a Greek dramatist?

An unknown philosophical work?

A forgotten scientific text?

An account of daily life written by an ordinary person?

Or a musical composition whose original sound has vanished?

And another question:

Would you prefer a fascinating but speculative reconstruction of a lost book, or a smaller fragment whose words could be established with confidence from surviving evidence?

Tell us what you would choose.

Hatta has already submitted a request for the recovery of his ancient family memoirs.

Unfortunately, the archaeologists have found no evidence that they were ever written.

He considers this a minor administrative oversight.

1. Vesuvius Challenge (June 25, 2026). “We Read an Entire Scroll Without Ever Opening It.”

Official announcement of the virtual unwrapping and reading of the surviving portion of PHerc. 1667.

https://scrollprize.org/

2. University of Kentucky (June 25, 2026). “The Day the Herculaneum Scrolls Began Speaking Again.”

Institutional reporting on the 2026 breakthrough and the scientific collaboration behind it.

https://research.uky.edu/news/day-herculaneum-scrolls-began-speaking-again

3. Vesuvius Challenge (February 5, 2024). “2023 Grand Prize Awarded: We Can Read the First Scroll!”

The historical breakthrough recovering more than 2,000 characters of ancient Greek text and the announcement of the winning team.

https://scrollprize.org/grandprize

4. National Endowment for the Humanities. “Students Decipher 2,000-Year-Old Herculaneum Scrolls.”

Background on the early challenge winners, virtual unwrapping, and the recovery of Greek philosophical writing.

https://www.neh.gov/news/students-decipher-2000-year-old-herculaneum-scrolls

5. Vesuvius Challenge (July 10, 2026). “Open Problems: Why Reading Every Herculaneum Scroll Is Still a Challenge.”

Technical explanation of the remaining difficulties in surface reconstruction, ink detection, automation, and scaling.

https://scrollprize.org/from_ash_to_text

6. Vesuvius Challenge Data Browser.

Explore scanned artifacts, virtually unwrapped surfaces, recovered writings, and publicly available research material.

https://scrollprize.org/data_browser

7. Diamond Light Source (June 25, 2026). “New Secrets Revealed from the Herculaneum Scrolls.”

Explanation of the contribution of advanced X-ray imaging facilities to the recovery of ancient texts.

https://www.diamond.ac.uk/

Research note: The June 2026 milestone concerns the surviving portion of PHerc. 1667, not the complete recovery of every Herculaneum manuscript. Machine learning, image reconstruction, and classical scholarship all contribute to the work. The laboratory scenes and dialogue involving Professor Parallax, White Rabbit, and Hatta are fictional educational dramatizations.

AI RABBIT HOLES | EXPEDITION #152

Where curiosity opens the unreadable, evidence brings ancient voices into view, and Hatta's library membership remains provisional.

#AIRabbitHoles #ArtificialIntelligence #VesuviusChallenge #Herculaneum #AncientRome #AncientGreece #Archaeology #VirtualUnwrapping #MachineLearning #ComputerVision #AncientManuscripts #DigitalHumanities #LostHistory #ProfessorParallax #WhiteRabbit #Hatta #Wondermage

📜White Rabbit 🐰 • Hatta 🎩 • Professor Parallax 🔭 • Q~ ♾️👈

Sometimes the strangest language is another way of seeing the world. 🌎👀

💡❔ Where curiosity goes slightly sideways, then comes back carrying a lantern. 🏮

🐰🕳️ Follow the White 🐰 Rabbit: AIRabbitHoles.com

🟨 Walk the Road: YellowBrickRoadtoAI.com

© 2026 AIRabbitHoles.com 🐰🕳️🕳️🕳️🕳️🕳️