Dr. Brent Seales, computer scientist at the University of Kentucky and co-founder of the Vesuvius Challenge, joins guest host Nick Tawil to explain how synchrotron X-rays and machine learning are making unopenable ancient scrolls readable, and why the work has now shifted from engineering to reading.
This summer in Naples, Brent’s team presented results from three sealed Herculaneum scrolls, including the first read end to end. Their announcement was covered everywhere from National Geographic to NBC News to The Guardian.
Brent and his work are the subject of The Vesuvius Challenge, an upcoming documentary from Infinite Films and Scott Free, in partnership with Calliope Pictures, narrated by Guy Pearce.
We’ve shared some highlights below, together with links & a full transcript. If you like what you hear/read, please leave a comment or drop us a review on your provider of choice.
Links
Show Notes
Brent Seales & Nick Tawil
Dr. Brent Seales: Stanley and Karen Pigman Chair of Heritage Science at the University of Kentucky and co-founder of the Vesuvius Challenge. Brent Seales at the University of Kentucky
Vesuvius Challenge: The open competition Brent co-founded with Nat Friedman and Daniel Gross to read the carbonized Herculaneum scrolls. Vesuvius Challenge
The June 2026 breakthrough: The announcement in Naples of new texts, titles, and authors recovered from the scrolls. The day the Herculaneum scrolls began speaking again
Nick Tawil: Guest host and head of Infinite Films. Connect with Nick
The Scrolls & the Discoveries
The Herculaneum library: Roughly 1,800 scrolls carbonized by Mount Vesuvius in 79 AD, the only library from antiquity that survived. Vesuvius Challenge: the scrolls
Philodemus, On Gods, Book 8: A title and author read from a sealed scroll, establishing that the work ran to at least eight books. CNN on the discovery
The En-Gedi scroll: The burned Torah scroll Brent’s team read without opening, containing one of the oldest known copies of the first two chapters of Leviticus in the canonical Masoretic text. Vesuvius Challenge on En-Gedi
The $1 million prize: The open competition to fully unroll and read one of the remaining sealed scrolls. Current prizes
Films & Coverage Mentioned
The Vesuvius Challenge: The upcoming documentary from Infinite Films and Scott Free, in partnership with Calliope Pictures, narrated by Guy Pearce. Deadline announcement
Secrets of the Dead: The Herculaneum Scrolls: The PBS episode following Brent inside the Villa of the Papyri. Watch on PBS
Cost of Glory: The Roman history podcast Brent and Nick reference while discussing what might still be buried. Cost of Glory
Dead Poets Society: The source of “carpe diem,” one of the two ideas Brent would give the world. Dead Poets Society
Highlights
Why Brent Launched the Vesuvius Challenge
The acceleration that we were going to get from just going open source was going to far exceed... it was going to far outpace our own ability to just get there. And what everybody wanted was to just solve the problem. It wasn’t so much about owning it or getting sole credit. It was really about delivering these things back to humanity. And we thought that we were right on the edge of being able to do that. And so I think the excitement of being able to get over that cliff, that marker of being able to say, “We’re going to do this,” that put us over the line we needed to be over to be able to say, “Okay, it’s worth opening this up.”
The Authors Are Coming Alive
So I feel like my part... not that it’s completely done, but it’s substantially done. We achieved this moment. And now the people who can read the language are not talking about the technology anymore. They’re talking about reading language, understanding the world. So I think there’s been a kind of change where I’ve completed the technical work, substantially complete, let’s say, where I don’t have to make the case anymore that this is possible. And of course, we’re going to refine the methods so that everything can be faster, everything can be better. But the conversation now is really about the works themselves, and the scholars who can read those works, and what we’re going to learn from the ancient world. You know, until the tech worked, we weren’t learning anything about the ancient world. We were learning about synchrotrons and virtual unwrapping. Now we’re actually reading the words, and the authors are coming alive. And that’s an amazing and a remarkable moment.
Edited Transcript
Nick Tawil
Hey, Brent. Good to have you here today. It’s so good to see you in person again. You’ve been incredibly busy and productive this summer. You are in the news literally all around the world, certainly here in the United States, but also in Italy, Germany, England, France, for the work you’re doing at Herculaneum in Naples, Italy. We had a big June announcement.
I’m privileged to know a lot about this because, you know, O’Shaughnessy Ventures, we’re literally doing the film on you guys. It broke on Deadline a few weeks ago. But I wanted to start my time with you now with an excerpt from an article in Scientific American.
So here it is: “How do you read a book you can’t open? That’s precisely what Brent Seales, a professor at the University of Kentucky, has spent his career trying to figure out. And on Thursday, his life’s work reached a pinnacle. Seales, alongside a huge group of volunteers and scientists working as part of the Vesuvius Challenge, has helped develop the technology to see inside books and scrolls that we can’t open without destroying them.”
Now, use the word... “Your life’s work has reached a pinnacle.” So I don’t think that’s hyperbole. I also don’t think it’s hyperbolic to say that what you’re doing here, and what you’ve been able to do with AI and the X-ray tech that you’ve devoted your entire career for, is probably a watershed moment for understanding our own civilization and our own history, our own civilizational history.
So I guess just to start, how has the summer been? And how has your work changed even in the last couple of months, just in terms of attention you’ve gotten around this? And I want to talk about the actual discovery as well.
Brent Seales
Well, thanks for that introduction. What a great piece. It is astounding, and to hear it read back... you know, every time we think we’ve found a pinnacle, there seems to be a new place. And this summer has been fantastic. The June press event in Italy kind of brought it all back home because, you know, we’re standing there talking about reading these scrolls for the first time using virtual unwrapping, and we’re looking across the bay at the actual volcano. And the time scale is crazy, right? 2,000 years. So it was a very high moment.
And we’ve had a series of those over the last five years. In fact, in 2020, which was the anniversary of the Apollo moon landing, I proposed to my research team, “This would be the year.” And you know what happened in 2020. Soon after those words came out of my mouth, the entire world, you know, goes upside down.
So from there really to now has just been this constant climb, and to get to the top... I mean, I think you’re right that the word “pinnacle” kind of captures being pretty high on this mountaintop of having solved the problem, and turning from the technology really more toward the lost voices that we’re restoring by reading these manuscripts. It’s amazing.
Nick Tawil
So talk to me about where these manuscripts actually live in Herculaneum and how many there are. Give me a sense of scope and scale.
Brent Seales
Sure. I did not know about this collection as a student, and it wasn’t until we started developing this particular technology that I started searching around for likely candidates. And Herculaneum came to the top immediately because it’s iconic. I mean, it was formed... The library at Herculaneum was formed by the volcanic eruption of Mount Vesuvius in AD 79.
Most people are familiar with the story of Pompeii. Pompeii is maybe the birthplace of modern archaeology, and it’s also this fantastical ancient lost city that was uncovered by humanity. Well, the reason why it was lost is the explosion of Mount Vesuvius.
What’s unique about Pompeii is the architecture, the scale and scope of the city. What’s unique about Herculaneum, which is where the scrolls were found, is that it’s the only library from antiquity that survived. We have all the history in the medieval copies of written material that tells us there was a library at Alexandria. People were into books. Books were traveling around the ancient world. But we don’t have any of the actual libraries left, and this is the only one.
So you’re talking about 1,800 scrolls, maybe a couple thousand books. They were all discovered in the realm of 1750 to 1760, when the Herculaneum site was first excavated, and many of them were badly damaged at excavation. But because they were being excavated by the royal family, King Ferdinand and others who were reigning in that part of Italy at the time, they were carefully kept. Some of them were gifted to other kings as diplomatic gifts, but for the most part, the collection stayed in Naples, and from 1750 until now, we have what we have.
And the state of affairs is that the scrolls were badly carbonized, which means that the volcano spared no expense in basically turning them into what looks like charcoal. So you have the only ancient library that is in existence in the world from antiquity, and then you have this profound physical problem that you can’t really unroll or very easily read any of the material.
Nick Tawil
So over the years, the easier scrolls were read and have been read and researched, but there’s just an entire portion that’s unreadable. And so that has basically stayed untouched until now.
Brent Seales
Yeah, absolutely. The ones that could be opened were. Some of the ones were pretty readily able to be unrolled. The volcanic ash and mud that covered up the library varied in how it impacted the collection. So early on, you had some that were just immediately opened. You have a lot of them that were destroyed because as they opened them, they just scraped it away and they made sketches, and now all we have are the sketches. The worst of the worst are the ones that remain. And for some reason, I latched onto that as the problem.
And I don’t know... looking back, you say things like “your life’s work,” but when your life is playing and you’re doing that work, you’re never really sure what’s going to become your life’s work, right?
Nick Tawil
Yeah. Well, what were you doing before this, before Herculaneum got onto your radar as a challenge? What were you doing before that?
Brent Seales
Well, I’m a computer vision and imaging specialist, and I did my PhD work at the University of Wisconsin in the field of computer vision. Computer vision can be defined as making computers and cameras work together to be able to either imitate what humans can do visually or to solve problems in the world autonomously.
So you see computer vision now every day succeeding with facial recognition, autonomous cars, all of the remote sensing that happens. The early days when I started that work led me into applications like manufacturing, autonomous navigation, space exploration. So I was doing work toward those applications. We were doing welding inspection, for example, where you’d have a camera help you watch a weld being made, and then the control loop would help you improve the process of doing those welds. It doesn’t sound super exciting, and I decided that it wasn’t that exciting for me. So I moved through some of those applications.
The thing I was doing immediately before Herculaneum and a lot of the heritage work that I started doing was medicine. You might be familiar with laparoscopic surgery. This is where surgeons will never actually open the operative field. They’ll do an entire procedure through scopes that have been inserted into the body cavity. And it’s incredibly hard to do that work, so the visualization part of doing that work is super important. So I was working with surgeons to try to help them better visualize not only the operative field live as they were doing the surgery, but all of the preoperative stuff that they were doing. You might X-ray a patient in advance to see what you’re dealing with, and then the surgeon would, what, memorize that preoperative stuff so that they have an idea of what they’re going to face when they go in.
Well, all of those things actually fed into what I eventually started to apply to virtual unwrapping, because the X-ray, right, I hadn’t had a lot of experience with that. But it actually became one of the key things that we’ve used to see inside of these books without opening.
Nick Tawil
And not just any X-ray. It’s, what’s it called?
Brent Seales
Computed tomography.
Nick Tawil
And where are these machines at in the world?
Brent Seales
The medical version of computed tomography is everywhere. Every hospital has one. In fact, tomography won the Nobel Prize in Medicine in 1979. The two inventors.
Nick Tawil
What about the X-ray machines you’re using for this project? You’ve had to travel to use these particle accelerators, haven’t you?
Brent Seales
Yeah, the X-ray for Herculaneum... We’ve settled on a synchrotron, and... that’s a very, very unique X-ray source that comes from a particle accelerator. And there are 40 or 50 of these worldwide. We’ve used a few of them in Europe and in America. There’s one at Oxford. It’s called the Diamond Light Source. We’ve also used one in Grenoble, France. It’s the European synchrotron light source. And what’s interesting about the synchrotron is that it produces a very carefully controlled, extremely bright X-ray beam. And brightness with X-ray means that you don’t have to wait as long to collect the images, because you have so many photons going through the sample that you can take a really, really quick snapshot because it’s so bright. So we can do an entire scroll at very high resolution very quickly.
Nick Tawil
What do you mean, quickly?
Brent Seales
Well, the kind of resolutions that we’re talking about now, which have really unlocked the problem, are kind of astounding. You know, the size of a red blood cell is bigger than the resolution that we’re getting to see the interior of the scroll. Every one of the little voxels that represents the scan of this...
Nick Tawil
A voxel?
Brent Seales
Voxel, which would be the 3D version of a pixel. You add that third dimension, you call it a voxel. The size of that little sample is astoundingly small, and that’s what lets us see everything that’s going on inside the scroll, including the ink.
Nick Tawil
Including the ink. And this is the key. Once you can lift the ink, the ink signature, I’ll say, off of the burnt scroll, then you’ve got something to work with. But it’s still tightly wrapped in this sort of cylindrical form. So now you need to... this is where the virtual unwrapping happens.
Brent Seales
That’s right. The X-ray gives you a model of everything that’s going on inside without having to open it, just like it does on the human body. But it’s still all wrapped up and compressed, and not in a form where you can actually do anything. So all those software steps that take you from that scan to the final readable thing, that’s the thing we invented to be able to do it.
Nick Tawil
And this is the Vesuvius Challenge.
Brent Seales
Well, before the Vesuvius Challenge, that was the research problem, which we essentially had solved to the point where we believed we had the formula. The Vesuvius Challenge took us over the finish line because we threw the contestants in there, broke through on the first-ever scroll, and then with the attention that we gathered and more funding, I mean, we just ran at the finish line after that.
Nick Tawil
And so this is co-founded between you and Nat Friedman. The Vesuvius Challenge is an open-sourced AI engineering challenge, so open source, like, presented and set to the public. Anyone could contribute to this project. And you weren’t really sure what you would be able to find when you opened it up like this. And how did your partners in academia, or other researchers and other lines of work, maybe the papyrologists, how do they all... Were they all involved in this at the very beginning? How did they react to this? Is this just helpful to them?
Brent Seales
Yeah. You know, the Vesuvius Challenge was not a no-brainer for me to jump into. It kind of evolved from an invitation that Nat Friedman sent me to just attend a conference into what became the contest. The conference was the chance where I could meet Nat and his colleagues. And in meeting him, I found out that it probably wasn’t going to work to just ask for investment, because there wasn’t that level of interest in the colleagues that he had gathered together for the meeting. But Nat himself was incredibly interested. And of course, he’s a really smart guy and very connected.
So he sat down after the conference, before I went back to the lab, and he pitched me this idea of a contest. And Nat’s an open-source guy. And what I knew that would require is for us to open up all of our research, including our data. So, yeah, I was sent packing back to the lab with this idea from Nat: “Hey, why don’t you just open up everything you’ve worked on for the last ten years, and we’ll do a big contest.” And you can see that maybe that isn’t a no-brainer, because you have to think, who’s been supporting me? What are they going to think about this? Can I just do this on my own? And maybe I should just do that, maybe just put everything on hold and do it on my own.
But we soon decided, as I talked it over with the folks in my lab, my trusted team, that the acceleration that we were going to get from just going open source was going to far exceed... it was going to far outpace our own ability to just get there. And what everybody wanted was to just solve the problem. It wasn’t so much about owning it or getting sole credit. It was really about delivering these things back to humanity. And we thought that we were right on the edge of being able to do that. And so I think the excitement of being able to get over that cliff, that marker of being able to say, “We’re going to do this,” that put us over the line we needed to be over to be able to say, “Okay, it’s worth opening this up.”
But, you know, I’ll be honest. Once in a while, we’ll see, even now in the press, an article that says a teenage college kid solved this problem. And while that might be true in a sense, it’s also monumentally not true, because we have this whole research program going, and then we had this entire Vesuvius Challenge happening. So all of that is the mechanism that has helped us be able to have that kid do that work.
Nick Tawil
Well, we were also talking about this before. You set up the challenge, and people that contribute to it in a meaningful way do get prize money for it. But what ended up really happening was it became a selection... it became an identity signaling, almost talent recruitment, function for you and your team, because the people that ended up contributing to it ended up becoming a part of the team, at least for a period of time. Correct?
Brent Seales
That’s right. And I love that we got the interest that we got, and I love that it became this huge recruiting tool. A much bigger funnel than I could have ever cast on my own. A combination of Nat’s network plus the notoriety of what happened with the Vesuvius Challenge and just the buzz. You know, we had two, three, five thousand people on a Discord channel talking every day. You can’t even... so many messages, you can’t keep up.
Nick Tawil
This is from around the world.
Brent Seales
From around the world. It’s global.
Nick Tawil
And so it’s still going today, the Vesuvius Challenge.
Brent Seales
That’s right.
Nick Tawil
And you run the team still. What’s the next challenge? Well, first, let’s maybe talk about what happened in Naples this past summer, because I think then the question becomes, what’s next? So tell me about the scrolls that were presented in Naples this past summer.
Brent Seales
Yeah. The moment that we signaled this summer in Naples was that we can do a complete scroll. So you give me a scroll, it’s never been opened, I can make a scan of it, and then the software process is now good enough that whatever’s inside, we can unroll and unwrap and present in a kind of complete way.
So we did that in Naples for three scrolls, and we were able to show also what we found in those scrolls. One is an extensive work of at least 140, maybe 150 columns of text. And I mean, that’s going to keep the papyrologists busy for a long time. This is all ancient Greek. It’s very dense.
Another scroll itself was partial, but everything that’s there we were able to unwrap and read, including the title and the author, which is kind of amazing. And a third scroll also gave us another title and an author, Philodemus, On Gods, Book 8, which means that there were at least eight books on gods by this guy, and no one knew that Philodemus had written those books.
So again, that can start more speculation about what’s in the books, what’s in Book 8 we’re going to read. So we unpacked that in June, and what I think it did is it shifted the conversation now from the technical, because we presented the complete technology that can do the whole thing, to the lost voices that are now going to be restored. So there may be 400 of these scrolls left that have never been read. They’re the hardest ones, and we’re reading them. And that is ecstatic.
Nick Tawil
You’ve gone from pursuing history to now actually reading history.
Brent Seales
That’s right.
Nick Tawil
So there’s a sense of completion now that you have around this project.
Brent Seales
Well, I’ve never been able to read any of the ancient languages because I haven’t had that training. So I feel like my part... not that it’s completely done, but it’s substantially done. We achieved this moment. And now the people who can read the language are not talking about the technology anymore. They’re talking about reading language, understanding the world. So I think there’s been a kind of change where I’ve completed the technical work, substantially complete, let’s say, where I don’t have to make the case anymore that this is possible. And of course, we’re going to refine the methods so that everything can be faster, everything can be better. But the conversation now is really about the works themselves, and the scholars who can read those works, and what we’re going to learn from the ancient world. You know, until the tech worked, we weren’t learning anything about the ancient world. We were learning about synchrotrons and virtual unwrapping. Now we’re actually reading the words, and the authors are coming alive. And that’s an amazing and a remarkable moment.
Nick Tawil
How do you refine the process that you basically pioneered here with your team? How long does it take today to read one scroll, and how long do you want it to take?
Brent Seales
I’d love for it to be an instantaneous thing, completely automated, that doesn’t take much longer than the scan time. We’re not even close to that right now. But imagine that you go to the scanner for your receipts from travel, and as long as it takes for it to scan is how long it takes for you to get the OCR and the thing to be uploaded. That happens right now for receipts and things like that. OCR, optical character recognition, was one of those first finally solved computer vision problems, and we would love for the scroll to be the same.
I don’t think that we’re that close to that right now. I think we’re still several years away from that kind of performance, because the technologies that we’re using are not that robust yet. But the computer vision and the machine learning that we’re using constantly improves. And we’re in this era now that is remarkable in that computer vision and machine learning, with more data, can get better. And so we actually see the ability to read the scrolls with the technology we’ve created as something that is going to create versions that get better over time as we read more scrolls. That’s never really been true with any other technology. Like, photography doesn’t get better the more photographs that you take. But because of the way the machine learning works, the more things that we capture, we learn from that, and then we get better at actually revealing the text.
So rather than time to text, which is one thing that we talk about, it’s quality of text, readability. We want to make the job of reading this ancient language the easiest possible. And overcoming the damage is really hard. I mean, this stuff is peeling, it’s fragile, so it’s kind of a miracle that it works at all. And then to say we want it to be pristine.
Nick Tawil
And instantaneous.
Brent Seales
And instantaneous.
Nick Tawil
So American of us.
Brent Seales
Yeah, well, it is. It’s always got to be better, right?
Nick Tawil
Oh, no. Actually, for the machine learning piece and the AI of it all that you and your team have been using to accelerate this process, in order to get it digitally unwrapped and handed over to the papyrologists, it comes across in the film in spades. But it also is something that other tech and VC investors have noticed, and even starting to make a little bit of waves in Hollywood, which is traditionally lukewarm towards AI. But in the film it comes across very, very cleanly, I think, and very authentically, that this is a great use case for AI, a positive use of AI, to further understand humanity. And I think it’s currently presented in the national conversation as either/or: you’re either for AI and against humanity, or for humanity and therefore you have to be against AI. And this is really a nice kind of marriage of the two in terms of a use case.
Brent Seales
Yeah, I see that. I think there’s irony that AI is so forward-thinking and premised on such modern and current technology, and yet we’re using it in this case to look backward. But it’s kind of a rescue mission, if not a sort of redemptive story, right? We’re kind of redeeming these lost works using AI as the tool to do it. And in fact, I like that message, right? That every new tool, including AI, that comes into human purview can be used as a way to redeem what’s around us. And so I think that actually is a really good story, and it’s one we should focus on.
Nick Tawil
So now that you’ve digitally unwrapped these three scrolls, is there a backlog of more scrolls, specifically from Herculaneum, or other labs or other researchers from around the world calling you up, saying, “Hey, you know what? I’ve got this parchment here that I can’t read. What do you think?”
Brent Seales
Well, it’s true that that is happening, and I’m glad that’s happening, because, congratulations, we have the technology to be able to look at some of that material. For example, there are scrolls in Amman, Jordan, from Petra, and they haven’t been opened. They’re carbonized. So there’s more material in the world that this could work on. But there is a backlog of just the Herculaneum stuff. I mean, we haven’t plowed through that. There are 40 scrolls that we’ve already scanned, and that’s 10% of what we think actually remains. So that’s a decade’s worth of work right there.
Nick Tawil
And the full villa hasn’t even been excavated.
Brent Seales
And that’s the next thing, right? The villa hasn’t been excavated. And I think people are a little bit naive about that. They think we have what we have and that’s it. But if you go down to the villa, which is not even open to the public... The villa is a part of Herculaneum that’s separate from the publicly available part of the town of Herculaneum. You go in there and you realize they really just sort of tunneled along the foundations of the villa, and that’s how they infer what the actual architecture of the villa was, and they just happened onto the artifacts that they have, including the scrolls, by doing that tunneling. It’s not fully excavated. And, you know, I’m not an archaeologist, but I would love to see, now that we know we can read the material, full archaeology happen there, not only to find the scrolls, but really to just better understand the villa and the scrolls we already have.
Nick Tawil
Well, it’s also, like you said, it’s the only ancient library that we know of. And what does that say about the authenticity of all these scrolls compared to maybe other artifacts that exist out in the world that academia or the research community currently has and uses?
Brent Seales
I love that you bring that up, because the provenance of scrolls from Herculaneum is unassailable. I mean, you can’t fake the carbonization. You can’t fake something being in this collection for 200 years.
Nick Tawil
So the carbonization actually creates an authentic fingerprint on this thing that cannot be exactly replicated.
Brent Seales
Exactly. Think about the Dead Sea Scrolls and how that all went down in the ‘40s. They were parted out by the Bedouin, so there were no archaeologists on site when they were found. And then you end up with advertisements in the New York Times for, like, a scroll from Israel. And what has happened since then is that we’ve seen a lot of fakes, where somebody will say, “I have a Dead Sea scroll, a scroll from cave number nine, Qumran, whatever.” And it’s really hard to prove sometimes that that’s true or not true. Not true with Herculaneum at all.
The other thing that’s amazing about a library from antiquity is that it’s not been copied by medieval scribes. So you have the original book that was actually on the shelf. So, fully authentic, verifiable, and the original book, not filtered through medieval scribes. And you know that a lot of the philosophy and the religion was tinkered with when the scribes of the medieval period recopied and preserved those works. And the tinkering is not always obvious. Sometimes scholars have to spend a lot of time to figure out, how did this change? They don’t have the original work, they just have the copy. Every book that we’ve read from Herculaneum has been a unique work. It doesn’t exist anywhere else in the canon, and it’s uniquely from antiquity. So it’s really a very, very compelling kind of collection.
Nick Tawil
So how long will it take you to do the next 40, 45 scrolls that you have in backlog?
Brent Seales
Right now we have a new competition that’s been announced, a $1 million prize through the Vesuvius Challenge, and that is for someone else, a competitor using our software, using their own, to do what we just did: read a full scroll from one of the 40. And we’re hoping that that’ll push our techniques forward. It’ll engage the community. Somebody will win $1 million for doing that. I think those 40 are going to be processed this year. That’s my hope. And then we’ll go after the remaining 360, the other 90% of what remains. A lot of those are fragmentary, so we know that we’re not going to have a whole book, we’re going to have a piece. But the game then becomes, if we can read all the writing, how do we put these works back together?
Nick Tawil
Why not grab a frontier lab here to just say, “Hey, we’ll just do this”?
Brent Seales
I’d love that. How do we do that?
Nick Tawil
We should have that conversation.
Brent Seales
We should do that.
Nick Tawil
And then let’s say these scrolls get scanned. How long does it take the papyrologists to analyze them?
Brent Seales
Papyrology is not fast, because what a papyrologist brings to bear when they look at the text is all of the other knowledge about the ancient world, because these texts are dense. Sometimes the texts aren’t fully...
Nick Tawil
You keep saying “dense.” What does that mean? A lot of text in a column, or the messaging is dense?
Brent Seales
Yeah, I mean, sort of intellectually dense. So you’re talking about parachuting into the middle of a book that is probably in the middle of a philosophical argument that a scholar is making, and they’re fully embedded in their time. So they’re talking about a person you don’t know whose name is maybe missing from the text, and about an idea that is only partial. And the game is to try to figure out, from the limited words, all of that stuff, all the context. And it can be really, really tricky and take a long time. Sometimes it ends up just being a mystery. It’s like, we don’t actually know what’s being talked about right here. We have all the words, but it’s really hard to know what the story is.
We’re hoping that most of the virtually unwrapped texts that we can provide are good enough in terms of readability and quality that the scholars will be able to unpack that density and figure it out. But the extra knowledge that the papyrologist has in their head, besides just the vocabulary and the structure of the language, is immense. And that’s what they’re applying, sometimes as a group and not just individually. Sometimes they have to work on it together.
Nick Tawil
So I want to talk more about the tech, and the tech throughout the entire workflow, to include what participants were doing at the Vesuvius Challenge and the challenges that they faced, either individually or as a group, in mapping out aspects of the scan, but then also the actual scan technology, the synchrotron you mentioned a few times. Give me a sense of scope for what a synchrotron is. Maybe start there, because that has to happen first, before you even engage people, your lab and other third-party participants through the Vesuvius Challenge, to even look at this stuff.
Brent Seales
Yeah, that’s right. We started by using a bench source. A bench source is one that you can move around. A bench source for X-ray would be one that you’d find in the doctor’s office or the dentist’s office. And so the first scans that we made, way back in 2009, the first time I ever scanned a scroll, it was done with a source that we took to the place where the scrolls were. It was just so much easier to have access if we were within the perimeter of where the scroll lived, at the library in the Institute of France in Paris. But what we discovered was that that source wasn’t going to give us some data that was good enough to do all the rest of this stuff.
So we quickly moved to the synchrotron, which is a source that isn’t movable. A synchrotron is an X-ray beam that comes from electrons that are running around in a ring at the speed of light. And the ring is like a kilometer across.
Nick Tawil
Just for this one scroll?
Brent Seales
Well, so the ring creates the electrons that run around in a circle, and then when you manipulate those electrons, it creates all of the different beams that you can use from this instrument. So it’s actually a huge campus around this ring, and one of those beamlines would be used to do the work on the scroll at a certain point. And you can tune what that beam is going to be in terms of the intensity of the beam and the energy that it has, all of the things that you think are needed in the characteristics to be able to elicit what you want from the scan. Way more controllability than something that you can take to the library.
Nick Tawil
So you have a lot of control.
Brent Seales
So what it created, though, was this problem of: now I have to take the material to this facility, and I have to guarantee the custodial holder, the people who own the scroll, the institution, that that’s going to be okay, that the scroll will be cared for, it won’t be damaged. So there was a time period where we had to shift to be able to use the synchrotron, but we hadn’t convinced anyone who actually had a Herculaneum scroll that we could do that. So you can imagine kind of going a little bit nuts, knowing that we could get better data, but we have to be able to convince the institution to let us do this thing. Once you get the scroll into the beam at the synchrotron, at that point, it’s really all about how do you capture the best possible scan, how do you optimize over a bunch of different parameters.
Nick Tawil
Which was a learning process.
Brent Seales
Which was a learning process. And the first time that we did it, we got what we got, and eventually we converged on a way of capturing the [unclear].
Nick Tawil
How much does this cost?
Brent Seales
The synchrotron cost, if you’re not being gifted the time, and typically we weren’t being gifted the time, would be $10,000 per shift, where a shift is eight hours. So, you know, a thousand, $1,500 an hour, and add inflation to that over the last two decades, and imagine how much.
Nick Tawil
And it’s the first time you’re doing it. You’re definitely going to have to learn off of that.
Brent Seales
Yeah, sure. And you’re not able to do the scan all in one go. So typically we do what’s called grid scans. You do a small section of the scroll, and then another adjacent small section, and then another adjacent small section, and you do these sections individually and then have to assemble the whole thing. So it’s its own thing to just capture the data and then manage the sizes and...
Nick Tawil
And not ruin the artifact.
Brent Seales
And not ruin the artifact. In a way, we were lucky, because these artifacts are completely inanimate. They’re completely desiccated, and they’re basically carbon, and you almost can’t do anything to carbon with X-ray. It just goes through.
But it’s also the worst case for solving the problem that we want to solve, because the X-ray doesn’t show the ink very well, or at all. There was a period of time, I don’t know if you know this, Nick, that people told me we would never be able to solve this problem, because it would be impossible to see...
Nick Tawil
Who told you that?
Brent Seales
Colleagues.
Nick Tawil
You don’t have to say their name.
Brent Seales
High level other professors, who felt like I was the guy tilting at the windmill, and they tried to sit me down and say, “You know that you’re never going to be able to see the ink using tomography. Carbon is invisible. You won’t be able to see the ink on top of this other surface.” And we refused to believe that, because we felt like there would be evidence. We just didn’t know how to tease it out.
And the two miraculous things that occurred that let us tease that out in a pretty non-obvious way: the synchrotron first. Super high resolution, so you can see the effects at a microscopic level.
Nick Tawil
How microscopic? Down to what?
Brent Seales
Well, a couple microns. A micron is one one-thousandth of a millimeter. So at one micron, you have a thousand slices in the space of one millimeter. That’s tiny.
Nick Tawil
So the two miracles that happened...
Brent Seales
The second miracle was AI, machine learning, because the evidence that is there is subtle. And so you need a tool that lets you amplify that evidence, embedded with so many other signals, because there’s a lot going on inside the scroll at that scale. There are fibers going everywhere, and there’s damage from the carbonization itself, and there are rips and tears and folds. And then somewhere, you know, the ink is actually on the surface of this ancient paper, which is not like normal paper. It’s not a composite that’s really uniform. It’s fibrous everywhere.
And so you have this really weak signal, and you have a pretty diminished modality, which is this X-ray modality. You would love to have visible light, but you can’t. And then along comes this incredibly sophisticated and powerful tool of machine learning that lets you find needles in haystacks, and boom.
Nick Tawil
And this is where actually the human comes in, too, because now you’re focusing AI engineers onto this problem, the Vesuvius Challenge. What are some of the obstacles that they had to overcome?
Brent Seales
Well, we had this huge labeling problem.
Nick Tawil
Labeling what?
Brent Seales
To break through... you couldn’t see the ink in the scans that we were getting with the naked eye, easily or at all. But when we trained the AI based on ink we knew was there... and we did that by using open fragments that had writing visible on the open layer. And then we would scan those with the same tomography, and then we knew where the ink was. We would label it “ink,” “no ink.” And then we could see that the machine learning would learn the evidence of the ink, even though we couldn’t see the ink with the naked eye in the scan. So all of that was scientific control, to be able to prove to ourselves there is evidence of the ink. The AI is learning it, the machine learning is learning it, but we have to label it for it to learn.
Well, how do you label the ink in a completely closed scroll? This was where the competitors gave us a huge boost, because now we have 2,000 people working on this problem, and they’re trying to create labels. And what they would do is they would run a fairly diminished machine learning effort on an open section of the scroll, virtually opened, and they’d see a response, some letters coming out of the mist, and then they would relabel. And iteratively, as more text gets exposed, they get more labels. Now amplify that across the 2,000 competitors...
Nick Tawil
24/7.
Brent Seales
With people sharing stuff, because we did inspire sharing through progress prizes. And so now you have the acceleration that we hoped we would get. The techniques weren’t substantially different than what we had already been doing, but the scale of what was happening, and iterations in terms of innovations were happening too, for sure. People were building tools that were helping. And so then eventually you’re solving this bootstrap problem by doing this kind of one-way ratchet, everything sort of getting better slowly.
And by the end of that year, I think it was 2023, maybe, we were headed toward the final submissions of the Grand Prize. Things were just accelerating like crazy. I mean, people sharing, and text obviously being detected from inside the scroll was very exciting.
Nick Tawil
And that’s when we got wind of it. And so at that point, you have something digitally unwrapped, then passed off to the papyrologists. And that’s a whole flow of work.
Brent Seales
Yeah. And we’re trying to improve that, actually, because it’s not really a one-way street at that point. Papyrologists just have a lot of insight into what looks legitimate and what might be an artifact. An artifact, for example, or even a mistake, might be what we call a sheet switch, where the virtual unwrapping actually gets off of the layer that it’s on and it descends to a lower layer. And the text, if you can read it, is going to tell you that that happened.
Nick Tawil
There’s, like, a break in the narrative.
Brent Seales
There’s a break in the narrative if you’re reading the words, or you can see sort of an obvious artifact that’s not going to be in a normal open scroll. And so we can actually go back and fix those things if they’ve happened. And usually the papyrologists are the ones who pick it up.
Nick Tawil
Oh, wow. So they’re actually involved pretty early on in the process. They’re not just waiting for you to hand them something.
Brent Seales
That’s right.
Nick Tawil
Cool. And then could you tell me a little bit about the scroll En-Gedi?
Brent Seales
Yeah. Before we were making such good progress with Herculaneum, this scroll from En-Gedi came to us through the Israel Antiquities Authority, and in particular the director at the IAA, who at the time was Pnina Shor. She’d heard of our work... we didn’t even know that this was happening, but she’d heard of our work and had found in their collection a scroll from En-Gedi that had been looked at by a number of professors, but no progress could be made because it was too badly damaged, and no one wanted to take on the formidable task of trying to open this little scroll physically, because it just looked like it was in pretty bad shape.
So she’d gotten that scroll back into the collection and then made a scan of it, and she used a bench source, not a synchrotron. And then she gave us the data and said, “Give it a try. We can’t do anything with this, and we can’t see any obvious writing, but I’ve read your work and maybe something can happen here.”
So it didn’t take us long, because this scroll was actually the perfect candidate for the moment in the development of our technology. It was animal skin and not papyrus, so in that way, it was less challenging. And then, as we did the virtual unwrapping, we found out that the ink... it must have been some kind of metallic ink, because the density of the ink in the X-ray was different from the density of the animal skin. So we weren’t looking at the same kind of really, really subtle effects of Herculaneum, where there’s no density difference at all. I mean, it showed up. The ink showed up in the En-Gedi scroll as a very bright signal sitting on top of this damaged animal skin, and boom, the text just appeared.
Nick Tawil
And what did it say?
Brent Seales
Well, that...
Nick Tawil
This text that had never been previously seen before, even by the owners of it.
Brent Seales
The scroll still remains wrapped up. It’s never been opened. But we could read everything inside because of these effects, using virtual unwrapping. And what we found is that it was a Torah scroll. It was the first remnants of the first two chapters of the Book of Leviticus from the Old Testament.
Nick Tawil
Wow. So how old was the scroll that you were given? How old was that expected to be, versus the text you found on the scroll? Is that the right question to ask? Like, “We don’t know what this is. We think it’s this old.” And then you find this. What does that change?
Brent Seales
The text definitely helped to date the scroll. They had carbon-dating evidence that put the scroll second, third century. But it had been found at an archaeological site that was Byzantine, so a couple hundred years later. So those results didn’t make perfect sense. There was disparity between the carbon dating and the archaeological evidence, which... they knew this site is Byzantine, so it’s probably fourth or fifth century. Once we read the text, it made more sense, because they could see the writing. So paleography, which is the analysis of the handwriting, dated it to an earlier era. So it was probably a scroll that they’d had in their collection at this synagogue for a hundred or a couple hundred years.
And as such, it was probably sacred. Something likely catastrophic occurred at that archaeological site, because normally none of the remnants would be in the Holy Ark. I think it was tradition that, the genizah, you would dispose of these artifacts, because they’re holy. And you wouldn’t just bury them or leave them partially burned for others to have. So this was the first time in a synagogue that they had ever discovered Torah. And it also became known as one of the oldest copies of the first two chapters of Leviticus.
And from the text, the scholars were able to tell that it was the canonical version, the Masoretic Text. And it’s a really early example of that settled text, so hadn’t changed. So, pushed the date for the settled Masoretic Text way back. They do have examples in the Dead Sea Scrolls collection itself of variations from the settled text, but this one was dead on the settled text of Leviticus. Pretty exciting.
Nick Tawil
Well, as we talk about this era, and we think about the papyri at Herculaneum, we know that the eruption happened in AD 79, but the scrolls in this villa could be much, much older. Because it was... I don’t know, you tell me, is it the sort of melting pot of a library that was collecting wisdom and scrolls from just around antiquity at the time? Give me a sense for what you would hope to find, or what a historian would hope to find and hope to see, in the remainder of the library.
Brent Seales
I think that the library probably was a place where they had an archive, so they had older texts. Typically, the way I understand it is that resident philosophers would write new works, but they would premise those new works on an archive of other works, often quoting them, extending the arguments in the earlier works, debating with each other, and then writing. So I think it’s very likely that there was an archive of much older works. And if that’s true and we discover those works, you could have literally anything.
It was a Roman villa, so it’s a little mysterious that there’s so much Greek and not as much Latin. So a lot of people have postulated that that’s just an archaeological feature, that it hasn’t been fully excavated. If we were to excavate it, because it’s a Roman villa, we would find much more of the Latin work. And I mean, that opens up all kinds of possibilities. We know of so many things that have been lost from the ancient world.
Nick Tawil
Let’s speculate. Let’s start speculating. I’m not going to hold you to anything, but I want to know.
Brent Seales
Livy’s History of Rome, for example. This would be for Alex Petkas, right? Because he knows everything that’s lost and found.
Nick Tawil
Cost of Glory great podcast.
Brent Seales
Yeah, exactly. Incredible. And then you have all of the literary and philosophical traditions of the Greek schools, so there could be much more of that. It’s thought that this particular philosopher, since we have Philodemus, was primarily an Epicurean. And there were also the Stoics. They argued with each other, and other schools.
If you ask me what I wish we could find, because the first century was such an amazing point in human history, I’d like to find early gospel material. Anything Christian at Herculaneum would instantly become the oldest witness that we have to early New Testament Christianity. I mean, even the fragmentary stuff from the New Testament is second century, for the most part. We have almost nothing from the first century, symbols only. But we know that this is Rome’s backyard. And we know the Apostle Paul was flitting around the world writing letters to the churches, which became part of the New Testament. So you have more than just the synoptic Gospels, which of course would be amazing. I mean, AD 79, right? This is just shortly after the life of Christ. So if there were writings that documented anything about his life that didn’t make it into the synoptic Gospels, those writings could potentially be in a library like this.
In, you know, the recent results corner of the library. I mean, most people who are academics and scholars would argue that it’s far-flung to think about anything that’s really gospel-y, because these libraries facilitated the ancient philosophies, and Christianity was new and hadn’t really taken hold. And I get that, nascent things at that time hadn’t made their way into the canon, the library, the archive.
But on the other hand, it’s fun to think about what could be there. The lost books of the History of Rome by Livy would give us a window into so many other events that would be better than just a philosophical tome. So rather than somebody talking about what it means to live a pleasurable life like an Epicurean, you could have somebody telling you about all of the events that happened as Rome rose and fell, and imagine how that would be. I think the historians would want history books, the philosophers want philosophy books, and the Christians want Christian books.
Nick Tawil
That’s great. Well, here’s to discovering new human knowledge. Thank you, Brent, for your time.
