Loading summary
Transcript113 lines
- [00:00]
Tara Palmieri
Take the exit, turn right into the drive thru.
- [00:03]
Luke Igle
Nope, I'm making dinner tonight.
- [00:05]
Tamsen Fadal
You don't have time.
- [00:06]
Tara Palmieri
Josh has practice.
- [00:07]
Luke Igle
Oh, that's right.
- [00:08]
Tara Palmieri
I'll just get a salad and fries.
- [00:10]
Luke Igle
No, just the salad.
- [00:11]
Tamsen Fadal
But salad cancels.
- [00:12]
Tara Palmieri
Fries.
- [00:13]
Luke Igle
Salad only. Fries. Salad, fries.
- [00:15]
Tara Palmieri
Food noise isn't fair, but Mochi Health is the affordable glp. One source that puts you on the
- [00:21]
Tamsen Fadal
road to successful weight loss.
- [00:22]
Tara Palmieri
Hey, can I get the fries?
- [00:23]
Luke Igle
Salad. Sorry.
- [00:25]
Tara Palmieri
Learn more@joinmochi.com Mochi members have access to licensed physicians and nutritionists. Results may vary.
- [00:30]
Luke Igle
The only way that a single person can have 19 terabytes of data is through images and video. It's text alone does not take up that much data.
- [00:39]
Tara Palmieri
It could be videos of his many friends to compromise them.
- [00:43]
Luke Igle
I had one journalist say it might just be pornography that he downloaded from the Internet. I've had other people say that it's illegal pornography that he downloaded.
- [00:51]
Tara Palmieri
Welcome to the Tara Palmieri Show. That's how many terabytes were seized from Jeffrey Epstein's property. And the record suggests it may be even more. One FBI investigator wrote in an internal email that authorities expected, quote, somewhere around 20 to 40 terabytes of data, noting that the total storage capacity of devices seized from Epstein's homes was 40 to 50 terabytes. Yet what the public has mostly received so far in just 300 gigabytes are PDFs, email chains, flight logs, court docs, photos. So where's the rest? Today on the show, I'm joined by Melissa Du and Luke Igle, 220 somethings living in San Francisco who decided they weren't going to wait for the government to organize the Epstein files. They were going to build their own search engine. I'm talking about JMail. If you're an Epstein obsessive like me, you're spending a lot of time on that site. It now hosts more than 3 million documents. It's processed hundreds of gigabytes of DOJ releases and has been visited by over 150 million users. Sounds like there are a lot of Epstein obsessives out there. But in our conversation, we go beyond the search tool, why they felt they needed it in the first place and what they're doing to keep it running. But the question that really hangs over all of this is how much data are we really looking at when so much of it is redacted and we know the government maybe holding on to terabytes? So what's being withheld and why? If Epstein had cameras in every room, and even little pin cameras and Kleenex boxes. And so many of the survivors said they felt like they were being recorded. And again, we've got FBI investigators saying they may have 20 to 40 terabytes. Are we really looking at transparency? Just a curated slice of it. Take a listen here. Lucas Igle and Melissa Dew, welcome to the Tower of Palmieri show. Thanks so much for coming on. I guess I just want to say I'm impressed with your work. I feel like it is now in my. What is it? Favorites or on my browser bar, jmail. And I'm sure it's in the browser bar of many other people and you just like find yourself searching just random names that come to mind through jmail to see if they are in the Epstein files.
- [03:14]
Luke Igle
Well, appreciate that a lot. I've definitely found myself doing that. Every time someone comes up in the news, open up chain, look them up.
- [03:21]
Tara Palmieri
Or just like someone I know, I'm like, wait, could they be in the Epstein files? And you know, it's funny, my. My cousin mentioned to me, she's like, do you know that you're in the Epstein files? And I was like, I am. And she was like, yeah. And so of course, the first thing I did was I went to JMail and I put my name in. I wasn't in JMail, but I was in the Justice Department. They had reference some of my reporting.
- [03:43]
Luke Igle
The, you know, I think whenever they have reporters and they're especially like saying bad things about Julie Brown or others that I've. I'm very proud of Julie Brown and the other reporters who, you know, they showed up in those files in the right way, right? Complaining about them.
- [03:58]
Tara Palmieri
Hey, they're paying attention. So I want to go over your background. You guys are Both in your 20s, you're living in San Francisco, and this Epstein story explodes over the summer. I doubt you ever thought that you were going to be knee deep in this kind of muck. How did this turn into building an infrastructure project?
- [04:20]
Luke Igle
Well, to introduce myself a little bit, I'm one of the co creators of jmel and I work a lot alongside Melissa, who joined us as we work to expand the project. I first, you know, became aware of this whole thing in my first summer right after my freshman year over at mic. It went from weird financiers in prison is put into jail to tons of professors who I looked up to. Turns out we're spending a whole lot of time with them way after that 2008 conviction that happened. And I had friends upon coming back in the fall of 2019 to campus being like grabbed by reporters because they made the mistake of walking outside of the MIT Media Lab, just being like, asked various questions about it. That was the beginning of me finding out about this whole thing. And then over the next five plus years, I just became more and more engrossed by specifically the MIT side of it. And then like everyone else, it expanding beyond that. I made a movie with a friend while I was in college. And of course, this whole saga was something that we covered extensively. And it wasn't until the Department of Justice, or specifically the House Oversight Committee put out that first batch of about 20,000 emails back in November that I came up with this original idea of jmail. Told my good friend Riley. He says, no, that's then after 12 hours, he says yes. And we have then each month built it out more and more as more and more files come out now at about 3 million pages.
- [05:54]
Tara Palmieri
I think it's great to hear that you've had a long history with the Epstein story. You know, a lot of people have become fascinated with it in recent years, mainly because they see it as a way to take down President Trump and political enemies. But you have obviously been really invested in it because you saw how the Epstein raw infected academia and your world in particular, and you stayed on it, which I admire.
- [06:17]
Luke Igle
Well, I appreciate that.
- [06:19]
Tara Palmieri
You know, I've also been covering it for a really long time, and I always felt like it was like, hey, guys, look, this is a really big deal and it would get picked up here and there, but for the most part, it was only seen as useful or advantageous when it could take one party down or the other.
- [06:35]
Luke Igle
Yeah, I think. And the useful and advantageous thing to do was to take out MIT and Harvard, I would say. And it felt like the. The conversation back in 2019 was almost too small in scope. It was, look at this really bad guy that these two institutions shouldn't have allowed into the door. Right. Which I think is true. And I find myself, like, really disappointed by the Media Lab specifically. But it feels like it has taken about five years for the scope of this kind of national conversation to expand to what it perhaps should have been the whole time, which is there's this global element to it. There are well over thousands of victims and currying favor with really important professors and embracing pop science and embracing the TED Talk as the form by which elites gather. It's all very embarrassing. It's all very kind of. It's very humiliating to a lot of people who write a lot of airport books. Right. But I appreciate what you're saying too, like it's, it's important. The, the full scope of it, I think is what's kind of shocking everyone.
- [07:35]
Tara Palmieri
Melissa, how did you fall into this? Was it a side hobby at first, a passion project? When did you realize it was something bigger?
- [07:43]
Melissa Du
I mean, I saw Luke and Riley's first drop of JMail, which was just the email component, back in November, and I remember clicking into it and as you know, they have like the contact sidebar and you see names like Joe Ito and Noam Chomsky. And I think I had a little bit of a Noam Chomsky phase in college, as most MIT kids do. I'm sure you read Manufacturing Discontent and you read some of his linguistics work. And I remember clicking through Noam Chomsky's emails with Epstein and just marveling at how deeply Epstein was in the academia circles. He would just email Latomsky and Joey about random questions on consciousness and science. And the way they interacted was very friendly almost. And I think seeing Epstein's emails in such a familiar interface really brought that element out of like here was Epstein just in his native element, talking to extremely well connected and affluent people. And I think since that initial introduction, I got more and more engrossed into his world because he was just so widely connected. I think Epstein's network gives you kind of a very lucid representation of how a lot of the world behind the worlds work. All of the most powerful people, they are just deeply connected in these very, in these very interconnected circles. And that kind of shows a lot of how power is transferred and how they can kind of orchestrate different things in the world. Like he could orchestrate Harvard through Larry Summers. He had deep connections with MIT Media Lab and was doing funding there. And I was really fascinated by that as part of the story. So when the second round of the files dropped in December, I kind of asked them how they were planning to handle the court documents. And Luke mentioned that he was creating J Drive, Riley was expanding into jphotos, and then a bunch of other people joined on and did JFlights and Jamazon and other kind of additions to the suites. Yeah, so that's kind of how I got involved.
- [10:13]
Tara Palmieri
And so there's only 10 mil, 10 of you, and you've had 3 million documents in the past dump. How do you even manage something like that?
- [10:24]
Luke Igle
Well, I think it's a story of Melissa taking it on, I would say. So there are a few things only because of technology that's been around for a year or two is why we've been able to build all of this software so quickly. And for 10 people with completely different ideas to build into one code base and have a product that can sustain what is now 150 million visitors. Specifically, the document processing was the hardest part of this most recent drop. Comparatively very tiny drop of just a couple gigabytes of documents on the deadline of December 19th happened per the Epstein Files Transparency Act. We hear whispers that million more are going to come out in January. I kind of believed it. And then of course, completely interrupting everyone's Friday, we all gather at my house once again, over 10 people again. And that's where we started dealing with the hundreds of gigabytes that got released. And that's where all the craziness began. We were relying on so many different journalists to get all of the zip files because downloading from the government website just kept on canceling and failing over and over again. We worked with our partners over at Reducto, our very good friends, who run a document processing startup that uses AI to perform basic classical optical character recognition on those millions of pages. And then figure out the really cool part, which is this portion of the doc is definitely the subject, this is definitely the body, this is the sender, and you can do that for anything. So that's what we use to do flights. This is the destination airport, this is the origin airport. And of course, all the data cleaning that we've had to go through and all the identification of mistakes, identifying which emails are between FBI agents, which ones are from Ghislaine Maxwell and not actually part of Epstein's inbox. So much of this was a story of being very close to this very new technology that has allowed us to deal with all this new data. And because of all the previous, those two previous build outs of jml, we had our partners over at Dropsite, we had the many journalists that they know constantly just giving us download links. But that was a really tough effort. Was January 30th. We're all shooting a documentary with a good filmmaker friend about jml. So the whole crew had to already gather at my house. Now we're also gathered to do this build out. It was crazy.
- [12:39]
Tara Palmieri
Yeah, you know, I was, I was thinking about it the other day and this couldn't have happened at a better time. As much as there's been so much frustration around the files and not having access to them, I do think that if they had come earlier, before AI had been at the place to be able to handle it, and before we really Had a full army of independent journalists out there, I don't know that we would have been ready for this, for the number of documents that we have right now for the tranche and been able to really organize it. And I think in a lot of ways the Justice Department would have gotten away with it. Away with, you know, perhaps more redactions that they should have been redactions or more oversight of the true scandals in the story. It was just been such an overwhelming deluge. But with AI and with all these independent journalists and with the great work that you're doing, we see a, A real narrative and a, and, and the, and the micro stories within this, because there are so many stories within this, within these files.
- [13:41]
Luke Igle
Fully agree. You know, there are a lot of, I would say, AI, skeptical journalists who I've looked up to for a long time now in, in New York, kind of more in the east coast, who have been subtweeting and just talking about JMail in a positive way, which has meant a lot to me. Where all of this technology that has been coming out the past few years, it's particularly good at processing large amounts of text. And I think a lot of this was demonstrating, instead of us spending a thousand hours to pick out every single flight from this massive document dump, here is 10,000 flights we were able to extract in 10 hours. Right. You know, I think Noam Chomsky himself wrote in Manufacturing Consent about how modern governments really like to give too much information and too much content to their, their own citizens in order to, I would say, do what past governments may have had to do through direct censorship. Right. I think this is a great bulwark against that.
- [14:40]
Tara Palmieri
That's like what the Russians do. They just overwhelm with content. Right? Overwhelm with information, make the people feel powerless through the disjointed information. Some disinformation, some truth, and just an inability to really be able to process it and digest it.
- [14:59]
Luke Igle
Yeah. I think there was a really interesting guy who used to work under Putin who would do exactly that. He would fund youth, left wing youth groups and Nazi youth groups all at the same time. He would, you know, he, he spent a lot of the 2000s and the 2010s doing exactly that. I, I think the Russia similarity is pretty becoming quite strong now.
- [15:20]
Tara Palmieri
Yeah. Melissa, there's a donation link on your website. You've raised about $32,000. I mean, is that covering this? It must cost a lot in servers. I mean, how many users have you. Do you have on jmail at this point?
- [15:35]
Melissa Du
I have to check on the exact number. Yeah, I think Luke has.
- [15:38]
Luke Igle
It's more like a hundred. It's more like 150. We have not done a good job making sure that that counter is correct. We're at about $150,000, $150,000 raised so far.
- [15:48]
Tara Palmieri
And how many users do you have on JMail?
- [15:51]
Luke Igle
We have 150 million visitors. And we zoomed in, that was about 50 million people who have very substantively used us for more than just a few minutes. And we have served billions of requests and about half a billion pages have been viewed.
- [16:10]
Tara Palmieri
Wow.
- [16:11]
Luke Igle
The most number of users using us all at the same time was in the tens of thousands. So that means at any given time, tens of thousands of people are there, which then translates to about over a million people. At the peak, we had 5 to 10 million people in one day visiting us. So that's how it's been spread out.
- [16:28]
Tara Palmieri
Do you have enough money for the servers and the power and the energy that you need for something like this?
- [16:35]
Luke Igle
Well, you know, it's the. There are a lot of tech communities that have been very closely watching the JMail story on, like, the tech side of things. We. This is completely my bad. I set us up on the most expensive way to serve this very static, norm, simple website. And as a result, it was me personally who then had to make sure that we could cover the initial fees of hosting this. Our first bill was about $50,000. The person who, the CEO of the company we chose to host at the very beginning offered to personally cover that bill. And we have now discovered a much more sustainable way where it's not going to be in the hundreds of dollars, it's not going to be per month, it's not going to be, let's say, $50,000 per month, every single month going forward. Unbelievable number of people have reached out to actually help us with optimizing this, making sure that it's good. There was a huge fixed initial cost of processing those documents, another fee that is well over six figures of just processing 3 million PDFs, running large language models over each PDF to make sure that every single detail is captured. It has been unbelievable working with Redacto, specifically the best PDF AI company out there, to cover that fee. And so there are all these different pieces of software that, from hosting to PDF extraction, to translating into other languages, that by living in San Francisco, we have gotten to know the people who can help us there. Another company that I want to shout out is called General Translation. They have translated some of our ui to like 10 different languages. And now we're going to start working with them on translating tons of those documents to Japanese, Spanish, Portuguese, whatever.
- [18:17]
Tara Palmieri
Now the whole world can know. Has a site ever crashed?
- [18:21]
Luke Igle
Oh, yeah. I think Melissa and I were in the trenches making sure that the database didn't get overloaded. And you know, Melissa did a lot of that heroic work on spreading out the work to what was it like 10 different replicas, five different databases, rather than one giant node that has all this load bearing infrastructure that everyone is relying upon. Lots of crashes. If you look us up on Reddit, you'll see is JMail down? JMail's down. So a lot of lessons when you have two different cloud services that are trying to cache things, there are all sorts of ways that it can crash, but we're pretty stable now.
- [18:56]
Tara Palmieri
Wow. And how much energy or server power does this consume?
- [19:00]
Luke Igle
Do you have a good sense of that, Melissa?
- [19:01]
Melissa Du
Yeah. So I mean our main database right now is using like 50 plus CPUs at a time, right? That's kind of running all the time. We have five to ten rucksad replicas that are also running with tens of CPUs each. I don't know exactly how to give the number for amount of energy for server power there, but I think you can use the bill as kind of a proxy of. We were spending tens of thousands initially on the platform that we're using. We brought that down to hundreds of dollars, but I think that's kind of the upper limit.
- [19:37]
Luke Igle
So the electricity bill equivalent of hundreds of dollars per month is what we're stabilizing that.
- [19:41]
Tara Palmieri
Oh, that's amazing. Okay, here I was thinking it would be insane. So you've said before that the DOJ has been messy and sloppy with these drops. What are, what specifically are you talking about when you say that?
- [19:53]
Melissa Du
I think there's a few things here I have to say. In the initial and some of the later drops that we were seeing, so kind of volumes like 9 through 12, we had this great member on our team, his name is Diego, and he went through the effort of kind of essentially building out git for the DOJ website. And what that means is he was tracking the hashes of every single file on the doj and when they changed over time. So obviously the DOJ had their initial drop, but they kept on changing that first kind of hundreds of gigabyte file drop over the next two or so weeks. Endogo was tracking a lot of these changes, so the hashes and files were changing. You could See that they were changing the redactions of each of these files. And they made it extremely difficult, as you all know, to even access that initial drop. It was very difficult for any single public entity to download all the files at once. And we realized that part of that was likely just because they hadn't really finished their work. They had stuff online that wasn't properly redacted or that maybe had certain pages that they needed to remove and add. And so Dago surfaced a lot of the examples of you could just pull the files from one day and then compare to the next day and see that many more names in the file were redacted. And so we've done a lot of the work to ensure that all of our redactions are fully up to date with what the DOJ has redacted. And we've also been trying to do that.
- [21:25]
Tara Palmieri
Are you adding more redactions? Are you taking away redactions?
- [21:29]
Luke Igle
Yeah, we want to be really, really careful about this one.
- [21:32]
Tara Palmieri
Yeah.
- [21:32]
Luke Igle
You know, Thomas Massey screamed at Palm Bandy or Pam Bondi the other week about how they redacted a bunch of male perpetrators names. And they specifically had an email called List of Victims that was completely. So that's an example of what we have the responsibility to redact, which is names of victims. When it comes to achieving parity with those redactions, unlike the doj, we can very clearly in our website be like, by the way, you hover over a black bar, you hover over, and you can trust via J mail, this is the name of a victim. Please trust us in our choice to redact this. And like Melissa said, I think the sloppiest part of this was just the total failure to redact the people. You know, the one group of people that the bill says you are allowed
- [22:20]
Tara Palmieri
to redact and have they called you and been like, hey, this is me, Please redact. Have you ever wondered if the person that was calling you was truly a victim or not?
- [22:27]
Luke Igle
There's some really amazing lawyer groups that have done a very good job saying, here is every single file that has it and it's our responsibility to double check. There was at least one case. I think, Melissa, you understand the case better than me, where it kind of seemed like they wanted us to redact the name of a man of a male financier guy, but it also kind of seemed like they wanted to redact the name of a victim who appeared in that same email. So we left the. We left the financier guy's name completely open. We Redacted this name of who was clearly a young woman. There are a few kind of tricky, I would say, sneaky attempts that I think I saw from them that we've been, you know, I. I'd like to think that we've. I think we've done a pretty good job being not letting any of those tricks happen. There was just so many emails, particularly from Ghislaine Maxwell, that the new email Drop says was sent in, like the year 4500. And then another person replies in the year 1900, where the software they use is just so bad, where they take a scan of a email and then someone, the FBI or DOJ clearly took a scan of that scan, possibly with like a physical printer. And then that's what we received. And that's what appears in jmail in comparison. When Riley and I made that first tiny version of JMail back in November, very small number of emails, just 20,000 threads. It was so beautifully scanned. It was presented very carefully by the House of Representatives. It's a pristine archive compared to the 3 million sloppy, unethical, borderline unethical at times pages that were released by the doj.
- [23:57]
Tara Palmieri
Are you still going back and making sure that whatever was maybe redacted and maybe is no longer redacted because the DOJ is realizing they're not in compliance to the law? Are you going back and making sure that if they make any adjustments that it's matching your records?
- [24:13]
Luke Igle
There are a few things. One is relying upon lawyers who tell us things. We quadruple check whether they're lying or not. Almost all the time they are telling the truth. These are real names of young women that must not be in here. Then there's the secondary task of keeping track of the kind of, you know, insanity that the DOJ is doing where they're constantly updating files. We have an amazing member of our team, like Melissa said, named Diego, who built that system that constantly crawls what the DOJ is doing. And we have not yet published, like a diff viewer, a difference viewer, where you can scroll through the difference between the DOJ and us. However, we would love to make that available as soon as possible. And that's where you do the other version of this, which is there are lawyers that clearly reached out the DOJ and not us. We need to mirror the good redactions that the DOJ is doing. Those are the two different ones. It's our. We've, you know, we took on the responsibility of making this website. This is kind of like another. This is A burden we have to take on.
- [25:11]
Tara Palmieri
Yeah.
- [25:11]
Luke Igle
And I want to shout out Clay Higgins. You know, I think he's a senator now. Yeah, he's the, the Republican senator with a very funny kind of very. I've seen some people do some very, very funny impressions of him. He voted no against the Epstein Files Transparency act and everyone knew it was silly for him to vote no. But he did say something kind of funny because it turned out to be true, which is I don't think the government's going to protect. Going to do a good job redacting these victims names. And he was right. But I'm still glad that the files are. Were released and given the broader story here.
- [25:47]
Tara Palmieri
Yeah. Has anyone offered you guys money in exchange for removing or redacting a name?
- [25:52]
Luke Igle
I don't think any of us have gotten emails like that. We've received tons of invitations to make a crypto coin. You're going to make millions of dollars if you just give us your wallet, Luke. You know that kind of stuff. But tons of crypto scammers, but that's it.
- [26:06]
Tara Palmieri
Has anyone from the DOJ or anyone else asked you to take the site down?
- [26:10]
Luke Igle
No. No, not that I've seen. And Google has just been so kind in that there, there are some litigious tech companies that don't like parodies and don't have a sense of humor and we've heard nothing from Google. I don't know if anyone important at Google knows about us. We'd like to keep it that way. If the answer is no, I would like to keep it that way. But I do appreciate that they respect that we're doing. We're engaging in fair use right now. So I, I do appreciate that.
- [26:36]
Tara Palmieri
Have you ever asked Congress to unredact certain documents?
- [26:40]
Luke Igle
We're pretty excited about that. We're allowing people to vote in J mail for the things that they really think ought to be unredacted. And we're using. We're taking advantage of these relationships we now have with staffers sometimes and Congress people to. To do that.
- [26:55]
Tara Palmieri
We've already seen other countries for them. I mean, that's the thing. These members don't really know the case that well. They're not deep in there reading through emails all day long. It's obvious. You can tell that even when they're giving their presentations that you're the one who's really. You're. You're supplying the information for them.
- [27:12]
Luke Igle
Yeah. And you don't even need. We don't need to have a direct line with Any Congress person, anyone, can go on J Mail and look at the top most redacted, the top requests to redact. So if you're a congressperson watching us right now, just go to JMail right now. Click on, like most, you know, emails voted most to be unredacted, and you'll see for yourself.
- [27:31]
Tara Palmieri
Okay, so what email do people want to be unredacted the most? What's the contents of it or what's the nature of it?
- [27:39]
Luke Igle
Yeah, what do you think, Melissa?
- [27:41]
Melissa Du
Yeah, the top email is this email with the subject line ask dads from 2019, June 1st. And it's essentially an email between Epstein and someone who's unnamed. So the one who's rejected discussing Trump coming to London. And there's a mention of meeting with a girl, or there's a few mentions of different girls in there, but a mention of Trump meeting with the girl and descriptions of the girl.
- [28:15]
Tara Palmieri
What year?
- [28:17]
Melissa Du
2019.
- [28:19]
Luke Igle
A month before the arrest of Epstein.
- [28:24]
Tara Palmieri
What's the second one?
- [28:25]
Melissa Du
This one looks like a chain between Epstein and Reid Hoffman. Reid Weingarten. Not Hoffman. Weingarten. Yes. And it's just Epstein talking about meeting up with Bannon. So.
- [28:40]
Tara Palmieri
Talking about. What was that? I'm sorry, Melissa?
- [28:42]
Melissa Du
Meeting up with Bannon.
- [28:43]
Tara Palmieri
Oh, meeting up with Bannon.
- [28:45]
Melissa Du
Okay. So I think there are some that are better leads and some that are worse leads, but kind of we've found some nuggets of redactions that have been interesting. A lot of them are kind of. It'll be Epstein discussing victims or discussing bringing girls to his Manhattan estate. And there'll be weird parts of the email that are redacted. Maybe the fourth message subject or the fourth message contents will be redacted. So those are interesting cases to look into.
- [29:15]
Tara Palmieri
What's your gut feeling? How much do you think the DOJ is actually holding back?
- [29:19]
Melissa Du
I think for me, personally, and this is largely also through talking to some journalists that are looking at the case, it feels like there are a lot of holes to kind of the contents that we have. So one example here is we were talking to somebody who was looking into the contents around Epstein's death. And obviously, we have the recording around his death, and there's the suspicious footage that's missing, but there are also some details, like one of the chief medical officers who was investigating Epstein's autopsy. She is known to have had kind of a trial, but the transcript of that trial is not in any of the drafts. And so that's a case where it seems like we see kind of Transcripts of them discussing with other people that investigated his death and kind of the New York Police Department who investigated his death, but not of this particular medical examiner. Another example I think that's interesting but maybe more speculative is we found Epstein's 2017 tax returns. And he's filing, I think on the order of tens of thousands of dollars indicates that he's maybe in the upper six digits or lower millions income, at least for self reported. But we don't see any of his tax returns from any other years. And you can imagine if the DOJ went to his computer, actually took down all his documents, we would maybe expect to see more financial documents. We also see a lot of discussion between Epstein and his clients. So largely this is Leon Black and Leslie Wexner, but he also did a lot of financial engagements with China through David Stern and so on. And we don't see very, very many kind of financial transaction documents in the drafts either. I think the largest sum that's reported is a transaction between Leon Black and Epstein of like $185 million. And so there's a question of how much of his financial history are you still missing from these drops? And I'm sure, like you as a journalist, I'm sure you have other examples in mind of just strange missing parts of the story that we still have questions around. I have more questions than answers oftentimes.
- [31:27]
Tara Palmieri
It's interesting that you mentioned Leon Black because the head of the sdny, the lead prosecutor over there. I don't know why I'm blanking on his name. It's Jay. I don't know if it's Carney. Oh, Jay Clayton. Excuse me. Yeah, yeah. He was chosen by Leon Black to replace him. So it's kind of interesting to. And he's the one who's in charge of the transparency files, essentially and complying. And so we don't know. We don't know about the transactions with Leon Black.
- [31:52]
Melissa Du
Interesting.
- [31:53]
Luke Igle
I did not know about that connection. I think the one that's spookiest to me, kind of double entendre, is that the CIA didn't really. There's. If the CIA ever responded to Jeffrey Epstein's FOIA request, we have not been able to see what the answer is.
- [32:08]
Melissa Du
Oh.
- [32:09]
Tara Palmieri
Where they're asking if there's an affiliation with the NSA or the CIA.
- [32:14]
Luke Igle
Yeah, I think Rokana, they gave a glow response.
- [32:17]
Tara Palmieri
They said we can neither confirm nor deny because it would compromise sources.
- [32:23]
Luke Igle
That's the Glomar response, Right, exactly. Which I think is a good way to Say no to a lot of requests for transparency that happened here. I think Ro Khanna said that Congress doesn't actually have the power to open up those portions of the Epstein files, which is what the many, many intelligence agencies and members of the intelligence community have written about him. That is currently classified. Obviously. I think it's like the most important thing is what the CIA actually thinks about him and what internal records they have of him. We do have great records now of what a lot of FBI agents were saying about Jeffrey Epstein through this release. I am much more interested in what CIA agents were saying about this guy on the other side.
- [33:02]
Tara Palmieri
Yeah, I mean, that's what I was wondering. Do you feel like a lot of the documents that are being withheld aren't for national security purposes or intelligence ties? Because 2.5 million documents seems like a
- [33:13]
Luke Igle
lot, I would say so. There's also a huge missing portion of just Epstein's personal files. It was very well reported that about 19 terabytes of data was seized from Epstein's property. If we can assume that 10, maybe 100 gigabytes of that is text, which is an unimaginable amount of text, I think we all know that the remaining 98% of that data is video. Like, the only video can allow any single person to accumulate 19 terabytes of data like that. Unless it's, like, video game data. I'm sure a lot of people's imaginations have run wild. Melissa and I reached out to some journalists to talk about it. They say it might just be that he had downloaded an incredible amount of pornography, and that's all it is.
- [33:57]
Tara Palmieri
He had cameras all over his house. They had a room, a security room full of cameras.
- [34:03]
Luke Igle
The only way that a single person can have 19 terabytes of data is through images and video. It's text alone does not take up that much data. It appears that if we've received a couple hundred gigabytes so far from the Epstein estate And from the DOJ, that leaves about 98% of the known data from Epstein's estate just not being in our hands. There might be a very good reason for that, because it might be illegal video of. And I'll let everyone put together the dots, perhaps through his own kind of network that some people assume he has. But the DOJ hasn't really said what that video is at all. Had some journalists tell me that it might just be legal and illegal pornography, that he had nothing to do. That's what I.
- [34:47]
Tara Palmieri
Or it could be videos of his many friends to compromise Them to have them videos of them in compromising situations. I mean, you know that they, that according to his pilot Larry Vasowski, they were putting pin size cameras in Kleenex boxes. The place was entirely wired for video. And you saw they had video rooms for security. And you look inside of the house, there are videos and there are cameras in every, every corner. I don't know, seems to me like the video footage would tell a pretty interesting story, but it could obviously reveal victims.
- [35:27]
Luke Igle
I would agree. And I think we ran into a similar ethical quandary which is like if there is a photo in JML that just cannot be seen by the public, what do we do? And what we did was black it out, but then write a great description of what it is. It's all, all that we ended up getting from like the Yahoo emails that we were able to exclusively publish, it was these headshots that Epstein's executive assistant would send over in this very like gross way. And he would sometimes reply and instead of just blacking it out in a very non transparent way, we just have text descriptions like this is a headshot of a, of a young woman, right? That's what Leslie sent to Jeffrey Epstein. And I think that builds up trust and at the same time it allows you to protect the victims. And I think the, that's the DOJ really wants to kind of play this hyper redaction game that they're playing right now. At the very least describe what it is that you're redacting. And I think a lot of trust from the public might be recovered slightly.
- [36:26]
Tara Palmieri
It's interesting because like Antoine Verglass, he's this photographer that was taking photographs for Jeffrey Epstein after his conviction in 2008 and he sent him a file, files that said white and black nudes. But I would like. And then I asked Antoine about them and said, who, who are these nudes? Because they were also apparently doing a nude shoot and he claims it wasn't. Even though in the emails that said won't need styling because it's nude. He said, oh, they're famous nudes. Like the ones you would find in my studio. If they're famous nudes, wouldn't they just be? Couldn't we just see them? Why are they blacked out? Unless the DOJ is just being slow, like, just like, oh, it's a girl, black, black, black, black, black. You know, woman, woman, just block it out. But if it's like a nude model, that is a famous nude model, why would it be blocked out? Right?
- [37:17]
Luke Igle
It's to just as I was saying kind things about, about Clay Higgins for correctly identifying that the DOJ would do a bad job. I do think that after one portion of the story that we didn't get to talk about as much after we were given access to Jeffrey Epstein's Yahoo account that the DOJ and the House of Representatives did not put out, this was through Dropsite News who reached out to us and gave us this very like carefully stewarded data set. We ran into like this oddly similar situation that I think the DOJ ran into, which is how do you find out if a woman who is mentioned in email is a victim, an accomplice or just a random person?
- [37:56]
Tara Palmieri
Right, right.
- [37:57]
Luke Igle
And it was tricky. And the only way we were able to do it, we did it on a scale 1:1000 of the size of the DOJ has had to do with an entirely volunteer team of people loosely connected in single group chats with journalists. We just built a tool in JML on the other end that allowed journalists to find that out for themselves. Once they press the button, once we have a team of people who say that this is a safe group of people, then they send. I don't envy anyone who has to go through that redaction process. But I think we all know this has been a comically bad outcome. The exact thing that I would say the people who are critical of the White House thought and claimed would happen does it does seem to have happened.
- [38:35]
Tara Palmieri
Yeah, you know, I've thought about that. The fact that he doesn't really email before he gets out of prison in 2008. Right. Are there any emails from Jeffrey Epstein before that?
- [38:49]
Luke Igle
There are tons, especially in his Yahoo. The earliest we found was 2000 or 2002 where he's sending a draft to himself about how he's going to yell at Vicky Ward to kill a story in Vanity Fair and he's saying your boss Grading Carter would never let this happen. This is bad journalism. I think Ghislaine Maxwell was an early adopter of technology. Her sister and I think multiple siblings achieved a great exit for a startup that they built in the Bay area in the 1990s. There are emails from Ghislaine Maxwell before the year 2000. There are probably also emails from Jeffrey Epstein before the year 2000. These are very tech forward people. They were constantly sending emails. There is significantly more that.
- [39:31]
Tara Palmieri
Do you think they're on their own private servers?
- [39:35]
Luke Igle
Yeah, I mean Yahoo and Gmail are hosted services. I bet they had their own cool, expensive, cutting edge email servers back Then.
- [39:44]
Tara Palmieri
And we'll never have access to those,
- [39:48]
Luke Igle
I don't think. I don't expect the DOJ to come out with it, but I'm not sure. I. I don't have any insider information there.
- [39:56]
Tara Palmieri
You haven't gotten any tips about that?
- [39:59]
Luke Igle
The best tips I got is the fact that there are around, you know, 18 and a half plus more terabytes of data that exist from as a state. And I have also seen, just like everyone else has seen, comments from Rokhana. Really, the congressman who's done a great job with all of this, claim that there are many classified documents that he, as a congressman, does not have the authority to release tips about other emails a bit less frequent. I do know that amongst the Yahoo data set that only JMAIL has been publicizing, there aren't a ton. There's such a weird gap in the record when you go all the way back to the beginning is that it does seem like Jeffrey Epstein or someone else deleted a lot of those Yahoo emails, but still left a lot of incredible things. And to shout out Ryan Graham of Dropsite, who originally reached out to Riley and I. It was through this very interesting set of emails that he allowed us to add to jmail that the authoritative link between Jeffrey Epstein and Iran Contra was revealed. And it was also through those emails that we received that we saw this very damning email from Les Wexner on his Yahoo that he managed to not delete, where Les Wexer says, uh, I warned you, or, you know, you broke your number one rule, which is always be careful. Yeah, this is something that Les Wexner emails. That was JMail. No one knew about that. JMAIL is the one who publicized that email. Uh, and I think a big. What felt like a big moment for us is the other week where they finally deposed Les Wexner, a congresswoman I wasn't able to tell who asked him live on camera about this email. And he's just squirming in his chair trying to figure out how to answer this. The curious explanation he came up with is he said, oh, I said, be careful about your finances. Congresswoman says, well, it seems like you were referring to being careful about getting caught for soliciting a minor for prostitution. Right. And he has an okay answer after that. This entire saga, specifically of this very, very small data set, which is the yahoo stuff that JMail hosts, is what gave me a lot of perspective on how bad of a job that DOJ did on their own data set. And also it's what allowed it to feel like we are doing some original work here and that, you know, we're helping contribute to this whole corpus.
- [42:13]
Tara Palmieri
Well, thank you for your work. We appreciate it. Without you, I would be spending a lot more time doing reporting. And I really appreciate it. I really do. And I believe that everything in life is supposed to line up exactly when it's supposed to. And we couldn't have done it without you, without the AI structures and without the legions of independent journalists and online sleuths. So I want to thank all of you too. Sometimes justice has to wait. Unfortunately, this has been a long, sad story, but I am glad that we finally have the structures to be able to pour through all of this. Thank you so much, Luke and Melissa for your time. That was another episode of the Tara Palmieri Show. Thank you so much for tuning in. If you like my reporting, just hit that subscribe button or follow like share. Tell your friends about it. Leave me a comment. I love hearing from you. I want to thank my producer, Dan Schiffmacher. I want to thank Abby Baker, doing my social media booking research. She's an all star. I want to thank Adam Stewart on the graphics and Dan Rosen, my manager. To you again soon.
- [43:31]
Tamsen Fadal
Hi, I'm Tamsen Fadal, journalist and author of how to Menopause and host of the Tamsen Show, a weekly podcast with your roadmap to midlife and beyond. We cover it all, from dating to divorce, aging to adhd, sleep to sex, brain health to body fat, and even how perimenopause can affect your relationships. And trust me, it can. Each week I sit down with doctors, experts and leaders in longevity for unfiltered conversations packed with advice on everything from hormones to happiness and, of course, how to stay sane during what can be. Well, let's face it, a pretty chaotic chapter of life. Think of us as your midlife survival guide. New episodes released every Wednesday. Listen now on Apple, Spotify or wherever you get your podcasts.