 Okay, I'll go ahead and get started. Okay. Hi, I'm Greg Conti, and thanks for coming. I think this is a topic that we really need to start talking about, the amount of information that we're giving to a handful of very large information service providers. If you think back to how long these companies have been around and we're going on a decade for some, and the different vectors that we provide information through web searches, free blog hosting, email accounts, image search, scholarly search, the list goes on and on. As an example, Google has 56 currently different primary applications. So if you think back from 10 years ago when you first started using variety of these different companies, you've got 56 different parallel trails that exist, and I'm going to argue that I doubt any of that's being thrown away because of the business emphasis. So it's, I think, a topic we need to talk about, and I just hope to do a couple of things during the talk today is, one, go through some of those vectors and look at what are we giving away, as well as look inside some real world browser caches, look at can we be fingerprinted beyond what you normally consider as fingerprinting, you know, there's the cookies and there's registered accounts and IP addresses, but can the semantic information we provide, such as that browser cache you just saw, if you go somewhere else and everything else changes, can you be re-identified and take a look at that? And then, you know, I think some of the people in this room use privacy countermeasures and just what is the current state that, you know, how useful are they is something I hope we can get to at the end with some discussion, as well as, you know, where we need to go next. So I am, I'm a fan, you know, personally I'm a fan of Google and I use their stuff a lot, they're my default, default web page, but they're also, you know, they're the elephant in the corner, they have the, in my opinion, the best set of tools and services out there in this space, so I'm not saying we need to do away with companies like Google, but I think we need to find a middle path that balances allows those companies to flourish, but at the same time allows, you know, our privacy to remain, you know, appropriately protected. So I'd like to, you know, we really ought to all work to be part of the solution, we're all in this together, but I think this is an issue we need to address sooner or later, because it's going to be, you know, an issue and I'd rather address it sooner, so with that, these are my own thoughts. I'm here as a free citizen and not the representative of the United States government. One way to look at this is it's kind of the inverse of Google hacking. Google hacking is all about interesting and creative ways to pull as much information as you can from companies like Google. Well, we're looking at the other way. You're turning around 180 degrees and putting, looking at, you know, kind of like me, say, I'm Google, you're all interacting with me, what do they see? And what, you know, what does that look like over time? And to start looking at, you know, start looking at the question. So just to get you thinking in the right direction, you know, imagine various scenarios that, you know, we're investing a lot of faith and trust in a handful of really these consolidated information service providers. Now, what could they do if so inclined, you know, for example, would your pat, if your company was researching a patent, would that be visible to one of these companies? I'd say probably yes. You know, IP address ranges are published and it would probably be pretty straightforward to start identifying these taking place. You know, if you had a serious medical condition arising in your world, would that show up? What if your company was going under? Would it, you know, would there just be the surge of job search searches going on from your company? Companies IP space. How about your anonymous blogger or say an anonymous hacker and someone knows your handle and your name and searches on it to see if they can find, you know, one of your friends, for example, or someone, you know, a fringe person trying to find your webpage. So you could lose, you know, anonymity just by what other people are typing in about you. And this is, you know, a true example. I had a friend who had a death in the family and they used a free online email service to email basically their entire family to let them know, give them an update. At the bottom, there was an ad for bereavement counseling included in the email. So it's just that you see things like that and that means that those are telltale signs that we need to start thinking about this. So I just have a few quotes from Usama Fayad, the chief data officer at Yahoo to just give us some statistics. And at this interview, they asked, what are some of the biggest data mining challenges you now face at Yahoo? And his answer, and I underline I think what's the important part, was that Yahoo users generate 10, and this is in 2005, 10 terabytes of data per day, which is the equivalent of the entire text contents of the Library of Congress. And he's not talking content of emails or anything like that, that's like user interaction data. And then he goes on to say, the first and largest challenge is the ability to capture it all reliably. So ask yourself, how much of that is being thrown away in an era when business models are built on a customized service, and that's a treasure trove of information. I'd say very little is probably being thrown away. So this isn't a surprise to these companies. They are very well aware that the privacy, you know, that people have concerns. And what he said is that we would never do anything to compromise our users' privacy. Their business model is built on that trust. But, you know, I think we know that sometimes things don't work out the way we'd like. So just the mere existence of this data can be a problem. And we live in a world where it's some unpredictability, and we'll get into that in a second. And there's also a tension. There's a tension between, say, your mission was to organize the world's information and make it universally accessible and useful, a good thing. But you also have shareholders you need to please. And, you know, profits be made to keep the business solvent. So there's this tension. You're being pulled arguably in two directions. And that balance is important. It's hard to have them both pulling in the same direction. And recent news hasn't been particularly encouraging on this subject. And some of the highlight, I've just pulled out a few random headlines about Google being granted a license to operate in China, but it was at the cost at the time of censorship. And it was just a requirement to do legal business in that company. There are several instances of subpoenas. And the Sony root kit, you're all well aware of that. And actually, that's a Dan Kaminski image in the background showing the root kit infestation in red. So, OK, let's say there is this tension pulling in both directions. But ultimately, they're really going to try and protect our interests and keep the information private. Well, there's always problems of things like insider threats, accidental information disclosure. Ask yourself if you've gotten a letter, or actually, who's gotten a letter saying your information has been compromised, gotten a letter. So, I mean, bad things happen. I know the Veterans Administration lost 37 million veterans information that whole instant, I'm sure you're aware. And other things, sales of companies, change in leadership, people retire, people are various changes take place. And then there's always the legal compulsion to have to be forced to reveal the data. Just so, again, back to the mere existence and also the possibility of eavesdropping at some point as well. And I mean, you could always wonder, well, what if it isn't an entirely altruistic provider and they have full access to the data? It's at least worth asking the question. So if you think about, and this is just a graph showing on the left-hand side are web services that typically you've released small amounts of information to. And then moving to the right, sites you've probably given more information to. And everyone's position will change. You can argue with me about how they're positioned on there, that's fine. And they will always shift to the right. I mean, you can assume once the information's been disclosed, it's probably not going away. So you've got these just sliding down more and more information being disclosed. And for example, URL shortening service, online bookseller, online auction, you start getting up into things like content, instant messaging, web search, file transfer services, and things like that. And then all the way then with the red dot is a company, like I said, Google has 56 core products and services that they offer. So they've got all of this cumulatively. And even something as simple as tiny URL, which is a URL shortening service, they've shortened 16 million URLs and have 325 million hits a month on them. So even that creates a tremendous social network of what people are interested in, what they've bookmarked, IP addresses of all those people. So I mean, just that something as simple as a tiny bit of information, what could be done at the farthest end. So I'd like to, again, I hope to make this a little bit interactive. And who is searched for things they wouldn't want their grandmother to know about? And you can replace grandmother with your spouse, your significant other, your employer, your government. Who's done that? I know the lights are in my eyes. But I would say probably everyone. And somewhere in some database somewhere, I think it's probably still there. OK, so if you think about your web search activity at some starting point in time, you've slowly disclosed information or rapidly disclosed information. And then maybe you've picked up, started using web-based email and picked it up and started using that as well. Instant messaging, say you used it and didn't like it, your disclosure over time would flatline. And then maybe you picked up internet voice over IP or using that through a service. So this is just one way to look at the problem. You've got these curves. This amount of data, they don't go down in my assumption that the data isn't being thrown away. So let's take a look at Google. And again, it's nothing personal. They're just the most influential and successful company out there. In my opinion, I think we ought to take a look. So I don't know if you're familiar with Google Zeitgeist, but it's an annual thing that they do where they list some of the top searches. And this one in the left column are top general searches for O5. The middle is top news searches for O5. The right is top frugal searches, purchasing searches. Ask yourself if it's technically possible to do it on an IP address to do the same thing. Is it possible to say, well, what is the Google Zeitgeist of IP address X, Y, Z? I don't know. It seems like a reasonable thing. And here's an example of this is search terms over, I believe this is O5, of Wikipedia. And you can see the growth. They didn't have a vertical scale on it, so you don't really know. But you can see that over the entire whatever subset of their data that they used or the entire set of data. But imagine, and here's another one with tied news. So imagine just looking at that, what about if you're a registered user, or you're at a constant IP address or something like that, that you could have these little charts, what were your top 10 search terms over time correlated to the news? And there's layers upon layers of these things. If you start fusing together these flows of data, those 56 flows of data, and you think about it from your individual perspective, on your social network perspective, connecting the nodes or connecting the business, the nodes either in a social network or grouping them into businesses or other informal or formal organizations. And then you look at the resources of a company such as Google, and this is what they were hiring for. And you see things like data mining, genetic algorithms, information retrieval, machine learning. And really, there's a lot of top tier talent there. And they're all working hard to get the most out of their business model, I would think. And I'd argue that companies such as Google and Yahoo have really world-class human resources, information processing resources, and stores of data that rival that of a nation state. And if you look at, again, a company such as Google, where does the information flow to this kind of black box? And at the top, you think, well, information others provide or retrieve. And that goes back to my example of, well, people are searching about you. Can that be correlated to you? So that's one source of information. Other people are providing information about you or your organization through their interactions, as well as information that you provide. And finally, that information Google directly retrieves and think Google bought. So you've got these flows of information. We'll look at a couple of, primarily focusing on what we provide as individuals and companies. And I'm calling those the disclosure vectors. But first, think back, when did you lose your Google virginity? And actually, this was a quote I found on the internet, but I thought it was pretty good. But it was probably a while ago. And if that five years or whatever or other, it would vary by the company. But that trail, in all likelihood, exists. And then you combine that trail over these 50-plus different services and then think about what information you've provided or retrieved. So for example, I consider you all close personal friends. So I decided to share with you the words I either didn't know how to spell or I didn't know the definition of. And when I was putting this together, I decided to primarily share my own data because it's actually very sensitive. When you see what's in your cache and what you've been looking for, it's not something you want to put on TV, necessarily. And I'll take that off. And if I put it on the web, that slide won't be in there. And the same thing, just a snippet of Google research that I'd done. So I sorted it and pulled out some of the search terms. So as I was looking at doing Google research, these are some of the search terms. So again, take it 180 degrees. You're the company. What could you see these constant flow of data? If you can partition that and cluster it appropriately, what's the picture that's being painted? And not just through search, through all 50 of these different potential lanes. So ask yourself, do you use Google as a address book these days? I rarely write phone numbers down. If a person has a reliable web page, I just constantly look it up. I use Google. So on my zeitgeist of Greg Conti, there'll be this spike of activity as I'm working with different people on different projects. And then it'll decline to nothing. Think about it with something like Google Alerts that you can have it automatically inform you about different news stories or web activity that occur. So your email address. And email addresses are one of those things that can then be used to connect nodes of data, as well as search terms. Who's looked at their home on Google Maps? OK. And when they up the resolution, the first thing I did is I looked at where I lived. I looked at my family members. I looked at places I'd lived, as well as military bases, because I want to see if they were on there. So they've found basically everything, very important areas to me. I've just told them where that was. So that's another stream. So these flows of data are cumulative. Your web search stream, you've got your mapping stream, you've got directions where you're getting directions. Of course, there's Gmail and things like that. Even if you regularly encrypt the payloads of your emails, let's say you do that, there's still the message externals. There's email addresses and sources, destinations, time and frequency, those such things that we'll still show through. And it says right there they don't throw anything away. So you'll never have to delete another message. And then another interesting area, and this is another connector type technology, that it allows you to sign up for Gmail via your mobile phone. So you're connecting the two clusters of activity with activity, disclosing information such as that. And one thing I was surprised was the number of people in the security community who actually were active Gmail users. I think part of it was there was a very successful marketing campaign with the invitations and all. But even if you say, OK, I want to stay out of this space, if your email address is even touched, if you've ever contacted a Gmail user, you're now part of what could be recorded. And think about it as far as your employees. If you have groups or when you say employees, think about whatever groups you're in, are they registered users as well? And you've got other information flows, such as looking up phone numbers, residential and commercial, if you're traveling, if you're collaboratively working on a document, if you have a calendar. So these are all flows. I mean, each one of them, if you put them all together, it's really frightening. And then there are mobile devices. And you start seeing technologies that know where you are, essentially. And then going farther up, I don't want to get into too much detail today. But we do place a lot of faith and trust down the road. We place a lot of faith and trust. We run applications on our desktop, and it's indexing our desktop, making it easier to use. And even doing the same for our enterprise. I just saw a good slash dot quote where they said, asked if Google was going to be Skynet. And I don't know. You think about it. Slash dot's good for quotes like that. Oh, and then information others provide or retrieve. So I want to take a look at that. And for those of you that maintain websites, you've probably looked at your logs and looked at key phrases. So these are how they've found you. Sometimes it's just your name. But other things, it's other type of activity that they associate with you. And it could be your real name in a handle, for example. So I've got just some examples here of what came in, some key words that came in through Google. And different spellings of my name, paper writing things I had done, programming projects I'd worked on, military units I'd been affiliated with, and then different writing things. So this is just a snapshot. I mean, if you pull it up, you'll see. Now, the secret sauce is connecting these nodes together. But I'd argue with the resources of a nation state and top tier talent, intellectual talent, that it'd be a very real possibility to start putting these together for their own client. And then, as I mentioned earlier, that there's the Googlebot piece. And I won't go into that. But that's constantly trolling out there. And again, if that could be connected to you in some way. So the idea then is, OK, you're thinking, I delete cookies, I use Tor or something like that, or I'm behind a network address translation firewall. And it's very hard for me to be uniquely fingerprinted based on traditional means. So one thing I want to, a research area I'm looking down, is the idea of, again, we provide a certain amount of information to a given service provider. And that somewhere along there, there's a threshold of information that we provide. And if we provide enough information, they will know who we are. They'll be able to identify who we are. So that's the basic idea. Again, this is just theory. But sure, I think it holds. Or if you cross that threshold at some point in time, then you're uniquely identified. In another example, if you are providing a certain amount of information, but you never cross that threshold, then you're OK. And I have the line sloping downward, that fingerprinting threshold, just to take into account for possible advances in technology. And this says, stop googling yourself or you'll go blind. OK, let's be honest. Who's typed their name into a search engine before? OK? So just something to think about. I mean, it's really mind boggling when you think about all the different flows of information, how it's continuous. And it's almost counterintuitive. You think about usability being a good thing. You have more usable tools. Well, what we have here are tools that usability lures you into disclosing more information. And the same thing, cryptography often is normally a solution we can apply, but disclosing the information is the cost of doing business. You can't encrypt your search terms before you give them to the company. It doesn't work that way. So just to look at some of the different ways that could be identified, you think network layer type activity IP addresses and things like that, operating system fingerprinting and the like. Browser environment variables, cookies, the traditional one, then you continue upwards the Google account. And then the area I'm really focusing on for this talk is the idea of the semantic information, all these flows bringing them all together. And then at the bottom, managing your IP address, managing maybe anonymizing your browser environment variables, and cookies are what people think traditionally. And you could probably include the Google account as well. But that's what people think of as anonymity. But it's that semantic information is very hard to disguise. OK, so this is your information space. And you use these different services. And these are all variants of search. There's book search, web search, like a scholarly search, and an image search. So you've got these types of activity. This will vary for each of us. So there'll be different sizes, small, large, depending on the activity. But I just want to give you a feel for the information flows, just what we're giving, what we're giving. And then you've got other type of flows, email, instant messaging, blogs, word processing, collaborative word processing. And what these are, they're connecting technologies. So you've got this cluster of activity that's you, but things like email, instant messaging, who goes to your blog, who's working on the word processor, sharing your appointments on the calendar, they can connect the nodes together. And then you've got other activity that gives you more of a feel for your location. And you've got mapping, language translation, local services. And then you've got financial information, as well as news and alerts. So I just want to give you a feel of those 50. Here are some reasonable ones that many of you might use. And just to think about, that's all you. That's the cluster of your activity. And in reality, that's the cluster of your activity from a work PC, or from a one given platform. And then think about the other places that you use. And you may have a mobile device, or a home PC, or a laptop that you use to varying degrees. So the secret sauce then is connecting these nodes. And when you've got things like these light blue nodes that give that type of information, allow you to connect them together, even if cookies don't hold, you'll have different cookies on there, other things well. Particularly again, going back to the resources available, this seems extremely possible. Okay, now we've gone from this is you, now that orange dot is you. Okay, so I'm just trying to think higher. Now what you've got, you interact with a bunch of other people in some group collaborative setting. It may be your employer, it may be a group of friends, maybe a school, something like that. So connecting on many, and this will vary by, there's many slices here. So, how can you connect the dots together? And you've got network addressing information is obviously a good start. What company owns this block of IP addresses? As well as, are these searches similar, or can you connect the dots again with that semantic information I mentioned earlier? So you're trying to group them together as being related in some way. Or it could be the inverse. You have one big cluster behind a NAP firewall, and you've got this big cluster of activity, and you're trying to split it up down to individuals. Who has a Google cookie on their computer today? Okay, so I mean the default setting, there are just, and I appreciate you answering the questions because I think it's important other people would be interested in just getting a feel for what, I'd argue we're probably the, one of the most security conscious groups on the planet. So if a lot of people have it on there, it says something. Now think about it, as you're trying to group these, or perhaps you've got behind a NAP firewall, and you're trying to split it out into different clusters of activity, ask yourself what's going on in your place of employment? Or if you are a registered, if you Gmail, you're a registered user. Now let's think a little bit about countermeasures. All right, we have this information disclosure, we're giving information away. There's the fingerprinting threshold that we need to cross. The idea then is to alter the information we provide, the value of the information we provide, so that we are delaying, hitting that fingerprinting threshold, delay or altering or failing to release the information. So that's what these countermeasures are doing. And there's a lot of them out there, I only have a subset here. In the work I've been doing, the people I've been talking to, there really is not a usable solution to maintain high levels of privacy online continuously. That's the feedback I'm getting. Does, just to show a hand, does anybody disagree that you can maintain continuously do this without, basically, continuously do this without a problem, without being a big pain, such that you turn it off? So, who actively manages their cookies? Okay, so a lot of people. And this one, I didn't know if people would wanna answer, but I guess I'll ask and see what it's like. Who regularly uses web anonymization tools? Okay, so a smaller percentage. I should have given everyone big yellow hands, you know, the, pull the finger up or something, because it'd be easier. Okay, so I've got a couple of examples of, let's see. So what I did is I went to co-workers and some of my co-workers' families and they kindly shared their caches with me. I would love to have gotten some student caches and had them, what I asked them to do is take their cache and I extracted the search requests. And it's in Firefox, there's some form data you have to remove, but it's pretty easy to tell what's a form field type, you know, a typical like, you know, name address, telephone number type stuff, because it merges search data with, it's not broken down by website, so you can't pull it out based on one website. But I extracted it and they went through and removed things they didn't wanna share and what was sensitive to them. So I have a couple of examples. The first one is the one I showed at the beginning. He's a relatively light Firefox user and it was over, from February 05 to July 06. And when he went through, 5% sanitization. So his opinion was, and again, this is a employer network that's monitored and the person's basically an Eagle Scout. So there really, you know, there wasn't a whole lot of, he just didn't wanna share names of family and friends and some other things that were like location-based type activity in the cache. And this is a whole other question about what's actually cached in honor computers, credit card numbers and social security numbers that he also removed. Now this is a spouse, you know, someone's family member at home from a November 05 to July 06 and this person was a heavier Google user and didn't feel comfortable sharing 24%. Okay, so think about this. 24% of the things you don't wanna share really, what about, we're providing 100% to somebody, so to the service provider and that's just again, one of 50 potential information lanes that we're working on. And again, they chose things to remove names, locations close to their home, phone numbers, a friend's blog, that type of thing. And this co-worker is actually me and I had about 4,000 entries and this is about 18 months worth of data and 31% I didn't feel comfortable showing you and I have this, a subset of this that I'll show you in a second and what I did is I removed things that I didn't wanna share, I put it in brackets. So I wrote what it was generically and I'll just give you a demo. And it'd be interesting to see, I'd love to see what, like your average student, what they are doing and then probably a couple of people in the room have really interesting caches as well. So I set the speed a little faster and hopefully it's not too fast but I wanna get through more terms quickly. And as you look at this, consider about what's your flow look like, what are you giving away and if you took a look across your organization, say you're in a company and you take all of these across your company, what picture is being painted that little innocuous search box is can be tremendously dangerous. I thought about adding random terms in there just to see if you're staying awake. But then I thought, I don't know, big brother or my employer would come and get me for putting it up. But as you're watching, when I say unknown name in there, for example, I just didn't remember, I searched for somebody and I don't even know who they are anymore. But things that shine through this pretty quickly, where my website was, people, authors, researchers, I wanna see their work, location starts becoming clear, general parts of the country for sure, what areas I'm interested in researching. And I have one more example. And this was, like I said, I had family members of various people kindly participate. So this is a family member. And you see a entirely different flavor. So I went through these things like 15 times to make sure, you know, didn't it say, well, you know, I don't know, things that would get me relieved. And it should be done in a second. I mean, everything from technology, research area, hobbies, I don't know who the destination of the black roses was, but very sensitive stuff. And this is the things they wanted to share. Recall that this person removed about 26%. And I removed 31%. So that one, this one you just saw, didn't, the individual didn't add the bracketed, you know, changed information. So it was just missing. And this slide is, I can't explain it. It's just a funny feeling that I'm being Googled. So just to recap, we've looked at, again, I wanna emphasize that just because I use Google as a case study doesn't mean I think they're evil or anything like that. They have tremendous tools and services that I use all the time. But I think this is a question though, like what are rising behind the dam as something we're gonna have to, you know, come to grips with. And the people in this room are, you know, five to 10 years ahead on the security front. And it's something that is worth having the discussion now and jointly seek solutions that, you know, are sustainable for the long term. And so we've looked at, you know, the 50 plus different vectors thought a little bit about what each of the vectors provides. And then the idea then of you've got these flows of information, how can they be grouped together as to you as an individual on one machine, to your set of machines that you use, to then that cluster of activity being grouped on different layers with you as a, in different configurations with you socially, with you at work, with you on an innumerable number of topics, how can they be grouped? And then finally a little bit on the idea of fingerprinting and how, you know, how this could be done. And there's, you know, some top tier talent there. You know, I'm sure they can do some amazing things. Okay, so with that, let's, you know, have some discussion or any questions. I guess I could ask a starter question. Am I nuts or are we giving away a lot of information? Okay, yeah, I'm right, good. But no, any questions, please? Okay, yeah, I'm blinded, so don't hold it. The question was, have I tried talking to Google about this? No. Sometimes you don't want to ask. I mean, and again, I don't think they're doing anything. Yeah, they want to make a profit and the trust is inherent to their business model. Okay, yes. I think there's actually a microphone, but that might be a big pain, but go ahead, sir. Yeah, what he mentioned is that the number of, that the people who had more information, more searches had more they wanted to protect. Yeah, and I think that's an interesting, what I really need is to get 1,000 people to do this and then I have nice statistically significant results. Okay, all right, good. Let's go over here. Yes, sir, a red bag, yeah. The question was, do I know if any law enforcement agency has successfully subpoenaed Google search terms by get from an IP address? Actually, I'll turn that around because I would think the hive mind here, if it's been done, someone would know. So has there ever been a successful subpoena, at least that we're aware of, based on IP address of Google search terms? Yeah, I'm sorry, I can't hear. These big honking rooms, actually, I don't know off the top of my head, but hard to say, and there's also public knowledge and then there's some private mechanisms to get information as well that maybe worth thinking about. I have no personal knowledge. In the Hawaiian shirt, yes. Is this a solvable problem? That's why I'm here. Hopefully we can find a balance. The question was, with everything from stoplights to just about many, many forms of technology, all these information flows coming in, can we stop it? Is the problem solvable? I think the first step is raising awareness and asking the question and trying to find useful solutions. There are promising things out there, but it's hard to say ultimately they'll be successful. But I truly believe there's a middle ground we can find. Sir. Yeah, I was just wondering, how well does Google lock down their information while they may not be evil? Other people who get access to that data might have less than honorable intentions. Well, I've never worked inside Google, but by all accounts, they take the protection of that data very seriously and almost like arm vaults and things like that. And then also mechanisms within their company by all accounts are very tight. It's tightly managed who has access to that because you're really talking national level stockpiles of information. This is like a national treasure in a way and it's extremely valuable. So thank you. Google information is not very heavyweight. It's just casual searching around on the internet. If you wanna find out what somebody's interested in, what do they spend money on? Like their searches on Amazon and what they spend money on Amazon or eBay, that's a better indication. Anybody can get on Google and search for anything. Yeah, I mean, so I looked at your cash search and it shows that you're a nerd, so what? Well, and I think that comes back to the larger question about privacy. I mean, this information exists. Well, McNeely said privacy is dead, get over it. So I think he's right. Well, is privacy dead? What do you think? Should it be dead? I guess it's probably the better question. I'd say I'd like to keep some privacy. Okay, thank you, yes. Can you hear me? There's a kid out there somewhere, he's a teenager, he's gonna be the president in 50 years from now. He's unaware that Google's tracking all of this. Should he be worried in 50 years when he's president that what he's doing as a teenager and searching and his connections will be revealed somehow? Well, it's larger than just like using a search engine. I mean, some of these people, some people are putting some things online, that's just a hot topic going on that's publicly available. It's not even hidden in the coffers. I'd say though, my opinion, the mere existence of stockpiles of information, really world-class, well, it's a threat that it exists. I mean, it can be extracted through a variety of different means that we kind of hit at the beginning. You know, legal compulsion, theft. I mean, their Information Disclosure Act, Disclosures occur very, very frequently. Okay, thank you, yes, five minutes. Okay. That was an interesting point. You mentioned a nation state at one point and their technological capabilities and that occurs to me, like even if somebody like Google is not evil and they are protecting the information, I kind of think about the possibility of some foreign nation deciding to buy a public company and what they might be able to do with the information that they gained by that purchase and if companies would have the ethics to actually erase the information, that's obviously destroying corporate property that's been bought. That seems like a very difficult problem. Yeah, that's a very good point. I mean, where corporate climates change, companies are bought and sold, information often goes along for the ride and it's in the world now, I mean, could very easily cross international boundaries. So yeah, it's definitely, times change. So yeah, it's a very good point. Sir, yes? Ah, what are my thoughts on the internet archive? I haven't really put a lot of thought into that. I've been more focused on just information flows. Personally, I like it and use it, but that's probably, we could probably find both sides in here as well that would debate one way or the other. But again, that's information we've put out there. People have made a conscious decision to post that. Yes, but I'm really talking about as well from the provider, the host of the blogs, which is, Google has that as one of their services. That information, from their perspective, it's not publicly available information. I think there's a line from what they can store from their server logs and things like that. Okay, good, we're out of time and we're actually on track. So I thank you for your time and attention and I'll see you around the con. Thanks.