 Wer hat letztes Jahr ein Security-Nightmare erlebt? Ich glaube nicht. Wer hat in der Familie ein Security-Nightmare erlebt? Wer hat ein Security-Nightmare in der Familie erlebt? Und wer hat dein eigenes Security-Nightmare erlebt? Jetzt kommt Ron und Frank. Jetzt haben wir Ron und Frank Security-Nightmares. Ja, willkommen bei Security-Nightmares. Hallo Leipzig. Ich glaube, das ist das erste Mal. Ich glaube, das ist das erste Mal, in dem ich Probleme hatte, das Ganze zu bekommen. Welches Problem ist das? Das Event ist jetzt ähnliche. Es ist das 18. Event dieser Art. So, das Event kann jetzt bereit sein. Und seit Oktober, so kann es allen auch. Der Audienz applaudiert. Sie darf jetzt auch selber Verträgen abschließen. Du kannst auch mobile Plans sein, Cellphone Plans, High Percentage Alkohol, Tobacco und consumieren. Und bleiben in places, dangerous to youth, like this place. And Lotteries. That means Bitcoin. Did someone say Bitcoin? And subscribe to the Datenschleuder. So we're the event that makes sure that everyone gets to the final talk. So how bad was it this year at Christmas? Who removed the Windows XP from a computer? One, two, applause from the audience. And who wasn't allowed to remove the Windows XP this year because it runs so well? Well then we can just go home I guess. What I found really funny this year, we always say that you can relax here a little because you don't have to support others, make tech support here. But I heard that one or the other teen-Ager didn't come here because the rest of the family is here anyway. And you already had them for three days and that's just not possible. And then you have to think about how you can deal with that because Stream isn't the answer to everything. So who had to get rid of one virus or Trojan? Who had to remove more than ten? Nobody? Hmm, there's something happening. Not more than five either. Who didn't know how many it was? Ah, the method has changed. You just set the machines up straight from scratch. So, let's start. This is a picture that my friend Pavel Maier made. His job is making machines be able to see and he determined that stop lights, to machines they look like this because if you take a picture of him at 60 frames per second, when it's changing between red and green, it looks like this. This strange effect that you're observing now. And that's what we drive our cars with. Sorry, no, that's what cars drive us with. So, the beloved Rubrik ten years ago, what happened ten years ago, citizen Trojans for more citizen participation, for reference to state Trojans. But the digital age didn't stop at the political parties. And for example, there were some leaks. You guys saw that, right? How the parties got hung up with that. I think that's worth, a hall's worth of applause. The other thing in terms of citizen Trojans and citizen participation is this Android app. We have Haven by the Guardian Project and Freedom of the Press Foundation. Edward Snowden made some advertisement for that, if I'm correct. Which is the thing that would be forbidden in Germany if we would place it here? Well, let's see where that leads. I think that's one of the first signs that evidence collection systems, for example, if you leave a hotel room, then you might want, you don't know, is that a public space or not? Even though you just left the hotel room, I think this concept of public space will, we have to deal with that for quite a while now. Remote government. Surveillance work, outsourcing them. There are supposed to be private persons, patriotically minded private people that brought down Iceland. Patriotically minded private citizens, they're able to take care of their electoral systems on their own and didn't need other observers. Putin claimed that it was independent individuals, as opposed to ones acting with government coordination. So that was the point. Und so, excuse me, it was Estonia and not Iceland. Modular superworms. The whole thing about superworms becoming a little more modular, like ransomware, as we've seen recently. And then we have the topic of biohacking. I think there were some people who had Able Law on a PowerPoint. Yeah, yeah, and since then. Some people on stage bowing to Frank and the other guy. Excuse us, we are pausing there in the translation a little bit because we're so confused. You can see that biohacking has made immense progress as evidenced by the people crashing the stage just now. I don't know if those people have been modified with CRISPR. But I'd like to see that soon. So this CRISPR thing is a method for gene editing. So, we can take it quite far, the FDA was just warning that it was trying out itself. And so I was hoping that we could get a little bit of that for the conference. I could imagine so many cool things, like some little, you know, elf years or change some colors of some things. There are lots of ways that you can kind of toy around with things. But for next year, that's what we hope for. Then industry IT, child room IT, and hospital IT. And I recall that we were laughing ourselves dead with that dinosaur that we brought along. I don't know if you remember. It was the kind of toy dinosaur that you could pet and it would sort of hum at you. And it had a camera in it and a microphone and it could make some little noises. And this year, the federal network agency forbid a children's toy and child watches with microphones were also forbidden by the Bundesnetzagentur. And the nicest thing about it, the nicest thing I think was the thing with the Furbies, you guys know that, right? They didn't stay stuck in the class, stuck in the past, they now have FurbyConnex that speak Bluetooth and it doesn't need any authorization. And it plays audio files so you can imagine where this is going. It just means you just need to get close enough to the children's rooms. And we would like an audio track from y'all. Also, if y'all for ten seconds to just chant Cyber. Dankeschön. Thank you. Beautiful. The loop isn't allowed to be any longer than that anyways. Wenn ich dann die Spielzeuglese... So toy shops, when they all start to cyber in there, you'll know why. You know these days, this Furby sort of seems to be the great grandfather of something Alexa. Did you say Alexa? Alexa, tell my therapist that seriously, that's a thing. There are multiple, for competing products as well, multiple implementations of Eliza. The original sort of therapist program and of course our machine learning based therapist software so you can talk with your advertising buddy or whatever. And so if you think about the human spirit is a stationary target. So if you have this kind of mix of therapy session and application in one thing, do you see the possibility? Well, what could go wrong? The cloud pat guys also lost 2.4 million files, sound files of children, many, many users. The next thing was the cloud predictions and the thing is if you sit there and predict the future and want to find out how hard things will be, then we think about things like Rohammer, things like escaping from one VM, going to the other via the hardware, via errors in the main memory. But how does the reality look like? Well, there are ASLR-Bucks where nobody set up security measures. Rohammer, you don't even need Rohammer, just download it and have fun. The biggest data leaks were just terabytes of customer data, just a variety of things. The other thing was the iPhone in 2007 like an iPhone, like a phone where you can execute stuff on it, but nobody is asking for that. So, but now security experts say that iPhones are the best thing for security-wise for the common customer. So, this is a market-driven thing. When exploits are expensive, then everyone has the incentives to keep them high. So, they have incentives to dispense them slowly because it's good for everyone who is sitting on a pile of exploits. It's also a good thing to add this one, too, because it was so good. Biometry, buffer overflow through smiling too wide. If you're, the corner of your mouth move too far out of this and your face, what happens? Is there a buffer overflow or not? MacOS 10, nichts zu sagen über das. I mean, Vista, Vista, Vista, was patched again this year with WannaCry, like that. And MacOS 10. And since that, it had two-factor authorization, username and password. Aber ich meine, die Dinge gehen einem nicht aus. Ja, sie gehen einem überhaupt nicht aus. Outlook hat ein halbes Jahr bei Esmime-Mails, also bei verschlüsselten Mails. These things never stop. So, like, Outlook for a couple years with Esmime-Mails, supposedly encrypted-Mails sent the clear text along with the message. And then the kick in the stomach, is it only happened if the emails weren't HTML. So, like, Mails, how you're supposed to send them, huh? All right, well, it's the normalcy update. Username and password combination for a working mail account costs 8 to 15 dollars. In Germany, we had 83,000 cybercrime cases in 2016 with 51 million euros worth of damages. I'm wondering where these numbers are coming from. I mean, does that include WannaCry and everything? And then the store of the 300 million, whatever, is that included? And are there stolen bitcoins in there? And that's the interesting question. There's this great joke in Hamburg. Ein Hafenbeschlagnahm, dann hieß es, ein Container-Kurzbrot. The police took a container full of cocaine and then they said there were four tons of cocaine, was put into the police of four tons of it. And two weeks later, all two tons of cocaine were destroyed. So, we have something similar here. We had somewhere, we sold between five and five hundred, sort of referring to the volatility of a bitcoin. So, ATMs, manipulation of ATMs, 476 cases, 56% in Berlin, was going on there. So, you see all these ATMs and you just see them and you think these must be scamming devices, right? So, you know these from other countries, but you can't just go around there without seeing like five ATMs. But you really can't resist the urge to do a scamming device. Who took money from the ATM up there? Does someone know someone who took money from an ATM there? At least smell good the money, was it fresh? Oh yeah, 50.000 MongoDBs, no, 27.000? No, just 27.000 in one week in August. It was a massacre. The DDoS attacks that were publicly documented was 510 gigabits. Last year I read something about 600 gigabit and in privately documented things we heard something about 800 gigabits. So, but we say that's progress in the cloud development. So, the typical way of monitoring a service is like you send a mail, nice business you have there. You have this day where you sell a lot of stuff and it would be a tragedy if the service was DDoS during that time. And then let me just demonstrate that to you and then you... Yeah, and so Bitcoin is a suggested way of getting out of it. They give you an address that you can send the money to. And so the traffic is exactly 15 minutes and it's coming from AWS or Azure because we know that people are using the free accounts from these cloud providers to come together, to pull together all of that capacity. And so there's no need to pay up. And so that's the logic of the internet these days. A round of applause from the audience for 100 million certificates from Let's Encrypt. And Encrypted Traffic in general has risen to 50%, if I'm correct, which is a great accomplishment in itself. And the crowd applauds again. Yeah, apps faken. Right, so faking apps is a thing these days. It's a thing now. So if you look at Google Play, in the Apple Store we had that too, with that Ethereum thing. And last year we said that all app reviews are fake, but not just the app ratings. Maybe the apps themselves are fake and maybe there's fake apps with real reviews. Depends on what you can afford. So a million Downs of the Fake WhatsApp Messenger are from Google Play. And then the sobering percentage of the day is that Google Android's security updates only reach about 50% of devices according to their own statistics. If you can believe it, is what I say. And so we come to the next point. Statistics that you haven't faked yourself. There were at least two zero-day studies. One of the interesting numbers was that a zero-day, if it's known as that, has been living for seven years. And so the vulnerability is seven years in the fields and use. And so it was discovered seven years ago and has been used and the manufacturer still doesn't have a patch for it. So this length of life has defined as the length of life ends when the manufacturer issues the update. And I find that absurd if you look at the number of people that still don't patch. The number of wounds that want to cry inflicted and the number of wounds that will still be inflicted because people still don't patch. And so, that Microsoft patches this and that or Google patched this and that or whoever patches this and that, those are headlines where I think that's nothing. Microsoft makes them available and whether they arrive anywhere or they land anywhere is another question. That isn't always one you can answer in the positive. And so, in this business, it causes these mechanisms by which manufacturers roll out their patches on certain days. And it's developed into a kind of routine. Microsoft rolls out a patch and then people open up Ida Pro and look out what kind of bugs were patched. And there's two markets. One is for developing the exploits. And then the other one is something about crashes. So the other reason that people don't patch is because software doesn't work. And so part of the reason you get people to patch is keeping the quality high. And so there's this rumor. And so there's this rumor that the rumor that Apple issues the really nice emojis with the important updates to motivate people to update. Sorry. And so this is of course a crazy conspiracy theory. But if you can find a way to make that work for IoT, then we should be thankful to him. To someone who could figure that out. So Apple has found a way how they can make humans update. But for IoT, I think we can only make that work by fixing the light after the patch. So the light works again after the patch. So maybe you could add a display to every IoT device that could display a smiling cat if you update. So maybe like collectibles. I think this is a good business model. Yeah, like Pokemon patching. We need a gamification of patching. Ja, das war Computer Massacre. The voting machine massacre went on this year. So the best thing of the... The best part is when people start looking for a metaphor for this whole voting computer village thing. It's almost like when you shoot fish in a pond with a Gatling-Gang with exploding bullets. And it's not filled with water, but gasoline. So there was PC Val and what was the name of the other thing? If we elect, yeah, that thing. We still have to wait for it yet. That's not out yet. How did that happen? You know, it was right before the federal election in Germany. Oh, yeah, shadowbrokers. So the phenomenon that we have malware with weaponized exploits where the secret service pay for opening up other people's computers. So we have that anti-e-government because it was the weaponshafts of the NSA that were emptied. Yeah, these were zero days that will accompany us for years. Yeah, these were known for ten years but not to everyone. So shadowbrokers make sure that these kinds of exploits are not used by other people. So the way this is handled when government malware is made public, how antivirus software adds them to their signatures or not, that has concerned us a lot this year. So the clear thing is national... So nation states are starting to see antivirus engines as national assets. ...daten und Dokumente, können alles mögliche sein. Und diese Strategisierung und Politisierung of the anti-malware business. And then there's this problem that attribution is difficult and so you've got the situation where this malware tech blog guy still can't leave the United States but at least he's a free man. What else? I think we should take a closer look at this. People on their Twitter accounts set their region to be somewhere in Germany because then certain people don't bother them anymore because the Twitter censoring mechanisms for the locality of Germany apparently share their tastes more closely because they've gotten rid of more of the Nazis. The place that you choose for your social media account determines the way that certain things are attached and we're missing some documentation there. And hey, maybe all of us want to become virtual citizens of, I don't know, maybe it's better there. Like what you have and when you don't have. We don't know that. Maybe you should reverse engineer that or something. Maybe machine learning could help. Machine learning to the rescue. Apache struts was supposedly the volume that was used with aquifax and then they were popped. It was sort of something like Schufa in the US. One of the big three with about 150 million US people in the database including in addition to ones from other countries. So complete profiles. Celebrate, celebrate. Wie nennt man so was? Das ist ein Dienstleister der Regierung, dabei hilft, analohte Telefone aufzumachen. They are a service provider of the government. That allows the government to open up other people's telephones. So therefore surveillance and not privacy. True data richness. So, and then there were a couple of AWS buckets of stuff. For example the Republicans had lost 198 million data rows. And how much was that? Well, that was everything. All voters. And 1.1 terabyte of data accessible and 24 terabyte data not accessible. Well, was that a very interesting thing? Well, we can compare it to 2015. We had that 2015 already. What has changed? Well, for example, they added the skin color and the political orientation of the people in the database. So we see this was used for targeting and this is the direction we can interpolate that a bit. We see that ransomware also does the crypto right like using really good algorithms but even for ransomware. Crypto stuff is hard. So rather use the reference implementation because every time you do it on your own it goes wrong or set it in another way. If you want crypto security for your data with the keys of other people it's better if these crypto malware with us didn't pay attention in the crypto lectures at the university. Right. So we see a development. The development always goes where the money is. So even two years ago crypto ransomware was a good business model because you can earn a lot of money with it. But there's some funny things like in WannaCry where it didn't really work or a couple of others where the question was whether it was real ransomware or just Viperware who just tried to erase all the data. So this is actually pretty good. So if you have the software on your computer and you don't know if it's ransomware or not but there's no customer support and it doesn't follow the standards of ransomware then you're yeah, it's a bit bad. So it should be there should be a good way to pay and customer support und so there are certain industry standards have developed which as a malware developer you should sort of follow and outside of these industry standards so the non-compliance that you could say are a kind of indication that they are total stimper. Total idiots easy because there's a lot of total idiots out there not everybody is what they're doing but this wonderful topic of customer support and as we all know customers are very annoying and you don't really want to have anything to do with them and so and so many people are sort of throwing themselves at coin miners with such enthusiasm because Computing capacity of is used to mine coins and so you don't have anyone on the phone that you need to explain why you're buying bitcoin and so the operations are much easier in this way ransomware is up to about a third I think a third of malware is ransomware that was a study about the middle of the year August or something and so I'm interested to see how far coin mining as a malware form will do because customer service is so much easier and so there been some all these new coins and how many of those new coins are ransomware friendly because if you think about it ransomware ransomware has certain limitations if you think about mobile phones you can only really use a mine on phones in the evening when they're connected to power and stuff so and it would be nice to design some coins that are malware friendly and so the question is how do they how do they these coins develop so lange sind deine Daten leider verschlossen situations where your data is only encrypted for two weeks while we're using your computer for mining and we want your power too or something from time to time these bitcoins are seized and then we have the question what do you do in those situations as a country you can turn seize things into money at some point so the first question is what's it called the the the state goes to the tax payer and says you should have sold that a lot later so how is this from a legal sense so we've been talking about this a lot in legal circles and of course there aren't a lot of cases of precedent so maybe you treat this like seize stocks so to be sure they apparently checked what about what happens when the when the seize goods are lottery tickets and that turned out to be an analogy that they found to be quite apt also die Antwort die Antwort lautest du dann halt verkaufen so the question is you can sell it gut jetzt müssen wir mal reinhauen okay we got to get into this now interesting things from more notable things from 2017 were these bluetooth broadcom and other radio vulnerabilities you mean the other computer in your telephone exactly this other operating system with the other thing that's also all over because in all computers there's a computer inside that has a computer inside it there's a operating system on it there's one of these things from Intel wait let me see okay I know it's got to come up with it Intel ME anybody that stands for linux embedded minix embedded oh minix embedded or was it the ninix engine okay wir können den auch zu Tode prügen wir können dieser immer die Millenium Edition bluetooth Broadcom es ist interessant wo du eine bestimmte Penetration die Menge bluetooth und Wi-Fi Devices du hast die Vulnerabilität hier und die dritte ist wie nahe sie sind und du musst einen Überlap damit es von Punkt zu Punkt und wenn du so einen Punkt reachst hier hier, ja hier anybody's Phone in their pocket get a little hotter nur einen Satz hier zu nur einen weiteren Satz hier hier ist ein anderer Facebook Schadoprofile Artikel ohne das Keyword Birthday es gibt keine neuen Revelation Facebook hat Schadoprofile und was mich erneut ist, dass es nirgendwo in diesem ganzen Artikel ist es ist ziemlich klar was sie für sie benutzen sie monetisiert monetisiert das Interessante ist das Ding, wo wir News in den Mitteln und wir würden lieber etwas machen als Security aber und so es gibt ein paar Dinge die da sind so wie Sensors being used differently than the Designer of them imagined for instance this Dolphin Attack where those advertising devices were like Alexa were manipulated to do things that they weren't designed to do und suddenly Alexa was a lot of cat supplies so so there's also the thing that these devices react to ultrasound so how long does it take to train a batch to emit the right ultrasound so what kind of attack surface do we have from a parrot so how long does it take to make the furby order the cat litter so the dolphin attack is really strange so there's also the question is how you can get around machine learning you just encode more things into pictures that the human can't see but the machine can so this was my favorite thing this year Ethereum ist such a smart contract system the idea is that you have contracts in front of software and that should be a good idea somehow and over the three years where the currency has over 3.5 billion market cap and people looked at the programming and checked whether you can write something that looks innocent but have bugs in them so the bugs submitted were pretty boring, nothing interesting in them well the public ones yeah the public ones well which kind of crazy person which kind of lunatic would send something in well if the price is only a thousand dollars well you can be a little late the bug bounty has to be similarly high to the thing you can reach with an attack and Ethereum isn't at the slow so the most interesting story was about this janitor who wrote the story on a block platform who tried to get rid of the mess of bad and cheap routers so they can't be used as DDoS devices so there was this discussion how useful it was but that's not the interesting thing so what happens if you let machines do that so what happens if there's an AI who tries to attack vulnerable machines and takes them down we're not that far from that so maybe we should just do that there are automatic exploit kits so we can build that and take down systems that are vulnerable and on the net i think we'll be there soon so this is just a data point there was a star wars botnet with over 350.000 accounts but it didn't really do anything on twitter so this is a really big thing this year so the robot or the AI or the algorithm collection that goes out and attacks systems and also gives patches to the vulnerable machines that it attacks there's an open source toolkit for that from the team shelf fish from the DARPA cyber grant challenge so in the cyber grant challenge the task was to develop a program that attacks other programs and also produces patches for the systems it should protect so this clears up the thing with the cheap routers so death by auto correct was the thing this year there was at least one place where the auto correct in a word processing software in a word processing software it was about an antibiotic that only knew one of the antibiotics in other and corrected that word and so it's kind of stupid because it's important to have the correct antibiotic because there are all these people with various allergies and stuff and so we decided to check into how serious this problem was it seems to go on a little further than that because doctors dictate a lot and so we asked Siri how many people have you killed but she declined to answer I find that very suspicious and car washing streets as well so one of the auto washing machines was tested and the videos were released so the door tried to smash the car that tried to get out and if the people if the people if the door got hold of a human then they could also spray the water directly at him so let's look forward so in the last 7 minutes that will be hard the center quality problem it's good if the cameras are getting better and better I think a few years back we had an app that allows to see the breath with the camera with the selfie camera that can also take the the pulse and it can see that at the cheeks and the refresh rate of the cameras gets higher and higher and we're recording micro expressions we see whether people are nervous feel good or bad a lot more is possible and it won't stop so the result is something like the thing with ultrasound and add bugs so you don't know what they react to to so adverse input will become more interesting so things that can damage you and can be produced by a computer will become more common so and they'll be controlled by a computer this leads us to the topic of what is built into this device how was this device built for the purpose that the thing is being used for now so there's a nice saying we're on the on the shoulders of giants so if you use the scripted library here we use this library over there we can use it to determine the number of centimeters to the pedestrian in front of us but if we include this library we need to have internet access because the phone is being downloaded from google for the temperature display it's displaying the temperature but we can now build such a temperature display in under 10.000 lines of code so we do stand on the shoulders of giants and can look further than others so there's the saying don't roll your own crypto use a library that's good but who patches it later on and that's the problem the giants aren't they're not healthy they only have one eye maybe and they have clutches I think we're missing the metaphors here we're not we're not standing on the shoulders of giants but of car houses I think we're standing on a giant dump now the bad story comes the bad news that's the best case scenario we've seen some great examples in this conference of what can happen giants, they're too expensive, they're too big and there's too much and the giant isn't even from here I decline giants for religious reasons I want to be a giant too but you're only one meter and 30 centimeters that's discriminating so if you get the idea you can implement yourself because how hard can that be implementing a payment system how can that be you could use classic for that it works great and classic means that it's a classic classics are always good this problem of what happens when we start with machine learning and releasing it on people has been sort of following the whole thing since the keynote of Charles Charles and there are these automatically generated child videos designed for children and we haven't seen the end of this when you sort of think about the technology technology components of here and we've got pretty good text to speech these days and pretty good speech to text and so we have a phone call the first time it happens will probably be an advertising call and someone from the audience says that it has happened before already and he asks ok did it have an understanding for you I think that will fail because no one uses their phones for making phone calls All this data is the oil of the 21st century, whatever they say, that was sort of the kind of things that we put up there last year. And so what was so great about oil, you know? I mean, we've been fighting wars about it, we sort of basically extinguished our planet over all of it, you know, the jokes right themselves. Data are the oil of the 21st century, and I think about the wars, obviously, yeah. We're gonna like, you know, fill the ocean with trash over it. I mean, it's the equivalent of AWS buckets. So Hava wrote something nice recently. Data isn't really the oil of the 21st century. They're more the tap with burning gasoline in front of it. And you can warm yourself at it a little bit, but other than that, it's basically just got disadvantages. No advantages other than that one. You can sort of go a little further with this. If data is the new oil, then data is the new snake oil too. And then now we're back to the topic, statistics is hard. And so what you can read out from data, you have to be able to get it. And so if you can't do that, then you can sell all sorts of things with data. And then what's sugar in the tank in this metaphorical sequence. Yeah, so we've got a little more here. Yeah, social media facelifting. It happens these days still. There are professionals that will sort of pretty up your Instagram account and make sure that you look beautiful and beloved. And so alternative IDs are becoming more and more important because there are countries where you have difficulties doing certain things without a Facebook ID. And so you need a sanitized Facebook Profile for that. And for those that don't want to buy it, there's an app to help them manage their social score. It would be best if that was driven by artificial intelligence. Ich bin total into the story. And the social score is being used to check whether you get credit or not. And a lot of that should go into that. For example, stuff from the government. If you didn't pay the parking ticket, that's a lot worse than if you didn't pay your phone bill. Also, I heard that health data should also go into this. So, you ask yourself, what will be the consequences? Definitely a chocolate black market. So, angesichts der Arbeitszahl an der Mediensektor. So, for the split of work in the malware department. Yeah, the value generation, the supply chain, value chain. Something like that. And so everyone in this value chain gets a little bit of a kickback, gets a little bit. There's the compliance and the malware industries. They all say that this Automization is not so bad, because there will be new jobs. But we're seeing all these new things and new jobs like compliance auditor in the malware sector. So, new jobs being created by Automization. And then Internet of Things, Ghostbusters. The IoT Ghostbusters. Last year we had the IoT Wunschelrutengänge. It's someone who finds the IoT devices that are even there. And so the Ghostbuster is like the movie, the one who catches them in a way that you can then analyze them afterwards. If there's an IoT device that's been infected with malware and you want to be able to sort of figure out the nasty stuff that's on it, then you can't sort of unplug it, because it's gone. All the data is gone. Nicht die Strahlenkreuze. So über 2,330 Volt, linke Leitung, rechte Leitung. So wie Strahlenkreuzen, nicht Strahlenkreuzen. 5 Volt, 12 Volt, Power. You have to sort of be clever about reading the RAM of the device, the memory of the device. Jobs gut, wo man klein anfangen kann. So, these are used jobs, needed jobs. Sorry, I need to take away your production hall. But I have to keep it powered because I need the malware sample. Oh ja, Quantum computers in Silicon. Quantum stuff, I think we can see progress in this. Anders, we have problems that we see today and don't know how to solve them. But if we can solve these, we might be able to build quantum computers with the traditional manufacturers or something that looks like and feels like a quantum computer. Let's see. So, we asked around what will go wrong next year. For example, in cryptography, where will be the neck apocalyptic. So, the things we heard were Bleichenbacher or more bad randomness and more bad randomness and hardware. Like Infineon and Estonia. So, what's new? Nothing. The old stuff isn't empty yet. Isn't used up yet. Applaus. Well, and we have Cloud Exorcism has just appeared on the screen. Cloud Exorcism is something. In the last years, we need a bucket for devices for these, you know, Alexis is a nice piece of hardware with the microphone and chips, but it has the problem of, you know, being evil. So, you can get rid of its bed, so we have something, because we don't know yet whether you can give food to family with this. Watch the talk where you can disassemble the vacuum cleaner. And where you can see that the vacuum cleaner has better securities than some financial banks. We also asked ourselves how do you get rid of the cloud in the vacuum cleaner. We'll get more of that. And if we did that already, maybe we can get a business model. So you can go there and give them your vacuum cleaner. And they get rid of the cloud. And I don't know yet whether they'll visit your homes or whether you buy them from there or via remote access. Or maybe little prostheses, prosthetics. Yeah, yeah, so there's something to it, to that. And the last thing, quite a few Bitcoins out there where people say, you know, I think I've heard this about five times, you know, I've got a few sitting around there somewhere too, but like, you know, the drive's dead or I don't know where the drive is. And you know, or there are people who say, I can't remember what the password is. So what we need is unlike a hypnotizer, which is unlike a lawyer as an actual profession. And there's not just a predefined wage that they earn. Yeah, I think that's it. We wish you a nice way home. Happy New Year, 1948. 1984. Yeah, very happy 1984 to you from the stage and from Austin the Translation booth. This has been...