EP 013 Global AI Influencer Gerhard Furter, Talks Chat GPT and AI - Can Humans Harness The Power Without Harm

Named 3rd in Top 10 IFSEC Global influencers in AI for 2022. Gerhard says, 'I'm paid to play with ideas, toys and possibilities all day while flipping the bird at the industry's concept of 'impossible'. He is the founder and CTO of Iris AL (Pty) Ltd. Specialising in Artificial Intelligence (AI), Machine Learning, Machine Vision, Deep Learning and Software Development. He is an innovator and problem solver, and he is passionate about Science and Technology, Animal Welfare, Children, Civil Rights and Social Action, Economic Empowerment, Education, Environment and Human Rights.

Founder and CTO Website https://irisai.co.za

Linked In https://www.linkedin.com/in/gerhard-furter-6360aa10/

Email gerhard@irisai.co.za

Twitter https://twitter.com/GerhardSugarpl1

Security Circle ⭕️  is an IFPOD production for IFPO the International Foundation of Protection Officers

Transcript

0:15

Hi, this is Yolanda. Welcome. This is an Ifpo production for Ifpo, the very first security podcast called Security Circle. Ifpo is the International Foundation for Protection Officers, and with me is a man whom I find most intriguing, named third in the top 10 IFEC Global influencers in AI for 2022, a subject matter expert, deep learning machine vision and artificial intelligence based in South Africa. Here today to talk basics around artificial intelligence and what it means to the security industry broadly. Welcome gal. Further, how are you? How's the weather where you are? The weather is actually excellent. It's, uh, it's now becoming autumn in South Africa, so we have, sunshine. We had a heat wave, we have sunshine and rain, but it's beautiful outside. Oh, good. I think it's important to have a good view, isn't it, when you're working. But look, we are here today to talk about some of the, the new, phraseology hitting our general, day-to-day working lives. Let's talk about, chat G P T. Now for those who haven't heard of it, I was quite late to the party I thought on this one., and, and I recognize that it's a artificial intelligence. it's a free source. I know you are gonna go into a little bit more detail, but it's now hitting LinkedIn on a regular basis. It's, coming up in conversations, tell us what's so special about chat G P T and why is it gonna be a huge impact for us going forward? Gohar. Okay, so chat b t is effectively a, a multipurpose ai. So there's all kinds of fancy terms, regarding deep learning bots uh, self-learning, cognitive machines and that kind of thing. But effectively what it comes down to, chat, b t is an AI that has mastered the ability to learn and is capable of communicating with humans in a fashion that is very coherent and easy to understand. Pretty much like the Alexa that you used to shout it in the background of your house, and yeah, you've heard me, haven't you, Yes. I So sorry for Alexa, but I'm really kind. I thank her all the time. I'm like, have you ever Thank you. Thank Alexa. She sings at you. I. I picture Alex at a quarter. They just quietly, oh, she's gonna ask me something. She's gonna ask me something. Yes, usually it's to turn the radio up or down. Chatty leverages the ability for machines to now communicate with humans as a human and to understand humans, not only in key phrases, but also in something called intent. So chat. BT not only understands the words you're using, but it also understands what it is that you're actually trying to ask, and that is extremely powerful. The race of the AI is effectively a massive learning bot that has been hooked up to the internet and is capable of using any knowledge source and any knowledge, so-called dataset that is present on the internet to answer and. What we call pilots in the industry, but effectively produce an answer or some kind of something that you've asked for. And the fact that it's also been, made like a sort of free source so people can use it at well, has just made it just that much more acceptable now to the, the average humans out there. How, how are people using chat? G p T and I mean, I basically did a refresher. I, I just know how to rock. In my spare time. I did a refresher in the Data Protection Act. Love it. And, in some of my questions, I thought I would put chat g p T to the test. And so I asked it a very complex question. tell me how I should be explaining the Data Protection Act to my friends who dunno anything about it. And it literally came back with two, three paragraphs on, well, you know, the Data Protection Act and it's very. Human in the way it communicates back. Now I can see students using this a lot for exam questions, and I don't know at the moment that that's gonna be acceptable. And I put it to the test and I asked it whether I should eat that piece of cake. And it knows it's limitations, doesn't it? It says. I'm an artificial intelligence. I can't advise you whether to use eat that cake or not And I thought, oh, crikey, that's useless. I'll have to ask Alexa, but in re in realizing its limitations, is that just because it hasn't been taught yet, how to answer a question about whether, you know, someone should eat cake or not? Ah, but that's the key concept. That's the thing. It's not being taught. It's. So now what happens is you have, let's take your cake scenario. That's a very good use case. We have a hundred thousand users out there asking the chat chip PT engine, uh, should I be eating this cake? And sometimes it will produce because it's quite a, it's, I'm gonna say random, but random is not a good word. because the communication is very personalized. The, the conversation you and I will have with chat will be very different from the wonderful instance my wife would have. And because of that you get very diffused or differentiated, interactions. So it learns differently from different people. Now, let's say it becomes an important aspect of your life. We have to know if we have to eat cake, it'll take the interactions with these a hundred thousand users and then learn what their best answers or based opinions, and this is the key opinions about eating caers. And it would then assimilate that to acknowledge it and be able to extrap. Future answers and future's responses, from that. Now the really cool part as well is not only is it dependent on humans, that's as a matter of fact, very little dependent on humans to learn. It can go and it can plumb the internet, uh, for nutritional information regarding cakes. So what you'll probably find is that at some time in the future, that's gonna come back and say, yeah, no, no, you should have a piece of cake. But you know what? Because of health reasons and because you're a good human and a carrot cake with some ice crossing or something. And, uh, that is literally what chat p t represents is evolutionary ai. Unfitted and I'm, I'm restrained, uh, trying to learn from us, but at the same time having access to the greatest darker source on earth. You did make me laugh when you said your wife would use chat g t in a different way to you. And, my mind went back to when Siri came out and we all kept asking Siri where we could bury a dead body. And Siri And Siri would come back. This, honestly, I'm an ex-police officer. This is the kind of wares of humor we have. And Siri used to come back and, and literally give you a list of swamps, woodlands, marshes, open hits, quarries, obviously tongue in cheek. We'll, I know, I mean, can you imagine, if murderers were using Siri to find out where to bury dead bodies, I think you just gotta watch an episode of Dexter for that. But imagine where chat G P D is gonna go next. I mean, is it gonna develop personality, do you think? Is it gonna be quirky like Alexa can be so. Can emulate personalities, it can emulate response patterns because we respond to personalities. if you, if the bot was to communicate with us like a third three year old child, we would probably ignore the result coming out of the bot. But currently it's communicating, as if it's an adult, without being sexist about this in any way, you sense a certain male angle to the onset. and also it's got a. Feeling of a, a guy around about 30 to 35. So feeling mature and schooled in whatever the daughter said or whatever knowledge that you're talking about. but it's an emulated personality. It's just a response button. It finds the right words based on your words, analyzes how you speak, how you communicate, and then responds in a way that would be the most. Effective based on your speech patterns. So in a certain sense, I'm gonna use a very negative concept here, but, in a positive way, that's possible in a certain way. Is a psychopath. It's capable of adjusting its personality to best utilize the current resource that's communicating with, yeah. To me that is one of the greatest aspects of the AI is that they manage to do that. It is going to become much more prevalent, much more relevant in the world when, we start looking at taking the AIS out of computers and putting them in bodies, which is coming very, very soon. And we will suddenly have, I mean, there's a huge, there's a world of study out there. I've seen some of the results. It's just fantastic with people have whole university faculties have been created just to make machines more acceptable to. facial expressions, body language. we found instance that for some reason all of us have this big thing about anime eyes, you know, so you'll see that all the cute little robots have big blue eyes and they have these little. but they also have fairly bland faces. We not want them to be too human. Cuz then they They creep us up. Okay? Mm-hmm. But we want this little balance between a cute little bot, looks like Wally, but yes. Still capable of communicating like it's your next door neighbor. Yeah. And that is where chat G P T will bridge the gap. they're making, making communication much more human. it also is, if you think about it, it is the perfect application of the internet. it is immeasurable data just accessible at your fingertips. I think humanity's gonna look back at the history of technology and chat. C PT is gonna be one of our milestones. Elon Musk landing a rocket on the sea. Yeah. Mmk. That was a, that was a milestone moment in, in, in our scientific history. The next one after that was the whole space station thing that they managed to get it working and now it's church. The fact that this exists, it's a m. So look on a personal level Yeah. I do genuinely be nice to, I am nice to Alexa because I think somewhere in the future we are gonna build credit or lose credit on whether we are mean to technology or not. So for example, uh, you talked about how. chat, G p t is male or its responses seem to be very male. I, have a qualification in gender inclusive approaches in, in technology, and it's interesting because a lot of new technologies aren't designed with inclusivity in mind. And, and with that, you, you know, if it is, uh, designed by a man and the, the responses are male, for example, then that means, you know, it's missing out 51% of the population. There was an episode of, uh, dark. you know, where, you know, you literally build credit up on whether you are nice to people or not. And, and I've heard, you know, responses around how people do generally treat Alexa, they generally shout at it. And because the parents shout at it, the children shout at it. And I don't find, I know it's automated, but I don't find it very courteous. I'm thinking, What happens when our children are learning to shout at technology and disrespect it and not appreciate it? Now, I dunno if I'm being over sentimental, but if you looked at this qualification, I did the amount of responses, globally for people calling Alexa a bitch. Were recorded and it was phenomenally high, alarmingly high. So I'm thinking, if you've seen the Orville, and I dunno if you've seen the Orville, you'll click onto what I'm saying straight away, but there's this Yes, I do. It's, oh, it's, we are Trek, we already preambled that, uh, before, but the Orville was very clever in relation to a specific species that was an. And, and how that species became an enemy is very relevant, isn't it? To how we treat technology in its infancy. Yes. Especially AI in its infancy. Isaac. Yeah, Isaac and I do worry. Is it a prophecy that in the future that that artificial intelligence could become so resentful of literally being a slave to a human. That it decides to take its own, mind into things. What's your view on that? I mean, I don't expect you to have the answer, No, no. It's ok. I, I understand. I've actually thought about that in the past because, some of Theis that we have patented in my, in my teams, have the ability to be weaponized. And we have a principle against that. We don't wanna weaponize our iis and. But then at the same time, you are actually doing exactly what we are currently being programmed to do by the Lords technology companies. We're starting to humanize, an anthrop, what's the Anth? Anthrop Morphs, what's the word? English word. I'm, so sometimes the English goes out the door. Yeah.. Yeah. Humanize you. Change the technology into something human and give it a human attribute. Alexa does not have a person. She has a series of emulations and harmonics in her voice, uh, prebuilt algorithms that makes us sound happy or sad or whatever emotion is they need to in imbue. but this unfortunately means that we are, as humans, we program to respond to each other, specifically with non-verbals and with subtle little things. Tone in your voice, you know, when you are. partner speaks to you in a specific way, you pick up, Ooh, she's saying she's happy, but Oh, I'm in for it. I'm gonna suffer for this one. You know, sounds, sounds like personal experience there. Yeah. Oh, definitely. But, uh, the big thing is that, uh, we've been programmed to respond to those things and the technology companies know this. I'm good, honest. We utilize it ourselves. Uh, and Amanda, one of our technologies, our little chatbot we use to interact with clients. Uh, we choose, we chose the language. Uh, even the voice. She's got a, a Scottish, uh, female voice, which is a very Nice voice. It sounds nice to people. And we chose these things because it makes it easier to listen to it. People tend to not ignore her when she speaks because of those attributes. Also the words that she use. Uh, non-combat, non uh, aggressive words. Even though you're talking about something like an alarm going, oh, for a bomb exploding, you still use words that do not incur or do not in some. imbue the conversation with an aggressive, uh, tone, cuz that makes it acceptable for humans. Mm-hmm. So the computer people have done that, they've mastered it with chat. They have, but I must be honest the personality part of it. We are, we are very far away from getting computers to have emotions. We also have, there's a huge debate. I actually accidentally stepped into it with a white paper or article I did on someone on LinkedIn a while ago where I said, but you know, at which stage do we decide that machines of souls, because they are now emulating the things that we as humans have always considered to be our sole property, the things that make us human. Have you seen some of the art. That some of these devices, some of these AI are producing. It is staggering. My two B daughter-in-law is an artist. She's a painter, and she's seen some of these things, and it actually drove her to tears, which she said she cannot paint like this. She's a human. She paints from a soul. This machine paints using an algorithm and it, it completely defeats her talent, defeats her ability. I've seen some of the music, somebody ambushed me into a very bad conversation where they played me at piece of heavy metal music. I'm a metal fan, those kind of guys, and the song that played was fantastic. I thought it was metall. It was an ai, it emulated Metallica. Wow. So the things we do now that, that we consider to be personalities and the things that make these machines human to us. We've become so good at lying to ourselves and emulating these things that they've actually now become a problem We now no longer can understand. Does this machine have a personality? Is this per this machine actually feeling an emotion? She's not outside of you. You threatening her with the solving iron. Alexa is not gonna give a damn about how you treat her. But what's important is it's not Alexa that's important, it's you, you are important. How you respond and how you feel about the response about yourself when doing so. That's important. It's similar to when we are really brutally honest. I'm just, I'm a big animal lover. My dog's lying right next to me. Scooby do right now. He's my life. I actually love my dog, but in all honestly, truly clinically lethally on the on the disc. If I was to shoot Scooby-Doo now, it would've not change my life. Okay. I would die as a human, but it won't make me less rich or take food away. It will take my ass away from me because in this grand scheme of things, Kube doesn't really have an impact on my life. Mm-hmm. But as an emotional being, it does. And that's where how we treat machines well in future. Still have an impact. We will still scold our kids for shouting at Alexa, but not because Alexa has an emotion, because we don't feel good about ourselves when we do those things. And for a couple of generations still that's gonna be the case. I cannot see how we are going to give machines emotion. We've all seen what happens when data puts his emotion chip in and it doesn't quite go on. I knew it. I knew you were gonna bring that up. I knew it. I was sitting here thinking, shit's gonna bring up data. He's gonna bring up data. And, and when you look at the evolution of data in the Star Trek journey, you know, they were very real about his emotion chip being experimental. And the, the best scene I love is the one where he's, he's in. They're in the bar. I think guy's there, and he's having a sip of something and he just pulls this very strained expression and he was asked, you know, I remember, yeah, he was. He was asked. He was, maybe you don't like that. And he's going, yes, that's it. I hate it. Give me some more. It's so Experience, more experience, more experience now. That's true. Yeah. And it's actually interesting that you bring that up because that's very close to our. Chat G B T is now and wants to experience and that's why they're being very clear. They're making it open. Every single human speaking to the system right now, every little sketch coming out of there, every little piece of code, every little algorithm, just a piece of homework finished on whatever, every piece of that data set is enhancing the ai cause we accept it. We, we, the best possible form of reinforced learning out there. Every time you say thank you or say yes, this is cool, you accept the answer that's been given. It teaches chap and say, oh no, listen, yes, what I just did was correct. Reinforce, reinforce, reinforce. It's genius. Thinking about getting yourself qualified. Try if pose CPO qualification. CPO stands for certified protection officer. The course includes Security risk management, effective communications. Anti-terrorism and VIP protection, security awareness. Crime and incident scene procedures, automation in protection operations. Interviewing and statements, legal aspects of security, ethics, and professionalism. And many, many more. Signup to if post CPO certified protection officer. Contact us now. At if Poe. Dot org. So let's talk about the vulnerabilities. We're gonna talk about exploitation. Yeah, yeah. We're gonna talk about the abuse of ai, which is already there. I think, journalistically and quite, very simplistically, everyone thinks that everything's going to be a situation of Terminator and Skynet going online and. Being able to defend ourselves. But I think, when we had our early discussion, I said to you, didn't I? That I think it's a shame because security by design is, is very much behind the evolution of technology. And we all wish in the security industry that there was more security built in at the design stage. And it isn. and therefore the security industry in general, including cybersecurity, will always be on the back foot and catching up. Now, I think in my simplistic way of looking at things in life, that there should be the Robocop law, you know, that AI should be, should be designed. I still love that term. Yeah, It should be designed to never harm a human, but when you consider. That AI is being used in military warfare and the future of military warfare. I can't see that happening in you. Now humans are, we hunt us. We predators by nature. It's, it's how we evolved. Unfortunately, that IBUs us with, an aggressor mindset and aggressive response to situations, the whole sciatic response, fight or flight. But we, we've been pre-programmed to actually be fight. We've done it by changing our world, taking control of our environment, making sure that it fits us, and then evolving into a creature that actually now, creates evolution instead of. So the problem with things like chat is, is if you go back, I think I used the image, the other name when we spoke, about the the nuclear bomb. Yeah. In the guys parked there and, put a little, what's his name? Opim. Was it? Yeah. Said that you know, I have become death to the destroy of worlds. He was quoting from the Indian script and when he realized that. Yeah. They did it for the right reasons. They did it because they needed to protect their own people, their country, but he understands enough about the human nature to know that they were going to abuse it. Yeah. Which they did. I mean, they did. Yeah. Short on the war by five years. Wawa. Yes. Okay. The point is though, Is even worse than that because it's easy to control somebody's access to plutonium and you need some knowledge to create those bombs. With chat. You just have to be able to tell it something in English, whatever language you wanna use, and it will give you back what you want. And in a certain sense, chat is very much like Sheldon Cooper. If I can quote the person from Big Bang Theory, I'm a huge Big Bang Theory fan, obviously. Oh, right, yep. We know who you meet. Yeah, definitely. And, since you two put on the glasses, I keep thinking Amy, but anyways, so Sheldon is tremendously gifted and intelligent, but tremendously naive. So there have been times in the series, and I'm using this as a, as a, an image that I hope a lot of people can relate to. There's been times in the series where the other guys exploited Sheldon's genius by convincing him that what he was doing was for the greater good. That actually he used that term once. Yeah. Or that it was something that will help them or benefit them or himself in some way. So chat is the same. They've created really good defense. Me. There's some good stuff in there, defense of, uh, you know, personal information and that kind of thing is built in. But if you're just a little bit clever, a little bit sneaky you can exploit it. I did it the other day. I made it bolt me the different blocks, seven different blocks that I put together myself in, uh, in a project. And I created a virus. Virus that can destroy a hard. And I did this by just asking the right questions, taking the blocks and compiling them myself. Now with a chat created virus directly or was used to create a virus, the result is the same. And I have a virus that I did not write that I can use to destroy a hard drive on a computer now. And I'm not, I wasn't even very. Clever about it. I wasn't being an AI engineer. I just tried to do it like a teenage boy would do. I literally used teenage language and I got what I wanted. Now that is the scary part. That is the part that we need to really think about because like any powerful thing out there, anything was created for good. It is possible for us, for humans to exploit it and use it for evil, and that is the part that we need to think about a little bit more. Well, you gave an analogy when we first spoke that, you know, it's like we, the handgun is, is not the weapon, it's the person's using. That's the handguns, that's the weapon. And I think you made the same parallelism that AI is, is the same. Yes yeah, that's perfectly correct. Handgun's perfectly correct. Or the rock. Yes. You've Point a gun at somebody to defend your family, or you're pointing at somebody because you're robbing your store. Mm-hmm. The end result is not the gun being evil, it's you, the person using it That's evil. And unfortunately as altruistic as the guys from Open AI and so on, or they're trusting our species too much, we are not a nice species. We have the power to tremendous good, and we've really done so. But, uh, we can do horrible things to each other as well. And, uh, in that sense, uh, chat is a too powerful thing to just give to humans to use. I have fully respect and recognize the mechanisms they put in place to try and use them, protect, protect it, but, There are far too many people out there who are very clever and will find ways to use it if they haven't already done so. And, uh, chat will be abused, AI will be abused, and that's where. Amongst all the little things. There's all these little tering mechanisms out there where all these companies like, uh, Elon and like, uh, open AI and Google and even Microsoft in these guys are all shouting and saying, listen, when need to start regulating ai, you start controlling AI now as a little AI company with some reputation here in South Africa, South Africans don't like controls. You wanna really see as annoyed put a put a leash on us. Yeah, we'll kill that. We'll bite and fight that leash. But this, once I agree, we should be leashing our technologies. We should be putting something in place to prevent abuse. Cause I am scared of, and I'm mean this, and I'm not being dramatic. I always get so annoyed about these people going on a Terminator and Skynet and you know, all this bullshit. But this is the one time, there's two technologies in this world that frightens me because of it's abuse. One is nanotechnology, the other one is Chay pt. And I'm scared of what happens when they when they actually combine the two to create something. Well, you, you, you're describing a kind of a Borg situation to me. Yeah. Pretty much. Right. But um, yeah, it sounds very sci-fi. I, I acknowledge that and as an engineer. I'll probably get scolded by my team who are all sitting here at the back going all pale facing. Oh yeah, he went there. He freaking went there. Yes, But the fact is that yes, we need to control technology like this. The same surgeon that saves your life by doing a heart transplant can cut your throat with the same scalpel. Why doesn't he do it? Because of the nature of the person doing the testing, he has a do no harm ethics, right? So we have to find a way to make sure the right people accesses the right technology. And that becomes even worse. Cause now we're on animal farm, you know, some, uh, some animals are more equal than others. So, where do you draw the line? How do you determine who's capable? Who's responsible enough to access a technology like chat or any of the other ais that are coming? Because you know, we, in the middle of a industrial revolution, chat's the first, just the first public one. The rest are coming. They're coming. Give it six months. Well, you be inundated with alternatives to check, wouldn't you call it a technological revolution rather than an industrial one? Which happened in the 19 hundreds, didn't it? This is the technical revolution. This is the, yeah. I actually prefer technology revolution. I like yours much for much more. This fourth IR thing. It's, it's so. Cliche, you know, the people all use the same term and you find it on the pamphlets, and every time I go to some conference, somebody's throwing it in my face. I don't agree with the industrial revolution. It's a technology revolution. You're perfectly great. Yeah, This time it's too accessible. It's not just outside, outside in the, in the production world or the financial world. This time it's in your house. The Alexa that's gonna switch on your light is part of that. So it's in our faces and which means we have to think differently now. In, in my network, I deal with a lot of people who deal with crisis management, for example. It's their job to manage businesses to have, ensure that they have the right crisis management plan. Because if you have a plan yeah, you never have a crisis. Yeah. And the, the most disrupting and disturbing thing from, from years and you've only gotta look at New Orleans, you've gotta look at for Kashima and Japan and for example, these crises happen because no one could foresee. It would happen and invest in any kind of mitigation to stop it happening. There was no plan. So, the kind of harbingers of doom, the, the risk managers, for example, they'd all say, well, this could happen. But then it was a case of, you know, the airplane industry, for example, it only got better and safer after plane crashes happened, and they learned from those mistakes. So when you look at the, and you can see where I'm going here in relation to AI and the technological revolution. I, I remember seeing this cartoon, and it was in a daily newspaper. It happened years ago, and, and it was so, it had such a lasting impact. It was about two aliens that are looking at earth. And one said to the other, NA, I don't think they're ready yet. And the other alien says, why? Why not? And the first alien, number one says, well, they're still pointing weapons at each other. And then I thought, how true This. If we can't, that's brilliant. Right? Right. That's brilliant. Right. This is, this is why I clearly maintain aliens will not make themselves known to us because we're still pointing weapons at each other. And, and it's that whole broad scheme, isn't it? Around if we can't get people around the table and make a clear policy around global, safety for ai, for example, what hope have we got? That's, that's my biggest. I'm with you on that. I agree. I agree with that. I think that that there's a, a huge potential, not just the now sound like one of the marketing guys, you know, a huge potential in ai, wa, wa. The fact is we haven't even scratched the surface yet. What these things can do. I mean, the advances in medical research now, the technology that identifying cancers and growth. So my minuscule that a human cannot see them and they're doing it effortlessly, just going, oh, there, there's a cancer there. Saving the guy's life. Done. It's tremendous. It's huge. My kids I've got a twin that's 15, they now use AI to teach these kids. They've got tablets and they've got this little nice Google app thingy and they use AI to teach them language. And, and certain other concepts. And unfortunately being my kids, they are not being very kind to these tablets. They've tried to hack them twice or three times now to get to the ai, but the fact still remains that there's a lot of stuff these things can do. I've seen the art and the music and the poetry coming from these entities and just that tells me, oh, this is gonna be good stuff. Let's just be responsible. that's all just limit access. And even then, that still is horrible cuz I'm taking freedom away from people that should be having it. Because one of my things that I propose that, I say that AI should be, uh, a big, this technology, everybody should have access to it. Mm-hmm. But then how do we control that? And it's a very difficult question and we're not gonna answer it. And unfortunately, the bad thing about humans, we have a tendency to, if we can't answer a question, we get bored with a question and we just run with it. And that's when the pop posts gonna hit the. It, it's, it's like anything, isn't it? Uh, the, there's a lot of things. boost the economy and crime is one of them. And I, I've, I fear is a good marketing campaign. Yeah. I think like with all things, I think the bridge will break before they realize the bridge needs to be stronger. And, and I can probably hear a lot of people resonating with, that's just the way it's always been, even though we wanna make the world a better place. But look, we have wow. How to make half an hour ago super fragile cast. B Doha sleeps fast. It's been a real pleasure, Gerhard, and I know that we had a meeting of mines because we are both deep trackies and, it's been really good talking to you about ai. Any lasting, uh, footnote for our listeners. Yeah, I think that your image earlier of data is an excellent one. But it should be expanded upon in the sense that the thing that I like the most about data is not just data himself, but also how the people around data experiences it. And that is something to, to you hold onto is that when AI really becomes part of our life and it's almost there, I think it's, if this was a progress part, we are looking at last little doctor just doesn't, doesn't want to go and. When AI becomes a pervasive part of our lives it is actually going to enrich us. It's gonna give us new things and new abilities we can do. It's going to be fun to live in a world with AI as long as the people who control it, who has the keys to the castle do it responsibly. I think we're already fairly dependent because we, here in the UK, we had an issue where Alexa went down. There was a denial of service and oh lord. You know, the newspaper stories were phenomenal, like people were, I remember for that stuff, people were telling Alexa to do stuff and they weren't, and they were just losing their rag with her. And I'm thinking, oh my goodness, we're already, where's the light switch? Where's the light switch? I know. Yeah. But listen. Thank you Gohar for getting in touch and for taking the time. It's been such a cool chat around AI and, stay in touch with us here at IF Podd this has been an if Podd production for Ifpo. Check us out on www.ifpodd.org for a list of all our podcasts. Thank you so much. Okay guys. Stay well. Thanks Yolanda. Bye-bye.