Habib [00:00:00] We should not limit people from using A.I.. Instead, we should train them.
Chris [00:00:04] We want to be able to leave a good impact. Our end goal is to be one of the guardians or the good people in the industry that are actually there to help.
Patricia [00:00:30] Welcome to the AI Purity Podcast, where we delve deep into the fascinating world of AI detection technology. Today, we are thrilled to unveil the secrets of AI purity, a groundbreaking tool that not only identifies AI generated text, but also uncovers the hidden truths, even when it’s been masked by paraphrasing. Our journey begins with an exclusive look behind the scenes as we sit down with the brilliant minds behind AI purity. Get ready to explore the future of text analysis. Unmasking the AI one sentence at a time. Today we are joined by AI Purity’s Chief Technology Officer and Chief Operations Officer, Habib And Chris, to give our viewers a little background. Our CEO, Chris, is a man of many talents. He is also a music producer and is signed to Universal Music Group and Def Jam recording. And our CTO. Habib, graduated summa cum laude and has taken master’s degree programs in computational linguistics and computer science. Together, they are the brilliant minds behind AI purity and we are so glad to have them on today’s podcast. How are you guys doing?
Chris [00:01:33] Awesome, Patricia. I’m very happy to be here. Thank you so much, Patricia.
Patricia [00:01:36] Hey, Chris, how about you, Habib? How are you feeling today?
Habib [00:01:40] Hello, Hello, Hello, Patricia. Hello, Chris. Thanks for having me and thanks for the introduction. That was great. Thank you.
Patricia [00:01:46] Thank you. We are so glad to have you today. And we are here to tell everyone about the great launch of AI Purity coming soon. So if you guys would just like to introduce yourselves to our viewers, we can start with Chris. Could you tell us a little bit about yourself and your role at AI Purity?
Chris [00:02:03] yeah, for sure. Yeah, I know. I love to talk about a bit more about what I do for AI purity. You covered pretty much everything in your intro, but just to share a bit more. Yeah. So I am the CEO for the company so far. I manage all our projects and pretty much make sure our day to day goes as well as we want it to be. As a bit of a professional background, I am. I graduated in marketing with a focus on psychology, so I like project management. I like working with people, communicate communication especially. It’s a very big trait of mine. So yeah, that’s a bit about me. Go ahead.
Habib [00:02:46] Harvey Okay. Thank you, Chris And I would like just give you a little bit of information about my background. I graduated from Sharif University of Technology, Iran, and got my master’s degree there in computational linguistics. And then I attended another master’s degree in computer science at university in Mexico City. And I got my second master’s degree in computer science there, but with a focus on natural language processing and machine learning. And now I’m taking my PhD in computational linguistics at the University of Calgary in Montreal. In Montreal, Canada. And about My– yes, thank you. About my activities in AI Purity. I developed the models where I pretty much work based on machine learning for detecting generated texts and AI paraphrased text. And mainly my activities were just developing these models. But except for that, I had a very great big cooperation with Chris and other staff. And maybe next year you can talk about those very, very great things we do together and it was a great experience.
Patricia [00:04:02] I’m so glad to hear that you guys have a great experience working together and really have you guys to thank for the amazing platform that we are sharing with everyone very, very soon. So Chris, I always like to talk to ask this to some of the entrepreneurs that I interview. We always talk about the importance of having a really good team behind you, especially when you’re running a business or like you overseeing the operation. So could you tell us about how you found Habib.
Chris [00:04:30] Yeah, for sure. I’d love to talk about that. Now, honestly, I feel I’ve got asked this quite a bit now, but I keep telling people I feel like it was like a match made, a perfect puzzle piece when we found Habib for the first time. So just to give you guys a bit of context, we came up with the idea right after Chat GPT launched. Then we felt there was a space and a need for something like this, something like our tool. And so, yeah, from the initial conception of the idea all the way up to our first interviews and it didn’t take too long, probably like 2 to 3 weeks, I’d say, before we actually decided to fully dive in and try our hand at something of this scale. And so together with me and our CEO, Ken, we started vetting out applicants. And like I like I said, I felt like I was the universe who nicked this together. Like Habib was the first. One of the first interviewees we actually saw, or applicants that we put out. And then as soon as we finished that interview, I remember that day as like something clicked and I felt like Habib was the one that would see us through all the way to that then. So far it’s been nothing but great working alongside of him, so I’m really grateful.
Patricia [00:05:49] That’s amazing. Habib Like you have all these amazing accolades and you could be working at any big tech company. Like, why did you choose AI Purity, which at the time wasn’t even called AI period. We’re a startup business and we’ve literally started this from the ground up. Like, what made you want to take this partnership and work to create AI purity? Even though you have these like you could have be working anywhere else?
Habib [00:06:15] Actually, when we started this startup together, I remember that from the first meeting I had with Chris and Ken, you know, I like, I like this team’s perspective towards the future and everything. Everything went in a way that I was always happy in this thing, and this was what somehow was a kind of of motivation for me to continue. Now, I’m very happy to be in this group, in this thing. But yeah, as I said, the, the, the main things that were really interesting and important for me to be here were first, they had a great Motivation and point of view towards future and then they they they really are great people great team really I really enjoy my working here in this team and this is something really, really important. And finally, I, I am and always have been happy to be here and enjoyed my time here.
Patricia [00:07:29] Thank you. I mean, it’s I definitely agree. It’s been a great team that I’ve actually been had the pleasure of working with. Ken’s really great. Chris is really great. And knowing you now Habib you’ve been really great and so pivotal to the creation of AI Purity and with your respective backgrounds. I mean, Chris, you’ve worked as a music producer, you’ve done many assistant jobs, even managerial jobs in the past, like a lot of people and a lot of people who want to, you know, start their own businesses or want to be successful entrepreneurs, they want to know, like, how does one get to a place where they can call themselves successful? Like, how do you feel like with everything you’ve done with your past work background, How did how do you think you got to AI Purity. Do you think it was luck? Do you think it was because of everything you’ve studied? Like, how did you get here?
Chris [00:08:20] That’s a that’s an awesome question. I guess it I, I would be lying if I didn’t say luck also played a part. But other than luck, I feel like, you know, other than luck and being at the right place at the right time, I feel my ambition and my work ethic also played a huge part of that. Of course. Yeah. Like you said, I’ve had different managerial roles before, so I’ve been a project manager, I’ve done human resource management before and a bunch of other odd jobs here and there, like I was a communications officer. But overall, the main thing for me and ever since I was like a young kid, I really like people. I love working with people. I feel like like I said earlier, communication is a really, really big deal for me and I always ensure that the people I work with, I communicate well and overall, I just like making things run. Yeah, making things run smooth. So I guess that all my previous years of experience in different roles kind of led me here to AI Purity it’s always been my dream to kind of work in a startup, especially in the tech space. Just a little bit of background. I wanted to take computer science as well, but you know, I wanted a different route. I went the marketing and psychology route, which I do not regret, but of course I would have loved to upskill a bit more in the tech space. But fortunately, I’ve had I’ve had I’ve been taking lessons and I’ve learned a lot from Habib himself to overall process over the entire process of building AI Purity. So, yeah, I think I hope that answers your question.
Patricia [00:10:00] Yeah, absolutely. I mean, everyone has like different routes that they take. You know, you never really know where you’re going to end up. You could take like a certain degree in college. And like Habib, he took computer science, computational linguistics, and now he creates this amazing platform related to what he studied. So, Habib, you know all about computer science. And in the past, did you think you were ever going to be at a startup that is about artificial intelligence? Was artificial intelligence even like studied during that time when you were in college?
Habib [00:10:36] Yeah, that’s a great question. Actually, I during my studies, I focused on many things that are directly related to artificial intelligence, like natural language processing and machine learning. But yeah, artificial intelligence is a very vast area. So we cannot say that everyone who is in our population artificial intelligence should work on everything. But yeah, these I mean, my expertise is closely related to our intelligence, which are natural language processing and machine learning. So yeah, I, I, I would expect that one day I would work in this area, but maybe the, the separation here. But this was somehow. This was not what I expected, really, to be honest. At the beginning of this preparation. I thought that so it can either be kind of a two week, three week, one month cooperation. But little by little, I think both sides, me and Chris and Ken we we understood that while we can do great things in the future and we continue with and we thought about new ideas and we decided to add new ideas to our previous ideas. And this, this went on like this. But yeah, regarding your question, sorry, I was just so deviating from your question about your question. Yeah, I could expect that I would work in this area.
Patricia [00:12:15] I love that and I can see that it’s been a really beautiful partnership and you know, it just goes to show that sometimes you expect certain things to go one way and they go another, and it’s actually for the best. So I think you guys are both equipped to tell me and the viewers what is AI Purity?
Habib [00:12:36] Yeah.
Patricia [00:12:37] I would love to hear from both of you. Yeah.
Habib [00:12:39] Yeah, sure.
Chris [00:12:41] Sure. I’ll go first. Yeah. So in its essence and its main core AI Purity is a company that focuses on AI content detection. So pretty much everything in that domain as well. We include paraphrasing text detection and. Yeah, so but I’m sure we’ll talk about the specifics more as we go along. But overall, that’s the main essence of our company. Obviously, we want to expand, in all things AI down the road and we do have some great plans ahead. So but at its most pure form, that’s what AI purity as we are, a AI content detection service.
Patricia [00:13:18] What would you say about it Habib, as the mind behind AI purity? What would you call it?
Habib [00:13:25] Yeah. From the name it’s somehow clear that the first, that the main idea behind AI Purity was to purify AI. Which means that, so, And this is this is the start. But of course we can have many great ideas in the future. But now AI generated text detection and AI paraphrasing text detection , which are the first ideas with which we have started AI Purity are somehow great ideas to detect what academia does not need that. I don’t know. Maybe not just academia, detective, whatever. Society does not need that which is cheating or plagiarism or those sort of things. So because now [00:14:19]artificial intelligence is changing the world. So this is good and bad. You know, always everything can be good and bad. So with artificial intelligence, that’s the saying, you can do great things, but you can do bad things, too. So AI Purity’s responsibility is now is to detect that. For example, I’m just if you want me to make some examples, for example, students can, can, can do, can use artificial intelligence to. Unfortunately, I don’t like this, but cheating you know, to you know, to to show whatever they’re not. I’m not going to say that they should not use artificial intelligence or tools like Chat GPT. I think they they they should use it, but in a good way, not in a bad way, because now I hear that some people are saying that, no, Chat GPT should be limited or students should not be allowed to use that. I do not agree with those people. I think everything is changing. So even academia should change it. So with the New World, even professors should change themselves to the new world. When there is a very there is there is such a great tool to use. Why should we not use that? But we should use it in a good way, not in a bad way. And this is the the main goals of AI Purity at this time to detect those bad things but to to provide great tools for the students and for other people to use in the future. I mean, the things that are related to artificial intelligence, but not everything but the specific things. [114.8s]
Patricia [00:16:15] That is so beautifully put. And I totally agree. AI purity purifies the AI, but at the same time we advocate its ethical and proper use. So could you tell us a little bit more about what makes AI Purity stand out? I mean, I know that there are already AI detection tools out there. What do you guys think makes it stand out? Why should our viewers right now or anyone listening to this podcast try out AI [00:16:42]Purity? [0.0s] What sets it apart?
Habib [00:16:45] Well, if I want to answer this question, because this is a very good question and part of this question is technical and part of this question is not. So AI Purity really tries to do whatever. That cannot be done easily. You know, because when we talk about AI generated texts, we cannot categorize all the AI generated texts in one category. We have some text can be easily be detected, but some texts are really, really hard to detect. What we are trying to do, we are trying to develop really complex models which are able or are capable of capable of detecting challenging texts, challenging AI-generated content. That’s a little bit of the technical aspect of this question. For example, we’re using the combination of paraphrasing and paraphrasing or ai paraphrase detection model and ai generated detection model, because when text is paraphrased, so it’s really, really difficult to detect. So to answer this question, sure, we are agreeing to do difficult and challenging tasks, which, to my knowledge few companies. Have focused on so far.
Patricia [00:18:22] Chris, any thoughts? What do you think about AI Purity? What’s its main stand out?
Chris [00:18:28] Yeah, well said, Habib. Great question as well, Patricia. For me, the main thing that I always like to say and I’ve been saying the past few months since we’re building the tool is at AI purity. We want to provide more than just data. We want to provide insight. So what I mean by that is there are a couple other companies right now in the space that claim to be perfect. The difference with us is we know this space is ever evolving. Always. There’s always something changing. And so there’s no really there’s no tool that can be perfect. What we can do is just do our best and provide our users with as much data as possible for them to hopefully look into and decide for themselves. So that’s our main goal. And I, I feel like that’s one of the main things that differentiate us from our competitors. It’s our transparency and our commitment to AI transparency as a whole. But yeah, what Habib said is perfect as well. We’ve taken measures to ensure our data is as accurate as possible. We’re trained on multiple data sets and made a lot of our own. And so yeah, at the end goal is to make sure our users have the insight that they need to make an informed decision. And I feel like that’s what is very important, especially in this new day and age. Now, ever since the release of Chat GPT, there’s been a lot of turmoil, not just within the academic sector, but pretty much everything related to writing. So while we’re hoping to start within the academic sector, we have goals to hopefully expand into different sectors as well, mainly journalism. We feel like the government, government officials would love a tool like this. Our lawyers, honestly, the sky’s the limit. And I feel like with AI purity, we can definitely reach a lot of heights together.
Patricia [00:20:17] I feel like there’s a level of sophistication. AA is trying to achieve that. Not all AA detection tools are able to do, as Habib said, like it does. It doesn’t just target AA generated text, but if I’m understanding correctly, paraphrased AA generated text on top of giving out all these amazing data that maybe aren’t given out by other platforms. So I mean tech talking about those technical aspects. What I really want to know actually, Habib, is what is natural language processing and how do these like technical and thus technological advances, how do they how does it even work? Like how do how do the A.I. detection part work?
Habib [00:21:02] Yeah, that’s a good question. But not that’s not a question which can be answered in a few words. So I try to to answer this question in a way that is related to what we are doing. Coming up. Idea A generated text detection feature. Natural language processing is a kind of bridge between linguistics and computer science, which means that. So if I’m going to just answer this question from the point of view of artificial intelligence, that natural language processing is enabling the machines or the computers to understand and generate non human language. So these are the two main ways of natural language processing, but it has many other aspects too, regarding what you’re doing now. Natural language processing together, big machine learning. Machine learning is not just a natural language. Processing machine learning can be applied to any classification task like image classification or. Whatever. It is, a kind of classification task that natural language processing is to process the language here. The language is the text in the written form, but the language can be spoken to. Or sometimes the language can be natural. Language processing can be combined with image processing, for example, detecting handwritten, for example, something which is mature. Natural language processing is helping artificial intelligence. It’s not only natural language processing. Natural language processing is part of that. And the other part is machine learning. But both of them together, they work together to build this model or this system. But if I’m going to go into a little bit more detail, they’re using language models. Language models are not what are changing the world of artificial intelligence, for example. Now everyone is talking about. Why? Because of those language models. Now they’re calling them large language models, which are behind behind all those great systems. Yeah. I think I hope that I answered your question. Yeah, it’s it’s, you know, it’s not it’s a little bit not easy to answer yet. Things are shortening.
Patricia [00:23:44] Yes, the technical aspects are a bit challenging to understand for the regular person. But I did want to ask, just based on all the stuff you’ve said. So machine learning is also a great aspect into the creation of the platform as well as natural language processing. Would you say that the more people use AAPT, the better it learns or the better it operates?
Habib [00:24:06] Yeah, exactly. And this is this is related to the machine learning. Part of what I said. How it works is that classification based on machine learning works with large data. And it depends on the tasks. Sometimes need larger data on some tasks need not that large data. But anyways, the more data you give to this machine learning based model and the more it learns how it can, we can say that it can predict future based on past observations. So if you, if you train the model with the data training data that contains, for example, apples in the future, the model can tell you, okay, this is an apple, I got it. But if there is no apple in the training data, this system doesn’t know what an apple is, for example. So but if you want the system but imagine how many observations because in the language we cannot count it. It is it is a very, very, very, very large number. And this is why learning languages is difficult, because I’m making a sentence that this is structure. You can make the same the same sentence, but in other structure you can change this word. In other words, you can just. So somehow we can create a word at the moment. So the observations are the number of observations are huge. So we can see that, okay, the machine, these are the things you have to learn. And please learn this, learn them. So that’s not possible. That is possible. But that won’t give you a very accurate model. So because of that no large language model, because you are using small tons of data, a huge number of datasets. So they didn’t we did with the very big number of data, the big data. Now we here, we’re having these models in this data and we could not have these great models. And the first problem that building a model that is able to process this, this amount of data is difficult, too, because it needs great infrastructure, super, super fast computers. So I’m not going to go into the detail, but this is why now free trade language models are available for the researchers, students for people to use. So you don’t need to go and train all the way if you can, if you can change your own model. But even if you cannot do that, you can use those pre-trained language models. And great companies have trained those language models and know that those language models are available for you to use. But you, you, you only go and get them and you find to take that language model to work for your task. So, for example, any kind of task which is which can be done with language models or primitives in natural language processing. So in general, yeah, the more data it has, the better it can work and the more it learns it gets trained.
Patricia [00:27:40] Amazing. So it constantly gets better. I honestly like it’s mind blowing to me to understand. So I really wanted to talk about a little bit of the business side of AI purity. So, Chris, as the chief operations manager, were there any challenges that you faced during the development of AI purity and its process, and how would you say you overcame them?
Chris [00:28:05] Yeah, that’s a perfect question. Yeah, obviously I feel like with something of this scale there are bound to be hiccups, but it just matters on, you know, how you bounce back and how you tackle those that hiccups. Yeah. And more specifics, though, overall. It’s been a really smooth process. I’d say 90% of it has been smooth working with Habib and the rest of the team. Ozzy has been perfect. I feel like, of course, the initial days we were like trying to get to know each other better. But I feel like going into the second week after that, it was like smooth sailing on from there. Obviously there are some spare parts are tough. There are some stages in development where we had to hire a third party integrators and developers to help us with our project. That was a long process, but I had to be done and I’m happy with Habib’s help, especially we did get it done in the end. But yeah, it was like a bunch of interviews, so many interviews, but in the end it was very, very well worth our time. And as of now, everything is going really smoothly. Other than that, though, I feel like the main part was just trying to help Habib as much as I could. Especially with my limited, you know, computer science knowledge. So the most part of what I do is just ensuring everything runs well. So I pity the middle man for Habib and the rest of our team. Our other read our web development team reps, Sharks. I shout out to them. They do amazing work. Yeah. So my main job is of making sure all the gears are running smoothly. Well oiled machine across the entire board. But yeah, this it’s been nothing but fun. I’ve upskilled so much and I’ve learned so much and I continue to learn so much every day from the brilliant minds like Habib and everyone else that surrounds us. And honestly, I couldn’t have asked for a better team as a whole. So yeah.
Patricia [00:29:59] Amazing. And you guys talked earlier about like, academia and its impact and what AI Purity hopefully does in terms of those industries. So I wanted to ask you guys, like how impactful do you think I will be to the interest industries it’s targeting? And would you tell us about the other industries besides academia that it wants to target?
Chris [00:30:23] [00:30:23]Oh yeah, for sure. Yeah. I’d love to touch on this question and I’ll let Habib answer it as well. But yeah, for me, I think we can all see pretty much how the invention of chat gpt kind of shocked the world this year. And with that, you know what I mean. It’s been kind of like a nonstop evolving thing. And for me, specifically for ai purity, I know our most of our competitors usually focus on the education side of things, which is fair. But for me and for AI purity, I feel like we have a longer vision. I foresee the AI Purity brand to be like a one stop shop for all things AI down the road with AI detection as our corner piece. We have no other than that other than academia itself. Just straight off my head. I feel that journalism industry especially would love something like this, especially with the changes of how SEO works now. So copywriters, marketers, even human people in the human resource industry, like, you know, people who vet applications and whatnot. The use cases are very extensive and honestly, the sky’s the limit AI [66.5s] purity, what AI purity is working on now really will have a very impactful effect on what is happening in industry in the future. You know why? Because, as I said, there’s no way people have to use this great tool, which we call now artificial intelligence or maybe better words. Maybe now it’s just exemplify it. So we cannot limit people from using that. On the other hand, unfortunately, some people use it in a way that now we can see that the senator is in the US and some people in the Parliament of Canada, they are trying to limit this great tool, which I understand. I totally understand their concern too. And this is why we say that this is this has a good impact in the future in the industry because what we’re doing now can somehow convince those senators, those people, people in the parliament, that okay, there. Are some ways to to remove those risks, to to to remove remove the dangers of using artificial intelligence in a bad way. So there are in these companies, they are doing great things to try to avoid people from doing from using this tool in a bad way. So I think what they’re doing now will help artificial intelligence grow in the future. And it is very important.
Patricia [00:33:54] Absolutely. I agree. I think one of the main goals, if anything, is that we hold people accountable for how the way they use A.I. and like if more companies and like if A.I. pretty garners that success within its users, like you said, hopefully it changes the way people use A.I. and that hopefully they don’t use A.I. for bad, for cheating, for plagiarism.
Habib [00:34:17] Yeah.
Patricia [00:34:18] That is one of our main goals. So right now we are focusing on AI text generation and detecting that specifically. I wanted to ask you like, how are we able to ensure currently the accuracy and the reliability of A.I. purity and its capabilities to detect A.I.?
Habib [00:34:38] This is a good question because we also thought about it a lot. So how can we develop a reliable system which does not accuse people of, for example, cheating wrongly, you know? And this is an important thing. Our systems must be reliable because we are going to target texts. And behind this text, there is the real person. So we should be very, very careful not to accuse a person of doing something by mistake. And this is this is our real goal indeed. So to this aim, we try to evaluate models as much as possible with different, different different types of texts. And then we are trying to train our models and build our models in a way that if they are going to make a mistake because as Chris said, we don’t have a model in the world which works with 100% accuracy in artificial intelligence. That’s not possible so far. So so we’re building our models in a way that is, even if they’re a rarity, the model is going to make a mistake. That mistake won’t hurt people. So this is this is something which we have set as one of our main goals for our models. So as I sit and through scan, maybe add, yeah, we evaluate our models really well. Very complex process with many, many different types. And this helps a lot in this process. Thanks again. And you can add this to to what I said.
Chris [00:36:39] Yeah, for sure. No, like. Yeah, very well said. It’s a really great question for Hub specifically, but as Yeah, I’d love to chime in too. For me personally, it involved a lot of reading and a lot of going through data sets and a lot of AB testing really until Habib and I were sure that, you know, it was making the results that we wanted it to see that, that the accuracy level that we were aiming for. But yeah, no, there’s no other way around it. It just involved a lot of hours, like numerous hours over the past hour, just reading and going through different data, different forms of text. You know, there are different kinds of text. There are just essays. There are like transcripts, there are poems. As you can imagine, there’s so many different types. And so, yeah, to ensure the accuracy and reliability of our tool, we had to go through all of them. There was no way around it. So I hope that’s our question.
Patricia [00:37:36] That’s really great perspective. So I really see the collaboration between the two of you, which leads to my next question about how important collaboration is, especially at such a large project. So how do you guys effectively bridge the gap between the technical side, which Habib is the best at and the operational aspect which Chris is the best at?
Habib [00:37:57] Yeah. Yeah. That’s a that’s a that’s a good question. I think that a great job is not possible without a great team. Imagine that I built, I mean, really great models, but Chris couldn’t help me in the other side and Ken couldn’t help the other side. I mean, the managerial side results in the managerial side. And so because. We face a lot of problems, a lot of challenges, challenges on the way. So those those difficulties, those challenges could not be overcome alone. Am I right, Chris?
Chris [00:38:36] Definitely. As as we know, all of us have different strengths, and I feel like that’s what Habib’s trying to get at. So Habib is like the brain. He’s the wizard behind the tools. And so, yes, so for me personally, bridging that gap between the technical aspects and operational aspects of the project, was it easy? But I feel like I grew into the role really, really, really well. A large part of that is attributed to how Habib works. His work ethic is crazy. So, so good. Everything is clear ever since day one. And so I have to give credit where credit is due. It wouldn’t have been easy to project manage everything else if Habib was on top of everything he does. And so, so much credit goes to Habib specifically. But yeah, other than that, though, I feel like, like Habib said, our team, our whole team is very integral to the success for the project so far. And so my job was just making sure all the technical terms were translated into here, you know, understand.
Patricia [00:39:33] [00:39:33]Layman’s terms. [0.0s]
Chris [00:39:33] Concept, exactly layman’s terms for the team, because obviously we decide as a team for any problems that we did face. We ran it by together because we really believe in, you know, talking it through in detail, proper communication. Like I’ve said, this a very, very big deal. And so, yeah, so that’s the main part of my job, is just making sure all Habib’s technical concerns gets translated into layman’s terms for the rest of the team. And so of course there were days where that got a bit tough personally for me too. So I had to upskill. I had to learn what certain terms were to, you know, ensure the the machine is well oiled and running at any given time. And so, yeah, overall, though, it’s definitely a team effort. I feel like each and every percent of our cells are very much what they do and so all credit goes to them. My job is just making sure everyone works together seamlessly. And so, yeah.
Patricia [00:40:31] And what have been some key lessons that you both learned during this collaboration process in creating?
Chris [00:40:37] AI Purity well, yeah. For me, just quickly, I feel like I had to upskill a lot. That’s a given. But I feel like patience for some, for when you’re trying to do something great is very, very underrated, I feel, you know what I mean? Everyone wants to be successful. Everyone has a goal in mind at the end, but without patience and, you know, a good work ethic, it’s very hard to get to that goal. And so I feel like that’s one of the things I’ve improved on the most. It’s just, you know, other than project managing as a whole, this developing that patience and having a clear goal, taking it a day at a time has been very, very good for all of us. And so, yeah, I’d say that’s probably one of the bigger things for me personally.
Patricia [00:41:24] How about you, Habib? Any key lessons?
Habib [00:41:27] Yeah, I, I would like to repeat Chris’s answer because, yeah, it is one of the most important things I have learned. And it was a lesson for me. All the process was a lesson for me, yet patience and not giving up. Because always there is a challenge. Whenever you want to do a project, you will face challenges and challenges and the more challenges you overcome just so you can go to the next step, you will advance somehow. And either you you want to do even small project or you want to do a big project. This is the same. You have to overcome challenges, problems. If you cannot solve those problems, you cannot, you cannot advance. So as Chris said, you must be patient. And if you think that, wow, again, a new challenge again, we have a new problem. So why the problem? Why we’re having a lot of problems this way. You cannot you cannot advance. The lesson was not only this. I mean, everything was lesson. And Chris knows that we needed to hire some people to do some is the proper app for us. And in the process and the process of finding those people, we had a lot of challenges. So I’m not going to talk about those things. But yeah, yeah, I think the most important thing is patience and problem solving.
Patricia [00:42:57] And could you both share some memorable moments while creating AI Purity? Let’s go with Chris. Do you have any favorite stories you’d like to share with us?
Chris [00:43:07] Yeah, thanks for the question. Yeah, there are there have been a few throughout the past few months. We’ve been working like the tool that stood out to me. I remember this was probably one month and working with Habib, this was the first time he sat. The code for me to test that my local machine. And I remember being so surprised and in awe of what he made. But obviously I had to help. I had to ask him to help me run it for the first time. And I had so much questions. And I remember being worried that Habib would find me annoying. But I know. And like, he was very helpful. And I remember it’s like core memory for me now was when I actually got the code running for the first time and this was like our first MVP model model. What I remember like getting the result to show up on my computer screen was like such a big moment for me. And like, it kind of showed me, Oh, you know what I mean? It’s the first real thing that we can actually quantify and see. That was the first thing for me anyway. So and ever since then, it’s just been milestone after milestone. But I specifically remember that first time I got it working on my ad that was like very, very hard to forget.
Patricia [00:44:17] I remember you sharing that to us today. We were all in awe of like your first little progress, and I thought you have. What are your memorable moments in creating AI Purity.
Habib [00:44:28] Memorable moments are related what Chris said. So, I mean, the evaluation of the data models were very interesting moments because Chris is a great person in the evaluation and the moments when Chris got back to me with great news and and those moments were very memorable for me because I saw that Chris was excited about the results. And also the team was very happy about the results. And at that moment, yeah, I was very, very happy. And this was I, I couldn’t change that moment with anything else.
Patricia [00:45:07] Amazing. They gave me sharing that. I can see like how happy you were with the progress little by little, and how this collaboration just went beautifully. Now, this is a little bit of a more technical question. I wanted to ask, like, how did you both find the perfect balance between innovation and practicality when designing the A.I. detection tool that we now know as a priority?
Habib [00:45:30] Oh, this is a great question because when you’re going to develop a model, you can think of several innovations. I mean, you can maybe when you think about new ways to do something, there are many ways to do that. And maybe theoretically you can. You can go with each of them anywhere. But yet the viability of the idea and if the idea is the if the approach is the business approach is very important. Yeah. And I myself tried to to, as I said, make a balance. I didn’t want to just go with innovative ideas which, for example, reduce the speed of the model. On the other hand, not just to stick to the, to the old approaches or to the. Yeah, this is very important and thanks for asking that. I think there must be a balance between innovation and practicality of the the system, the model. And I think that we somehow did that. But maybe maybe Greece can can can confirm that. But we. My opinion is that we did that.
Patricia [00:46:57] How do you think you’ve been able to successfully do that, Chris? I mean, obviously with something as huge of a platform as I do, you want there to be like the best features, but we also at the same time want it to be user friendly. So do you agree that we’ve been able to do that with the priority, as Habib said?
Chris [00:47:17] For the most part, I definitely agree with what Habib said. There’s a fine line between practicality and innovation, and I feel like finding that good balance is very crucial. And one of the was one of the things reprioritized ever since to start back during just the brainstorming phase and conceptualization of our period as a whole, we made sure to account for this specifically. So thank you so much for that great question. As you know, like building something of this caliber can be overwhelming because as as you said, we want to be, you know, offer more perks, offer more features that our competitors may or may not have. But at the same time, we need to keep it practical and user friendly and, you know, all these other technical things I’ve got to worry about. I know like in the end I feel like we did and are still doing a great job at keeping expectations well-managed. I’d say it’s very easy to get overwhelmed with a lot of all the things you want to do and implement. So I feel like it all kind of just revolves around the team once more having a proper team to discuss things with our feedback loop between me and Habib specifically is very good. I feel like we work really well together. And so you managed to eliminate what we don’t need and we manage to integrate what we do need. Then after that, once we, you know, share with the team, it’s easy to kind of for, you know, it’s easier to believe in what we believe in when they believe in us. I feel like so, yeah, overall, I feel like we are doing a great job at that and I promise we will be continuing to up keep that momentum and that success we’ve kind of built there. You know, I just want to end it. Well, we will keep dreaming very high. We keep aiming to be the best, but at same time, of course, we got to prioritize. I feel like that’s what I hope I’m getting out with it. So thank you.
Patricia [00:49:12] Amazing managing expectations. Absolutely. We want to give the best, but we also don’t want to put too much on our plate. So small steps, right? So I want to talk about A.I. in general. I want to know where you both stand on the ethics and responsible use of. I know you guys have already touched on it a little bit, but on top of that, I wanted to ask like, how do you think A.I. pretty addresses these concerns on ethics? Habib AI Do you want to go ahead?
Habib [00:49:41] [00:49:41]Yeah, sure. Well, what AI means in terms of the ethical aspects of A.I., because when you’re generating something. You have a maybe more responsibility, heavier responsibility to be careful about the ethical aspect of what you’re doing. But now we’re paying attention exactly to the ethical aspect of AI. Now, the ethical aspect of AI is one of the main concerns. But fortunately, especially in the area of natural language processing, I mean, whatever is related to the to human language, I mean, this aspect of AI, we’ve seen great advances, especially recently. Researchers, I think, are paying attention to the ethical aspect of AI more than before, especially with the new advances. The ethical aspect of AI has become even more important. As I said now, everybody, everyone is talking about these advances and how people should use that, how they should they should develop their models to generate authentic content because sometimes it gets very, very dangerous. And for example, we have systems based work based on artificial intelligence, which should make very, very important decisions. So if they make a mistake, so the consequences of that mistake are huge. So but yeah, what you’re doing is neither dangerous nor risky because we’re trying to remove those dangers and those risks instead of generating those risks. I hope that I was clear enough. I don’t know. [117.3s]
Patricia [00:51:39] Yeah, that does make sense. I kind of like working as, like, the balancing act between how people use AI and how to use it responsibly. Were neither here to do anything crazy, like how A.I. has changed the landscape of like, so many people are now afraid of like losing their jobs. A lot of industries have been impacted just by the use of AI alone, and we are here to kind of be the the guardians, so to say, of like how people should responsibly use. Yeah. Do you agree, Chris, or have you.
Habib [00:52:11] [00:52:11]Sorry. I just wanted to say that what I answered was the ethical aspect of AI from the point of view of the people who are working on the AI, not the people who are using AI. If you mean the people who are using the AI, yeah, that’s another question. And that’s a very important question too. I think that I’m sure that people in the future will receive training how to use AI. Now a lot of people or I think maybe in general people, they do not have a training how to use A.I., but in the future, schools, universities or other educational institutions, they will add subjects to train students. Or if the people who are in their system how to use artificial intelligence. So in the next maybe 10 or 20 years, things are like this. People know how to use A.I., too. And I think that everything changes. University systems change, schools change because with the new advances in AI, if you know how to use that. You can do really great things. And this is why I believe we should not limit people from using AI. Instead we should train them. We should teach them how to use AI. [88.1s]
Patricia [00:53:41] Amazing. Chris, what do you think?
Chris [00:53:43] No, no. Yeah, for the most part, I definitely agree with everything I’ve said. Yeah, it’s it’s a very loaded question. And I feel like, depending on who you ask, people will have different answers. Right. So speaking on behalf of our team and AI purity, I feel, of course, that we hold proper, you know, proper ethics and responsible use of a at a very, very high regard. At the core of it all. I feel like this is kind of more or less the reason why we even started this tool in the first place. Just for some context, I remember back during conceptualization there are a lot of students that were being framed by their profs, but also vice versa. You know what I mean? It’s like all your essay was made by an AI, but what can the student do if they’ve been accused? There are there is there really isn’t a solution. And so that kind of this question kind of plays a part into that. It’s like we want to be able to leave a good impact in the world. After all is said and done, especially in the realm of A.I.. As you know, personally for me, [00:54:47]I do feel like it’s here to stay no matter how we feel about it personally. I always tell my friends and colleagues, it’s like similar to the invention of the calculator or the invention of the Internet back in 1983. At first there are going to be a lot of people who despise it, but eventually people are going to come around and understand it’s something that we won’t be able to stop and rather just adapt with. [24.8s] You know, there have been a lot of times in history where something of this sort has happened, like something of this caliber. And so if we’ve learned anything from the past, we should just always be able to adapt, adjust and analyze and, you know, move accordingly with that. And so for yeah, but for AI to be as transparent as possible, our end goal is to be one of, like you said, the Guardians are the good people in the in the in the industry that are actually there to help with something, you know, something of this caliber. It’s very it’s, it’s human nature to find people who abuse it. But at the same time, the other side of the picture, it’s very normal to find people who are actually trying to improve on it and try to make everyone’s lives, you know, a better place with the work they do. And so I feel like we fall into the latter part of that. The end goal is just to, you know, give people something they can use that will be beneficial for everyone around them and of course, to them specifically. So.
Patricia [00:56:13] Amazing. Thank you both. And I also wanted to ask, what are the steps that were taken to ensure a period use ethical use and the prevention of potential biases and its algorithms?
Chris [00:56:27] Yeah. Specifically, our data sets that we use to train the model were taken from ethical sources, of course, and or sources that were properly vetted. This kind of ties into your previous question too, on how we ensured the model wouldn’t have any biases or anything like that, or, you know, it would have, I wouldn’t say any false information. So yeah, with that in mind, we made sure to actually read every data, go through everything. We feed into the model to make sure, you know, there aren’t any false information that we provide that we have no biases at all. As you know, a catch up even they themselves have been a bit more strict into what they what you can and can’t do. And so for us, we kind of more or less try to adhere to what they’re doing, especially since they’re the industry standard for everything. But whatever they think is fine. And of course we’ll do the same. And for now, hopefully down the road we’ll have a voice of our own to, you know, influence these other industry leaders. But for now, we’ve just done our due diligence. We’ve kept our head down, make sure everything we’re working on is properly vetted and everything and all the data sets that we made our own were also, you know, thorough and not show any bias at all. But yeah, by all means, have you, if you have any thoughts to share or.
Habib [00:57:45] Yeah, sure. Thank you, Chris. Yeah, I just want to add that the most important ethical aspect of our system is to make the best decision. That does not hurt the people who are using that or the people who are connected to that. For example, as I said before, if if the university professor using that to see an article written by his student students or a student is, hey, I generated. If we make a mistake, this mistake, and have a consequence for the student. So this is not ethical not to pay enough attention to this thing. So we’re trying to do our best to reduce the risk of making such mistakes as much as possible. So I think this is the most important part of the ethical part of the ethical aspect of our system, our tool. And as we said, you’re trying to use the best data to train our models, not to use the data, the data that biases our model and those sort of things. And I totally agree with Chris. But as I said, when people are using our tool, we are making sure we are doing our best to build a model that makes the possible mistakes that convert people.
Patricia [00:59:20] Thank you guys for sharing that. And I think I agree that AI period does its best to contribute to the responsible use of AI in society. Now I want to ask, what do you guys think are the future plans for AI purity? How do you envision us growing in the near future? Potential updates, Maybe two features. What do you guys think?
Chris [00:59:44] Without sharing too much of the specifics. But yes, of course, we’re just going to keep our momentum, our momentum up. Right. We have a really great trajectory and awesome plans for the very near future in terms of potential updates and new features that we want to introduce to high purity after our eye contact detection talk. I am not liberty to say any more. But all I can say is the future looks really bright. And like I said earlier, we were aiming to be a household name for all things A.I. down the road. Hopefully in a year or two and like a one stop shop for everything A.I. related. Of course, ethical, irresponsible. But yeah, I let all that help you take it over for me.
Habib [01:00:31] I think there is. Yeah. I just want to add that we do hear a lot about the future, but one of the. This. Things I just mentioned is that we are going to focus on the research and the academic part of what we’re doing too. So we will contribute to the students to this area in the future by publishing papers and attending conferences in the future, too. So this is maybe one of the best things that I love about the future of AI Purity.
Patricia [01:01:17] Amazing. I think we are really trying to establish that goal of becoming authorities in the AI detection world. And that’s what we can right now disclose to our viewers right now. Well, I did want to ask the specific industries that we wanted to target and we wanted to contribute to. But you guys already talked about academia and other industries. Were there any more industries you think that would be the best suited for you to target besides academia?
Chris [01:01:49] Chris Christie You know, honestly, that’s a good question. Like, yeah, obviously we want to develop a bit more in regard to academic. And so the academics sector, mainly because they’re the ones who need it the most right now, but down the road and very, very near in the future, we want to start helping out with marketers as well. Journalism industry will need a lot of tools like this, you know, copywriters, marketers. Like I said earlier, even government bodies like for cyber, because we want to get into cybersecurity at one point, make sure everything is in our guns well for them. There are a lot of use cases for A.I. related tools that they can use. But honestly, it’s a very broad answer, mainly because we have plans for each and every industry out there right now. This kind of ties it to, you know, being practical and prioritizing certain features first, of course. But eventually I see us down the road tackling every sector out there, mainly because I will seep into every sector out there. And so we’ll be here to stay for sure.
Patricia [01:03:00] What do you think, Habib?
Habib [01:03:02] Yeah, exactly as proof that I believe in many parts of every sector in the future. I think that all those sectors in the future need to use AI in an ethical way. So of course, there are some some stuff which must be purified and this will be our main mission in the future to purify those things.
Patricia [01:03:25] Besides, are there any plans besides AA text generation that we would like to purify? I know a big thing right now is like deepfakes and stuff like that. Have you guys heard about that?
Chris [01:03:36] Oh, yeah, for sure. Yeah, I love to talk about this as well. That’s a fantastic question. This also kind of ties into the ethical use of AI as a role. I feel like why that topic is brought up. The first thing people go to is, oh, you know, deepfakes, because that’s maybe one of the worst things you can use AI for. But at the same time, it’s also kind of one of the best. And the reason I say that is because, like I shared like Patricia shared earlier, I also work as a music producer. And so I has been kind of seeping into our industry as well. So there have been great things about it. There are some some deepfakes, you know, where they kind of emulate an artist’s voice. So, you know what I mean. There have been some great tracks released this year where the two artists never even met in real life. It’s all just generated, generated, but it’s such a good work. So but with that being said, of course, there are like a lot of copywriting implications as well.
Patricia [01:04:32] Yeah. Okay. Well, you were earlier you were talking about like deepfakes being used in the music industry. I mean, you said it’s done create great tracks so far, but it’s also equally as scary, don’t you think that. Machine learning doesn’t only store data in languages, it can now store data in images and voice, like how do we I mean, are you not scared of that? I mean, what do you what do you think you were saying earlier that, you know, now we can like create tracks from musicians that don’t even know who the producers are and they aren’t even the real singers of these tracks?
Chris [01:05:10] Now, you know, perfect like. Yeah. Just to expound on that a bit further. Like, like, like you said, deepfakes are very concerning right now, especially one of the most concerning things in regard to A.I.. And it’s one of the things that have a lot of negative connotations regarding it, especially throughout social media. But yeah, for me, it’s a great question for parody as well. I’m sure down the road we will want to start delving into, you know, Deepfake generation as well. I’m sure like artists and other creatives will need this down the road to like properly vet whether something was human made or AI made. And like, you know, with with everything in the art world and how A.I. is progressing over there, it’s been crazy. Like Photoshop has been on it. Like they’re generative, fearless, insane. Yeah. So there are a lot of you know what I mean? Something to verify whether an art piece was made by a human or even a mix of both will be very, very in demand very, very soon. And I feel I feel like Hollywood itself and, you know, the Western media and culture are slowly catching up to this as well. Like I said earlier, there are been a lot of tracks where artists collab together, but they’ve never met each other and it’s just overall really crazy. And so yeah, that from there you you can go into so many topics like who? Who does the copyright belong to when it’s made by an AI, even though it uses someone else’s voice, You know what I mean? There’s so many implications regarding all this. And so of course, eventually we hope to start delving into tools that can help in this regard as well. Just pretty much verify where I started. A piece of art or media comes from will go a long way. I feel especially in like one, two, maybe three years from now.
Patricia [01:07:03] Habib Do you think that’s possible for us to do in the future to not only detect text, AI generated text, but AI generated images, videos or even songs? Do you think that’s a possibility?
Habib [01:07:17] Totally. Totally. Yeah. Yeah. And I think we’re even able to do that now. But we at the moment have focused on this part of our project. I mean, mostly you mostly focus in the natural language processing part of the project. And in the future, of course we can focus on different aspects. Yeah. And as you said, everything can be can be generated by AI. It is not only text. There’s not going to be language and language has different aspects to language is not only written, it can be spoken to. Now, for example, you must have seen some. There are some applications that can imitate what you say or, I don’t know, generate a voice that is exactly like yours and speak instead of to you and do a lot of things. So, so those are other aspects that we can focus on to in the future. Yeah. Yeah, of course we can focus on every aspect of artificial intelligence in the future.
Patricia [01:08:18] And I wanted to ask you guys, like, should users of A.I. be wary? Because if I’m not mistaken, AI stores the data that is constantly fed to it like we do when we fed it. All these, like, scripts and like basically anything that they can learn from, like essays or books. Right now there are many trending applications online where you can upload your photo and then it gives you other variations of a photo, basically your face pasted on a different body. To our viewers out there, you guys have any advice like should they be wary of using AI in this way? Is it okay? What do you guys think They should?
Habib [01:08:59] [01:08:59]They should. There are many applications that are built solely for data collection or there are some projects based on something which is called crowdsourcing, which means that you work for a company or for a group or a project without even knowing that you’re working for them. So you must be very careful what you’re doing, which application you’re using, which data you’re feeding to the application. Exactly. Sometimes it won’t hurt you, no problem if you help others do. But some? Sometimes. No. You must be very careful about that. I even have an example in my mind. But that example didn’t hurt people. It was useful too. But maybe it’s interesting to to note that there was a handwritten, old handwritten text they wanted to use for machine learning, but they wanted to actually, if I remember correctly, they wanted to build OCR systems or it is optical character recognition to convert those old books which were for a library. I think the United States to convert them to digital data which can be used in, for example, e-books or something else. But some of the books, they’re very old and the OCR systems could not recognize the text and they collected all of those unrecognizable data and they gave those texts as CAPTCHA to the people and they asked people to enter whatever they see. And for example, they gave one to, I don’t know, 1000 people. And if, for example, out of 1900 people, for example, entered the same thing, they could somehow be sure that, okay, this is so instead of hiring people to do that, they use them as CAPTCHA. So this is this is this can happen in a bad way too for example, there are applications. Okay. We can add some features to your photo. Please provide your picture but maybe you’re giving your data for for for some stuff for you don’t know for what. So yeah, the question is yes, you must be very, very careful. [142.8s]
Patricia [01:11:23] What do you think, Chris? I mean, because you’re in the music industry and I know you’ve had your hand in creating songs and tracks out there. Aren’t you afraid that one day, you know, that would be fed to AI and someone out there will be producing music using your instrumentals or your vocals in the future?
Chris [01:11:41] [01:11:41]How perfect. Such a good question. I feel like Habib nailed it on the head as well. In the end. My my answer is yes. You should really be wary, especially in this day and age. But with that being said, it all does boil down to data collection. Like Habib also said that, you know, I’m sure we’re all on social media, we’re all on Facebook, we’re all on Twitter, we’re all on Google. And so we’re we’ve pretty much consented to sell our data anyway. So the scary thing about A.I., though, is they’re just sometimes taking this a bit a step further to where it’s kind of borders into the Unethical line. so I guess my advice for any listeners or viewers is just read the fine print. Know what you’re getting into when you download a certain app for more often than not, like the 30 seconds of fun you have with it is not worth the data they collect from you in return. And so, yeah, I guess that’s just my take on it. You should all be wary when we’re on the internet. Be wary on what tools you use and what you consent to like you know when you see a cookie pop up. Read that fine print. See what they’re actually taking from you. And yeah, but yeah, for the most part, I guess in regard to media creation and whatnot, just personally speaking, I wouldn’t mind if I no one used my instrumentals or like my voice to make art of their own. I’m sure, depending on the artist, it’ll matter. You know, their their opinion will change depending on who you ask. But for me, I feel like art is meant to be collaborative anyway, and so if they feel like they want to collaborate with me but choose to go that route, then by all means it’s fine. At the end of the day, it’s more art and never hurt anyone then. So I guess just in that regard, just be ethical as best as you can. It’s really the only thing you can do. I always try to ask from, you know, I always try to credit the original artists as that goes a long way, even maybe try to reach out and give them credit where credit is due, like in the form of payment depends. There’s so many ways to go about it where it’s not unethical, but I guess that’s just my advice. It’s be a good person, give credit where credit is due and yeah, we should be fine. [130.4s]
Patricia [01:13:52] Absolutely. I mean, I doubt that you would feel the same if someone were to take your work and capitalize on it without giving you credit. So obviously that’s where the line is of being an ethical when you’re using AI. So now that we’re talking about the data security, how can we ensure our users that once they’re, you know, giving out their information and using our tool to check if their work is AI generated or has traces of paraphrase in it, like, how do we safeguard their data? How do we ensure our users that know we’re not going to use anything that you tell us and sell your data? How do we safeguard their.
Chris [01:14:36] Now for sure. That’s a fantastic question. I’ll answer this sort have been, if you don’t mind.
Habib [01:14:40] Sure, sure.
Chris [01:14:42] Yeah. So going into AI purity at the start, we knew confidentiality and privacy would be a very, very big thing. And if it’s not, it might as well be the biggest thing, actually. So our promise as a company is that we’ll never sell or never intend to sell any data that you do that our users consent to give us for any samples that you may scan through our system. It only remains within our system. It’s where all we have the necessary security methods and compliance in place to make sure nothing is leak. We’ve been serial sharpeners and we’ve been in the consent writing space for almost, what, 12, 13 years now. And so a big part of that work involves privacy and client confidence, confidentiality. And so nothing changes for ad purity. We hold that at a very high regard. It’s really one of the most things we we’re very serious about because any sort of breach in our system can have massive implications. So yeah, I guess for us, we’ll just keep updated on all industry standards. That’s our main problem is it’s we’ll never sell you or the user or any of your data to any third party companies.
Patricia [01:15:59] I mean, on the technical side, can you also ensure that our users are we can safeguard basically their data and not infringe on their rights to their own work?
Habib [01:16:09] Yeah. Chris, answer this question. So I’m not going to talk about it a lot because it was good enough. I mean, Chris’s answer talked about everything, but regarding the technical aspect of the model, the model does not share anything with others. The model only receives the data and then processes it. And then finally, I mean the current models, at least they do not generate anything or they do not. They just decide or classify the texts. That’s it. So I think there is no risk in using the data in a bad way or, I don’t know, using using it or generating something. For example, if a person scans a text including his name and the person is worried that maybe his name is going to is going to be shown somewhere else and in other texts, no, that’s not going to happen. So we’re not going to use that. What what they give us to to generate something we’re, I don’t know, to use it to, to, as Chris said, to to sell it in the future. No, we’re not going to do that.
Patricia [01:17:25] Amazing. And I’m sure our listeners and potential users of purity are assured by that. I know the question I want to pose to you guys. How do you plan to stay ahead of the curve? There are other platforms that do similar things as AI Purity, but how do we stay ahead of them?
Chris [01:17:41] Yeah, for me, I feel like the most important thing really is listening to our users and putting our users first. So in that regard, I feel like if you’re grounded and actually listen to what people need. You can go a very long way. It’s all about keeping an open mind and keeping an open ear and actually implementing the ideas that people want to see and the people what the people need. And so I feel like what a lot of different companies do sometimes is the easy to get lost in the whole grand scheme of things and just start implementing what they want to do. And for us, I feel like ever since the start and so the foreseeable future, we’re always going to be putting our users first. And I feel with that in mind, we’ll go up a very long way in terms of, you know, keeping everything going smoothly and implementing new features as we go along as well. So that’s my take on that.
Patricia [01:18:33] What do you think, Habib? How do we stay ahead of the curve?
Habib [01:18:36] Yeah, from the technical aspect of the company or the service, we try to. Because artificial intelligence in particular, machine learning and natural especially natural language processing is going very fast. So every day, every week you you will see a new thing, a new technology, new advancements. So we should do research every day, try to use the most the state of the art algorithms. This the most this state of the art approaches in natural language processing and machine learning. I mean, our research goes hand in hand with our technology, our service. So this is what can elevate us and some somehow this can help us be or to provide the best service always in the future. Because if you want just to stop and say that, okay, now we have a model, this model is working well, but you you don’t know, maybe next month your model is, is not good enough for for the need based on the new advances. So my answer is that we should we’re and we are going to do both research and improve our models based on the new advanced.
Patricia [01:19:54] Amazing. Now for our viewers, like, what would you like to share and what can people expect from AI period now that we’re about to launch soon? Chris, you want to go ahead.
Chris [01:20:06] They have by all means. Yeah. So I feel like we shared a lot in this podcast. And so for anything that’s not been mentioned, I highly recommend check out our website. You can see firsthand, you can try it. And I’m sure without a doubt in my mind that once you do try it, you’ll probably be here to stay for the foreseeable future. But yeah, I guess I’ll leave it at that. A lot more to come from us, for sure.
Patricia [01:20:32] And how about you? Have you anything to say? Potential users.
Habib [01:20:37] Year. I encourage the people who are listening to this podcast to try this model. This is an amazing model. I’m sure that they will enjoy it and try and use it get.
Patricia [01:20:51] Now, you guys can both answer this. I want to talk a little bit about like specific features. Like are there anything that you’re most proud of about the features you’ve built on AI purity, something that has never been seen before or something that is just adding so much value that hasn’t been seen in this landscape before?
Habib [01:21:08] Yeah, I don’t know. I’m not sure if those features are not used in any other applications in the world, but as far as I know, we are using really, really unique features in this model. For example, we’re providing both information that the users need to know if their texts are authentic and some extra. I’m just going to call them academic information, which means that we are going to provide some information which can be useful for the people who are familiar with this area. For example, some statistical analysis of the text, some statistics, I mean some numbers. So yeah, these things I think are unique for for AI purity. Except for that, I don’t know if I can say that now here, but we are going to provide. We are going to explore that PDF word for the user, which contains really detailed analysis of their text with different models and they will provide really interesting information about their text in PDF. They can both see the results on the website and download the PDF and look for more in-depth information in that area.
Patricia [01:22:31] How about you, Chris? What can you share about the lovely features that are so unique to AI Purity?
Chris [01:22:37] Yeah, I think you did a really good job expanding on the features that I really particularly liked. But yeah, I know personally, just to add a bit on top of what he said is the PDF export option is amazing. It’s well crafted and it does include so much data that an average user, but also a more scientific user will find it very, very important to see and very, very useful. Like I said earlier, we aim to provide not just data but insight. And so I feel like the PDF really does that justice. And so that’s probably my favorite part of everything. It’s a great it’s over.
Patricia [01:23:15] Do we want to share with them like what they see on the PDF?
Habib [01:23:18] Yeah. So I don’t know if I’m at liberty to say that, but we are going to. I mean, our model provides information for two models and in the PDF the text will be analyzed with two models. One model analyzes the text based on a model text, paraphrasing text, and another model, the text generated text. And this is a really unique think. I’ve never seen such a thing anywhere else. Did the overall which is based on the performance of both of these models. I can give more details about this, but yeah, this is something really unique.
Patricia [01:24:03] How about you? Is there anything you want to share about how in-depth our analysis are for using our AI detection tool?
Chris [01:24:12] I can add on to what Habib said. Without similar to him, I’m not at liberty to say or disclose anything too much. But what I can say though, is the paraphrasing text detection feature is amazing. It’s crazy. It’s it’s like the closest thing to magic right now. As in all, like, paraphrasing. Tools out there are very rampant and they usually, more often than not, get through other detectors. And so we’re really, really proud of what we’ve made in this regard. I’m so proud of Habib especially, and our whole team as well to, you know, that helped make this coming to fruition. But yeah, like just, just to say there’s in the PDF Export Report, there is a feature there where you can actually see which certain standard sentences we think are paraphrased. And that alone, in addition with all the other stats you see on that PDF report, is bound to blow any regular users minds. So I’m really excited for the world to see what we’ve made. I’m really, really proud.
Patricia [01:25:18] I think our viewers can see that our developers are being quite vague and they’re pretty much saying that if you want to see what the PDF looks like and you want to see what the report looks like, you have to go to Apia today and use the tool itself so you can see how amazing Habib and Chris and the whole team at AI Purity has created this tool specifically for its users. You guys want to share where our listeners can connect with AI purity and stay updated. Where can they find us?
Chris [01:25:45] So we’re all in the big social of the top two Facebook, Instagram. You can look at us. It’s AI purity with a space in between. I believe we’re also on Twitter or X or our LinkedIn as well. The best way to reach us is through our email. It’s info at AI security.com for any inquiries or any just questions or even just generic emails that you want to you want us to read and check out. By all means, please feel free, feel free to do so. We’re very active on there and we try to answer each and every single one.
Patricia [01:26:17] So thanks for sharing that. Chris Habib Any final words you’d like to share with our users before they try out the platform that you created?
Habib [01:26:26] Well, I’m happy that we successfully created this rule for our users, and I’m sure that they will be happy using that. And I just want to again convince them to and try it. And then I’m sure that after trying that they can use it to to make sure that what they are writing is authentic too, because this is not something only the professors use or someone wants to use to the case. Someone red handed. No, even a student can can use it to get ensure that what they’re writing is authentic. Because as I said, I believe that everyone should use Chat GPT and other great artificial intelligence based tools, but they should use it in a good way so they can use these tools to ensure that what they’re doing is true. And yeah, that’s it. I’m happy to see in the future how satisfied our users will be with us.
Patricia [01:27:31] Thank you so much, both of you, for being on today’s podcast and for sharing these amazing insights. I’m really excited for the future of AI Purity and I’m excited to see how successful our launch will be. And to everyone who’s here, thank you for joining us on this enlightening episode of the podcast. We hope you’ve enjoyed uncovering the mysteries of AI generated text and the cutting edge solutions offered by AI purity. Stay tuned for more in-depth discussions and exclusive insights in the worlds of artificial intelligence, text analysis and beyond. Don’t forget to visit our website at www.ai-purity.com. Share this podcast to spread the word about the remarkable possibilities that AI Purity offers. Until next time, keep exploring, keep innovating, and keep unmasking the AI, goodbye guys.
