Ep. #1218 - AI’s Ethical Challenge

Matt Watson  0:00

And we’re back for another episode of the Startup Hustle. This is Matt Watson, your host today excited to be joining you today with Saul Liao, who is the founder and CEO of OneMeta. This company does some really cool stuff with AI, which we’re going to talk about today. We’re also going to talk about some of the ethical challenges of AI. And as we see all sorts of stuff with, you know, crazy government stuff and military stuff, and driving car cars that drive themselves like AI and ethics is also a big conversation. We’ll touch we’ll talk about that a little bit today, too. Before we get started, do remind everybody that today’s episode of Startup Hustle is brought to you by Full Scale. Hiring software developers can be difficult, but Full Scale can help you do that quickly and affordably. We have a great platform to help you manage that team. Please visit FullScale.io to learn more. Saul, welcome to the show, man.

Saul Leal  0:47

Thank you. Thank you for having me. It’s a great honor. It really is.

Matt Watson  0:51

So, I’m excited to talk about what your company is doing today. But before we do that, I’d love to learn more about your background. I think you are a serial entrepreneur. You’ve done a few different things, right?

Saul Leal  1:03

Yeah, yeah.

Matt Watson  1:05

So tell us a little bit about the highlight of your career that got you here before OneMeta.

Saul Leal  1:12

Yeah, that’s, that’s a great question. So I’m, I’m mainly from South America. I’ve been here in the US for about 20 years. And one of the things that, you know, as I was, as I grew up, my my family was kind of a military brat, my family moved from different countries, live in different places. And I was, I was not in the same school since fifth grade, every year, it will change.

Matt Watson  1:43

Oh, wow.

Saul Leal  1:44

And kind of that adaptability of connecting with people and understanding different cultures. Keen as, as a great kind of personal spray or or behavioral trait to become an intrapreneur. Because I think at the at the core of entrepreneurship, there’s, there’s a couple of aspects. One of those, in my mind, is empathy, to really understand the problem and solution and extrapolate that through the second aspect that I believe is faith. I believe that at the very core of intrapreneurship is not necessarily belief, but faith can, like, have a stronger knowledge that this song will come up tomorrow, and things will get better. So when combining the two, empathy and faith, I think those are really strong ingredients for entrepreneurship. And I’ve been grateful in my life to have a lot of experiences in those two aspects.

Matt Watson  2:48

I love it. And I think one of the ways I always describe entrepreneurs is they also have to have a lot of tenacity. Right? And I think that’s, that’s part of that faith of like the drive and like being positive and hoping and praying. This is going to work—the tenacity of it.

Saul Leal  3:04

Yeah, I agree, and is have that personal conviction. Yeah, it will happen it’s you act you react at as if it is already there, even though others cannot see it, and that’s kind of the core definition of faith.

Matt Watson  3:20

Yeah, yeah, you gotta, you gotta think two or three steps ahead, too, and drag everybody with you there. That’s correct. So tell us about some of the other entrepreneurial things you’ve done before your current company.

Saul Leal  3:32

Excellent. So one of the company’s I started fairly young. I built a company and installed company back in Venezuela that’s that’s where I’m from. And right after graduating as a as an engineer, and those were one of my my first opportunities and of gain equity of building a solution, especially at technical, a technical solution. And then after, after that, I came I went actually on religious mission for two years to Portugal to learn a new language. And I already knew English at the time, but understanding now Spanish and Portuguese and spending two years, you’re serving people provide tons of empathy and I see that experienced as a foundation for entrepreneurship. Then work in corporate America did an MBA, MIS, Masters Information Systems, here in the US work for a company called Deloitte and then was offer interesting position. And that was great part of my career has been on what people call intrapreneurship. Yes, partnership within organization established organization do you the beauty of is that you’re still given a budget, but you have to build it from scratch. So that was also a great training and I, the alma mater where I graduated called Brigham Young University, BYU, allowed me to start a division called BYU Television. So even though it was an engineer with some background in information and finances, they they show me the give the opportunity to build a full television station in, in a few languages, and the idea was simple just to get it out there for the students. Well, well story short, after a few years, we put it in 27 countries, wow, three point a million households, and we put it on 50 million cell phones across the continent. And that opportunity, that’s when things were really starting to say, Oh, I’m an entrapreneur, I am a builder, I can, I can work on different aspects. I can work with different teams that can work on product creation, I can sell, I can I can grow sustainability, eliminate churn. So that was a great lab to look at the basis of it. So that was one of the most most important aspects of it.

Matt Watson  6:28

Well, and so it sounds like you’ve done a lot of things over your career that have to do with language and communication.

Saul Leal  6:36

Yeah, so later in my career, I worked I built a agency, a marketing agency, and called The Uncle. And it was a fantastic organization with data, inspirational marketing, we also did a lot of the was kind of the golden days of viral videos. And we worked with, you know, with hundreds of millions of views, that company was acquired within a year and a half. And then I came to another company, DMC, and again, as an intrapreneur, as an intrapreneur, we built, a product called the Family share network. And he was Matt, he was just fantastic. We had the opportunity to grow several pages on Facebook from 10 million to 256 million will be well, the largest publishers in the world, in the family category. And we own FamilyToday.com, it’s, that’s the website, we’re in the other languages, about 80 languages, we have family.com in 80 languages in English. At the time was family shirt.com. Today, it’s called family Today.com and building content, creating content, interpreting content, it was, it was a great time to be a publisher during those days,

Matt Watson  8:10

You really have done a lot of different things with with different languages and stuff. So what what then kind of led you to start One Meta? And I guess once you first tell us a little bit about exactly what one meta does today? And what was kind of the inspiration like the light bulb moment for One Meta?

Saul Leal  8:28

That’s interesting. And that, you know, I, I did cover intrapreneurship, as I mentioned, empathy, faith, tenacity, which I will call you on that one in the future. No, that’s that’s a very, very point. I do think it’s a consequence of your of your faith. But tenacity is a very important aspect. But there’s, as you go through that journey of being an entrepreneur, and that was my case, how one MIT is started was, we have a piece of software that we built to create one to many meetings at scale. So a Zoom meeting, for instance, after the pandemic, we all know how these grew in the world technology and that industry really preys on Web RTC events. The, the maximum amount of people that you can get there is about three to 5000 because of the architecture, we built technology that allowed over 100,000 people to be connected on an event. So Wow. And that was that is a fantastic piece of technology. As some having connections with potential clients and discussions, they realize that after 1000 people, language was challenged. And it made sense. And as soon as I saw that problems like oh, I can solve that. Any entrepreneur will say Oh, I can solve that have some experience on languages. And as we build on started building that engine, it started realizing some of the challenges with AI. So at the time, this is about two years ago, we connected with to Microsoft to be connected with a company and our service called Open AI. This is our two years ago, right and, and this was the very beginning of the of the faces of what we know now as a chatty btw. And as we start connecting and looking at Microsoft being an incredible partner of ours from the very beginning, as we are putting all those together, we’re seeing Oh, there’s more to the model to the architecture than just the AI model, the AR models were extremely strong, you have to do a lot of training, we provide some proprietary and patterns on how you train it, that became really, really strong. But realize it’s not only about the AI model, there’s a pre AI works that needs to happen at post AI where work needs to happen. So we start working, let me give an example noise cancellation, right? Your voice isolation, which is actually identifying who’s the speaker, detaching it from the background, and increasing the resolution of the cell. And as we start working on all of those aspects in that architecture, we saw an increase of accuracy on transcription from 92% to 97%. And it was consistent. This look, go ahead.

Matt Watson  11:33

So when you originally started this was it, you had already built a software that was for these large events? Like you’d already built that software before you started doing the language part of it.

Saul Leal  11:43

Correct.

Matt Watson  11:44

Okay. So we’re, what kind of events were those? That would have like, 100,000 people?

Saul Leal  11:50

Good question. There are several of those, but I was I was looking into multi-level marketing or direct marketing companies. Okay. So it’s specifically, you know, companies like new scheme, or doTERRA, or Amway, where they have distributors, and they want to connect from all over the world. And they have millions of individuals who want to get that in in one session. Okay, pharmaceutical industry, where they have, again, 10s of 1000s 40 60,000 doctors across the world, they want the simulator a specific type of information. Yeah, it makes sense. And these organizations are not only massive, but they need a secure system to do so. So that was kind of the nature was going to, well, I build this technology, I was going kind of a rabbit hole with those. But as they build these pre-AI and posts AI architecture, I put it and they start presenting to people. And this was kind of the aha moment, Matt, people start saying, Oh, we want to buy the product. We like the whole architecture, the events, but these little features that you have on interpretation, translation. It’s so fast is so incredible. Can we buy that today? Can we get this tomorrow? Like, like, we’ll do a contract with like everything. But can we get this today? Like we’ll use it today. And Matt, that starting happening one meeting, two meetings, three meetings, four meetings, five meetings, six meetings. And that’s when I started realizing that we have something very magical. Then we took it to Microsoft engineers, and they freaked out, they say, how is it that you’re using someone else’s model on other servers, and you’re faster and more accurate than us using it directly? So then I went to other organizations at that level, and their feedback was the same. And that’s when we realized we have something very, very special. And then we start building a whole architecture of gait. How do we go to the market? How do you create a category through this? And we have changed kind of our mission and vision throughout the beginning without the world would probably our translation company or interpretation, then we say, oh, no, we’re much more larger. As we work with hospitals, we start seeing that the budget of translation was going to the doctor. But when we provide our service because it’s a lower budget, now the consumption of translation interpretation was going to the nurse who was going to the CNA, and they saw Oh, the need is much more larger is just has been constrained by the budgets. Then, and we and we say okay, we’re in a multilingual environment, we enable multilingual conversations. We’re increasing that market. And after a few months, we realize that much more than that. I’ll talk more into it. But ultimately, we are in the understanding business

Matt Watson  15:13

So, so let me ask you this is the is the core business translating, really doing transcription of the audio to text or you guys are also doing converting audio from one language to another language and audio.

Saul Leal  15:28

So we do all of it. So we would do a speech-to-speech, audio-to-audio life interpretations, or I’ll talk to you in Spanish. I will try it here in a bit. I’d like to do in Spanish, you Herman English? You talk to me in English. I hear you in Spanish on the fly in 150 languages in Waco nation. We even put that technology man, it’s so magical. In a phone system, I will call you while I’m walking on the streets of Shanghai in Chinese. And you will be walking in the streets of Germany in German. I will call you in Chinese, you will hear me in German. No phone. Sorry, no app, no internet. Just amazing. Use a phone number. I will call from my 801 for number two, your 212 phone number. And it just works like magic. No, no internet, no app. Anyone can use it, HIPAA compliant, sought to compliant. All security measures, use words. A lot of people might think that inclusivity and a half. We can talk a whole show about inclusivity. Or did you have race, color, skin religion, and yes, those affect but the number one factor of exclusion, Matt, that no one is talking about?

Matt Watson  16:52

Its language. It’s your accent, right? Yeah. And

Saul Leal  16:56

his language overall, like the fact that you’re not even speaking, the language that is excluding you to understand and to participate. We saw that.

Matt Watson  17:08

I saw I think I’ve heard I don’t know if there’s other companies that are doing this. Or maybe it was you guys sounds like this something you guys do. But having having essentially companies kind of call center employees that work in India or wherever it is, that can speak you know, sort of okay, English, but having a rough accent or whatever. And then your guy’s software basically takes English to English, but without the accent.

Saul Leal  17:31

Yeah, that’s not what we do. There’s another company that focuses on that, specifically, we could talk about ethics for whatever you want to because that is, I think that touches on the ethics and personality aspect. They, anyways, it they that’s, that’s not what we do we do more of the language specific. Okay, we would do it on the on the fly, we process content and context and sentiment analysis on the fly devil with the accent, I think he has a lot of merit into it. But But there’s more that you can achieve.

Matt Watson  18:12

So what is the delay in your guys’s technology? Like, so you’re talking in Spanish or whatever? What is the delay before I hear the English?

Saul Leal  18:21

It’s 1/8 of a second. Wow, that 0.1 25. Now, in the demo that I have right now because I do the demo with a phone number. There’s two type of interpretation, you call it consecutive or simultaneous. Simultaneous interpretation is, you know, a word or two delay, in our case is not even that is 1/8 of a second. And consecutive interpretation is you say a sentence that the sentence gets repeated in the oral language, the other person says their sentence and the sentence could have been your language because it’s a phone system. And because of the user experience, we decide to do consecutive interpretation on this demo. But again, we’re in a live event on our software, you will have simultaneous interpretation with less than half a second delay.

Matt Watson  19:18

Well, that’s amazing. Because I mean, I don’t know about every language in the world by sure know about English. A lot of times the way that you say things and the words and all that the meaning can be totally different. So I would think you’d have to like complete sentences before you can figure out like exactly what they were, like how to translate it, you know,

Saul Leal  19:37

That’s a good point. And that’s the kind of the beauty of these these models that we’re seeing these large language models and these predictive models as well. So we put them all together, and they’re phrases like raining cats and dogs, for instance, that you understand, I understand where it is because of the culture and how we grow and language, but you’re trying to translate that literally, and just doesn’t make sense. So those are, that’s when we take things in context. For instance, we can, when someone uses our technology, they know time of the day situation where people are being connected, so on so forth. So you can almost predict, oh, they’re gonna say good morning, instead of good afternoon. So now you’re you, you what you want on this models, especially ours with our specific functionality, is you want to get the information and predict with high level of accuracy what it’s going to be set, we did native language before you translate it.

Matt Watson  20:38

I love it. This is this is super, super cool. I do want to take a moment to remind everybody that finding expert software developers doesn’t have to be difficult, especially when you visit FullScale.io where you can build a software team quickly and affordably. You can use the Full Scale platform to define your technical needs and see what developers are available to join your team today. Please visit FullScale.io to learn more. So tell me, where would where would somebody like me where would I see your technology? Where Where would I use it? Is it baked into other things that I would just never know about or is it a consumer facing app? Like how would we actually use your technology,

Saul Leal  21:14

All the above. Our strategy of go-to-market and I think it’s something important that interviewers need to think about I’d certainly they do we do I do is go-to-market technology. So how you build something at scale in what different areas and how you move towards it. So our technology can be used by other developers. So we created an SDK or API, where any developer can just connect to our technology and in three lines of code. And without hundreds of 1000s of dollars of investment, they can do commit to services through our technology. And just implement everything, they can put it on a chat. And you can write on a chat without touching anything, the city will recognize the language and said to the other person without touching anything into their native language. And they can have a conversation with 234 or 510 people or more in the chat. And nobody will know that everybody else is speaking another language. Everybody will consume the language in their native language and just happens real time. So you can do that with in three lines of code. That’s kind of one area that developers can use or technology, and other companies can use it behind behind the scenes. And you will note that it’s one meter. And you can do that with also a speech to speech, or a speech to text or text to speech or translation. So we offer all that, we also offer plugins to call centers to your point. So call centers can actually connect or technology pick up the phone and that agent, that subject matter expert is able to speak 152 languages on the fly. Then we have all technologies where you can go to consumer, just go to our website, onemeta.ai Look at the different products and you can buy a video conferencing platform that allows you to speak in different languages on the fly in a multilingual environment. Someone can speak German, Spanish, Polish, Arabic, Hebrew and use just talk and everybody will understand each other. Yesterday, I was talking actually with a gentleman from from China, he lives in Palo Alto, and we’re just playing around his English was actually way better than my honestly, it was just fantastic. My English guy was soil. So I told him, Hey, you want to speak Chinese, and I’ll do Spanish and that you’re not gonna do that we’ll talk for over an hour. And he’s in Chinese, I speak Spanish. You were so fluid. It was it was fantastic. was magic. All really was. And then we have our technologies on events. So you walk into 100,000 people stadium, and you QR code. And you can look at the announcer of the game or event or the training in any language with subtitles on your phone, or you can put the earplugs and just listen to it.

Matt Watson  24:07

That’s very cool. And that’s a great use case.

Saul Leal  24:10

It really is. And the last one is the phone system. You can go to our website. Yeah. So there’s a phone number, you get a phone number and that is you set it up is going to be Spanish-German or English-Japanese. And you set it up and then just call that number and then call your friend who speaks Japanese and doesn’t speak English and then you just talk to the phone and the merch call the other the third party AI interpreter will just translate the whole conversation on the fly. Schools are as very cool. They love it this cool use cases for school days fantastic.

Matt Watson  24:51

So as your guys’s software built into Skype or any of those kinds of things as well?

Saul Leal  24:55

That that’s a good question. It isn’t yet. It is in a roadmap sooner sooner rather than later, I will make some announcements, some, some great things will be happening fairly soon.

Matt Watson  25:07

And it makes a lot of sense to be built into Google meats or zoom, or Skype, or all these sorts of things would would take a lot of value in it. So I’ll be right. So, this whole time I’ve been sitting here wondering, I have to ask you like, so, at any point, you guys consider changing your name because of Meta because the stupid Facebook like what out? Is that? What do you guys think about that?

Saul Leal  25:31

You know, the name was was interesting. As we’re building all of this, Matt, I realize that what we can do, and I think a lot about languages and about humans, and about AI, and they spend a lot of my time into it, ethics of it. And ultimately, what I want to build is one universe, one world, yeah, where we all can communicate. And that’s how the name came about. Okay? Meta means goal in Spanish, okay? Or universe in or place to be in Latin. And that’s how we came. So we expend a lot of resources with our, with our legal team. And to be more specific, the fact that it’s not at the beginning allows us to don’t have a lot of conflict. We’ll see. We hope that we won’t have any conflict with any other company out there. But that’s what we started with Rivoli, we’re looking look at the risk and seeing since fire up to this point, and I don’t see any problem

Matt Watson  26:41

Doesn’t create any, any confusion because Meta and knows meta being better than so

Saul Leal  26:48

As long as long as we keep it all together. And it’s one word, OneMeta? Yeah. Then it’s, it’s fine. Yeah, but things are tense for actually, that’s a good question. Were you?

Matt Watson  26:59

So did you guys already have the name before they changed their name?

Saul Leal  27:03

No, no, it was after the fact. I mean, I already had it on my mind written but it was it was after the fact.

Matt Watson  27:11

I was just curious. Oh, that’s a good question. I’m sure. I’m not the only one curious. That’s listening. So there’s a lot of

Saul Leal  27:18

people that asked about that question.

Matt Watson  27:19

So let’s talk about the ethics part of AI. And, you know, obviously, AI is being used for a lot of different things, right, like military related stuff, no doubt, and places like Ukraine on the battlefield, right? AI can identify objects and track them and do all sorts of things. And you’ve got, like you mentioned earlier, AI of removing people’s accents. I mean, you have lots of AI stuff related to driving cars. And, you know, people, you know, the classic example of this for the self-driving cars, right? It’s like cars going too fast around the corner. Is does it? Does it kill the people walking in the street? Or does it kill the driver? Right, like, but one of these days? That’s like a real world scenario that will happen? Oh, yeah. Yes, I agree with. And but my question for you is what you know, for in your, you know, you’re the founder and CEO of an AI company, how do you how do you think about ethics, when it comes to AI?

Saul Leal  28:23

The most important thing is to realize its place, Matt, and realize its purpose and what it does, very similar to algorithms, and to behavioral or regression algorithms. A lot of people will say that, you know, the algorithms of network effects, Instagram, Facebook, or, you know, TikTok that those algorithms change. And a lot of I wasn’t that world for many years. And as we build advertising and content, and we’ve tried to disseminate vital pieces of content, people will go back and say, Oh, the algorithm change. Well, reality happen is that people change. Algorithms is an extension of human behavior because they are attacched, they’re attached to human behavior. And when we talk about AI, I think of AI something very, very similar. And of course, there are two sides of the tilgang could go on both sides, but I’m up. I’m optimistic realistic towards AI. And what that what that means is that I believe that AI is is basically a superpower. I had, we live in this space that I’m in which is language understanding, including interpretation transcription. There’s a lot of human jobs that people are I know, white collar families in Latin America that the older leaving comes from interpretation, right? And I think I’m Matt and I, and these keeps me up the night is, am I going to It’s makeup, I got to take their jobs, who’s going to pay for this school, the private school of that family of the kid have that belongs to that family? And that because I know that, right? I know it takes time. And there’s a time and place for everything. So, a client of ours, asked us sold your company is a huge competitor for us in our industry. And this was my answer. I said, No, we’re not. Your number one competitor is your employee who owns the relationship with your clients. And if you have a view have a non-circumvent or non-compete with them, they will hire our services, and they will take your client that is your number one competitor. Your second competitor is your competitor, who will hire our services, and we’ll take away your clients. We are not your competitor. We are, we give superpowers now does the world need 7 billion Superman’s no it doesn’t to save the world? It needs a few Superman’s and, and I think that what AI is going to do at least in our space and fully kind of office tasks. And there’s a lot of that is that is going to segregate the business is going to disintegrate monopolies, which is in some cases a good thing when we talk about in the internet, we thought the internet was going to be the greatest equalizer of of income is available throughout the world. And if we look at the United States, they actually created more and more monopolies. The internet created more monopolies in terms of socio economic income, then than before. Like we are more of the 1% like it actually monopolize a lot of the resources. But when we look at the internet, across the world, it actually fulfilled its purpose. It was a great equalizer. We helped economies, we have people that increase their average income, thanks to the internet. Oh, yeah. So I think that this scene is going to happen with artificial intelligence, where now you have a virtual assistant in Indonesia bots, and these virtual assistant is you’re paying them $300 and is doing the work of one. Well, that third to assist. And now through AI is going to become a manager, and is going to now to charge $1,200 because he’s going to have for AI virtual assistants now that virtual is human versus becoming a manager. The question is what’s going to happen with a supervisor here in the US or in first world countries, that is hiring the staff outside, who actually has increased their capabilities because of the superpower AI? I believe that that individuals in first world countries will now have their own company and compete with their employer. And then the monopolies will be set these disintegrated or disseminated. That is kind of my approach approach to it. So I think that that’s, that is one way to look at the technology, how is going to help us all, I think is going to provide a lot more opportunities.

Matt Watson  33:29

I think a lot of these technologies make it so it’s available for more people, right? So there will be more people that will be able to do translation related things that just weren’t able to do it before because they didn’t have a translator. They couldn’t afford to hire a translator. They couldn’t do the jobs like you described. It just it makes it more easily, you know, available to a lot of people that couldn’t afford it before. Right. So you know, obviously we have tools like other examples, we have tools like Photoshop, we have so many like Canva, who creates like an easy way to edit photos, new different things. It didn’t put all the graphic artists out of a job. It didn’t put Photoshop out of you know, out of business. It just means I can use Canva now.

Saul Leal  34:14

Most exactly, you said it right when we think about the computer, the computer replace typewriters, it did did some manufacturers went out of business because of the computer some type writing, like writer manufacturers they did went out of business. But if you look at the amount of authors and coating creators that pure allow the amount of content that is now being created because of the computer that is much more value for humanity and humans will will adapt. So I think there will be some areas where we will adapt but the reality is that the the increase of of content creation We’ll, we’ll be substantial in the years to come. I think that the core is, what are the ultimate objective of all of this. And I think that’s a little bit mad where we may have it wrong. In terms of a we just focus on creating quality of life and what quality of life means are more focusing to bring you financial resources to the NDB, on to one, and that that is where there’s a shift happening to the world, especially to the pandemic, where we’re focusing more on what matters the most for the individual. And we’re allocating more and more time to things that matter the most, when parents and that we’re at work, start spending more time with their children. That’s when we realize, oh, this is a whole focus of businesses and capitalism is so I can spend money buying ice cream to the ones that I love on a Friday afternoon, or go and and we sometimes lose that aspect of what what matters the most.

Matt Watson  36:03

So let me let me ask you this. So one of the other big issues with AI, especially we see this with open AI is all of the sort of ethics moderation part of it that has to go into it. Right? So how do you how do you guys deal with that part of it, where somebody in one language is speaking something that, you know, like, cussing, bad language, terrible things, whatever it is, like, how do you guys have to deal with that part of it? Are you have you guys translate that part of it deal with that? Do you guys have to, is that’s something you have to deal with?

Saul Leal  36:32

There’s a two part answer to your question. There’s one about the bat, Ward’s translation more about tactical aspects of it. And then you talk about moderation of the model. Yeah, and then the answer each of those on the first on the first one, it’s called CCD, that’s kind of the AI name, or language, when you refer to those, and there are specific models, where you’d recognize as being pronounced and hash it, you change it up by an alert, and you can actually provide a filter to the interface. Okay. So before it gets transcribed before he gets hurt, you can actually create mechanisms to protect those environments. Now, toxicity Interesting enough, is not only about that language, in terms of the specific bad words, but it’s also about the tension. So if I start saying something, even in sophisticated words, and I’m bullying you, I can catch that and I can catch it before it’s the message being sent.

Matt Watson  37:32

And basically, you have like settings to do that or not do that.

Saul Leal  37:35

Yeah, correct. And again, that is called toxicity, there is a component with kind of this same methodology that is called personal identifying information PII. And that if someone says, you know, their social security number, or verbally, they will be cash, and it will be transcribed to protect information to keep a compliant approaches. So that is that aspect is very tactical, we

Matt Watson  38:02

do have some of that you need to translate, right? Like if I’m calling my mother-in-law in the Philippines, and I’m trying to give her my social security number, or I’m cussing and I want her to hear it like I would want to translate it.

Saul Leal  38:14

Oh, yeah. In those cases, you can, you can turn it on or turn it off.

Matt Watson  38:18

Yeah, but a call center maybe wouldn’t, right, you’re like, No, we don’t want this.

Saul Leal  38:22

Correct. That’s exactly what you say. Right. The other aspect and probably the most interesting aspect of your, of your question was the moderation. Right? Yeah, you level and they only share a couple things. And they have to do with with the ethics of AI. Let’s say that we have the best couple that are you know, I’m from Venezuela, so I’ll talk about baseball. So we have this couple that are the best baseball players in the world. Husband and wife, Mel V when they marry and they conceive a baby and this baby is said to become the best baseball player in the world. Okay, they these parents hired the best coach in the world and train this baby to become the best baseball player in the world. He grows and they they put the becomes really strong has all the to all the knowledge, all the training, and then they put them in an environment at the World Baseball playoff. They do the perfect environment, the perfect team playing with the perfect competitors, and they win the World Series. And these because of this and ultimately, these baseball player becomes the best one in the world. And he wins a prize lets you set up an amount of a million dollars wherever it is, to Who does it belong the money? Does it belong to the parents, who conceive the player? Does it belong to the coach who train them? Does it belong to the environment and the competitors that allow him to become the best one? So interesting of the legal term, is that over, it’d be the answer is easy belongs to the baseball player. But this Vegas thing, how we build the legality and accountability, and the entity of, of an intelligence. Whatever that intelligence does, regardless of who created them, their parents who taught them their coach, anything as an entity, the work that they do belongs to them. Matt, this is at the core of the dilemma with AI. If you build a model that can stand on its own, and can build something and create something, does it belong to the individual or issue that create the model? Does it belong to the person that train it, but they belong to the environment? Because they’re so in order to come up with that output? There’s tons of variables. And this is part of the ethical dilemma of like, Oh, I’m building a model. And those the outputs as these models Green Book, does that belong to the person that did the prompt? Does it belong to the people that build the model? Does it belong to the one who trains it? And these are? Or does it belong to the source of the training? So this is part of the questions that AI that we are as a community trying to figure it out? Because there’s some legal aspects, as I mentioned, to what belongs to who does that make sense?

Matt Watson  41:57

Yes, it’s usually the vendor who created the model, right? I mean, at this point, they’re the usually the ones that end up owning, owning all of it.

Saul Leal  42:04

But that’s my point. But it’s is that the right thing? Because they didn’t when they train it from, from check speed or rail, they trained from a specific singer, that style singing, that didn’t belong to the person created model that belongs to that individual who actually is the source of the training.

Matt Watson  42:23

Yeah, their input their input into it. So you got this, like somebody like Reddit and Wikipedia and all that, right? Complain, like, Oh, you took all of our content, and you created this model, but it was our content. Right?

Saul Leal  42:35

Yeah. And that’s where you started.

Matt Watson  42:37

Yeah, that makes total sense. Well, I do want to remind everybody, if you need to hire software developers or leaders Full Scale can help we have the people in the platform to help you build and manage team of experts when he visit full scale.io can give us a few questions about what kind of developers you’re looking for, what kind of technology skills you’re looking for, and we can show you exactly what developers are available to join your team today. At Full Scale, we specialize in building a long term team that works only for you. We do staff augmentation. And we have hundreds of employees. Check us out at FullScale.io. Well, this has been a really fun conversation. And I love AI and it. Have you seen the movie The Creator yet?

Saul Leal  43:20

No, no, I haven’t.

Matt Watson  43:22

So it came out like a couple weeks ago, and my 14 year old son wanted to see it. So I went and saw him. But it was about how like, basically, we’d created AI and we had like robots and these AI robots, you know, eventually, like, take over the world begin a big fight with them. And they nuclear bomb goes off and we’re fighting with the new the AI robots or whatever. You know, that kind of dystopian future we’re fighting AI. But it was it was super fascinating. And it it does make you think about man, where is AI going? And at what point in time, are we, you know, you’re building the software, right to take this software you built and put it in a robot and now a robot walks around my house? Like, how far away is that?

Saul Leal  44:00

I don’t think I don’t think we’re that far. And at the end of the day is how do we use it as implemented. It’s a tool. It’s an extension and it’s also the intent. Yeah. And the question for me is who is getting access to technology with the wrong intent? And I think that’s part of it. As we as we speak, you remind me of something that we haven’t done on I wanted to just do a quick show of the of the technology.

Matt Watson  44:36

So I’d love to see a demo.

Saul Leal  44:39

Yeah, you can hear it fairly quick and that our audience can get a little bit of a taste of it. So as we’re talking, I am going to my phone, and I am calling a phone number. And, and I’m clicking on it, and it’s a it’s a demo on consecutive interpretation. Okay.

Speaker 3  45:05

Hello and welcome to Fairbank call. You have 90 seconds to experience our over the phone interpretation trial. English and Spanish are already set up. Please begin speaking after the tone

Saul Leal  45:18

Excellent out of Turkey Camino amigo Matteo. Yes, Tom haciendo negative assume yes, click on the book with Mara, we’re also taking a look here.

Speaker 4  45:31

Excellent. Now I’m here with my new friend Matteo. And we’re making a recording and explaining a little bit of our wonderful technology.

Matt Watson  45:40

That’s very cool.

Saul Leal  45:44

Year. That is that that is fantastic. And this is one of the technologies that we have. And as we built this technology, something that we did, Matthew, is that not only within 152 languages, but we did several technology advances in terms of putting more than one language on the same microphone. So we can we can try it again. But basically, I will speak Spanish or German or English on the same microphone, and it will just recognize and translate it. Now, this is what happens behind the scenes. There are about 10,000 permutations for every millisecond as we’re pronouncing the words, and we’re matching the Deaf form of the wave to a phonetic dictionary in English. Okay, so we’re saying hello, and it’s matching the word hello as it’s been for nouns into that phonetic pronunciation of the word English Hello. But if we add another language, it’s Spanish, let’s say. And we’re saying hello. It’s matching two languages to the phonetic dictionary. So it goes from 10,000 per millisecond to 10 million per millisecond and grows exponentially. And we build certain technology that we can do up to five languages on the same microphone. So I could be in a panel at an event and talk in Spanish, and I pass the microphone to someone to talk in English, and they pass the microphone to talk any Japanese or Italian. We will be working on a new client, it’s the Vatican. As we were working with them, it was fascinating to give the presentation they were talking Arabic and French in the same environment. And that to see how not necessarily dilution or eliminating language barriers, which is kind of a cliche phrase, but it’s more of a building the bridge. And we talk about dismantling language barriers. But in reality, it is way more than that is going to bridges. We focus so much on the words that we forget about the intention. If you think about it, Matthew, right now, as I’m connecting with you, you’re connecting with me. We’re talking with our audience. We’re using words as the protocol to converse. But really, what we’re trying to do is just, if you think of it as is, it’s a way to encapsulate our thoughts and emotions. And this is what we’re discovering: that our company creates a more understanding world because we’re connecting thoughts and emotions from the one, two thoughts and emotions from the other. And that’s ultimately what AI has been helping us to do.

Matt Watson  48:46

I love it. It’s awesome. Well, I appreciate so much having you on the podcast today. Reminder, buddy. This is Saul Leal. His company is called OneMeta, it’s OneMeta.ai. Sounds like you’ve got some free tools are free, a free demo thing. And some things are probably relatively inexpensive for consumers. And then sounds like your software is built and all sorts of other things. Right? So probably a lot of different use cases for this. I want to try it and see if I can talk to my mother-in-law. I don’t want to figure out if you’re if you support Cebuano. That’s her language, so.

Saul Leal  49:20

We do have Cebuano

Matt Watson  49:22

All right, here we call my mother-in-law. That would be fantastic.

Saul Leal  49:27

Actually, we can do that right now if you want to.

Matt Watson  49:32

She’s asleep right now. But I may I may try this later.

Saul Leal  49:37

I’ll we’ll have it, and you could go to a website, or we can use a phone number. It will be fantastic. We’d love to hear more about that experience. You will not

Matt Watson  49:45

I’ve never been able to talk to her before because she doesn’t speak English. So. All right. Well, thank you so much. As we leave, do you have any final words of wisdom or tips for other entrepreneurs out there? It could be about AI, or it could be about entrepreneurship, or anything else.

Saul Leal  50:00

No, it’s It’s focusing on what matters the most. I think that, indeed, intention really matters in life and understanding that one and, and just empathy, faith, and tenacity.

Matt Watson  50:13

I like that tenacity.

Saul Leal  50:15

Me too. Thank you.

Matt Watson  50:17

All right. Thank you so much for being on the show today.

Saul Leal  50:19

Success. Thank you. Thank you all.