Why AI Companies Are Selling at a Loss on Purpose

**[00:00:00]** But the tech strategy was always give away their products like the drug dealers give away a free bag of crack. That was like Oracle's sales strategy. And we're seeing that now with AI. They are trying to get you hooked. You know, once you're hooked, what are they going to do? They're going to put up the price. But that that time you built all your technology with is deeply integrated in the back end. Hello everyone. My name is Armando Perez-Carreno and welcome to the Web Talk Show. Today with us is Joshua Gould from The Big Word. How are you Joshua? Great Armando and thanks for having me on your show. It's a pleasure to have you. Can you tell us for our audience what you do at The Big Word and how you got here? So, uh, I am the CEO of The Big Word, but I did not start out life as the CEO. In fact, when I was 18, and I won't give you my whole long history, but when I was 18, I went to college and I sat there in my first lecture and my thoughts were, "Oh my god, how do I do this for 3 to four years?" It's a lot of knowledge and not a lot of understanding. and I just wanted to get out there and make money and be in business and talk to people and yet I'm being talked at and I've got to retain all this information and my memory is not that great anyway. So I took another two weeks to uh pluck up the courage to tell my parents that I'm going to drop out. So I call myself a college dropout, but I only went for two weeks. So, um I don't know if you can consider yourself a college dropout then, but I took a job because my dad said, "I'll pay your rent for another week. After that, you're on your own." And it wasn't punishment cuz my parents were not necessarily big believers in in the university or college system, but it was just the way **[00:02:00]** I was brought up that you have to take responsibility for yourself. So I took a job um out of desperation with cause and cause being the beer company and I learned to sell. They gave great sales training. So from there I took a night shift working at The Big Word call center selling beer. Uh sorry not selling beer. I was selling beer at cause selling language services, mainly translation services to the banking system in the US. So it was a night shift for me, but it was a day shift for everyone that I was selling to. And that's how I got started in The Big Word. I worked my way up, got onto the dayshift and then became the shift manager and all the way to CEO of the Big Wood, which um, you know, wasn't that hard back in those days. It was 50 people. It was actually a family company uh for me. And um I never wanted to work with family. It was not my intention. It wasn't my father's intention. But uh I was selling and I was enjoying it. And then after a couple of years, I decided, you know, I had been to every fun party in London and it was time to go and experience the parties in New York. Of course, I told everyone it was strictly for business reasons and we're expanding the business, but um nobody, you know, it worked out well for a for a hot minute, as they say, because I got to New York and I sold deals to HSBC, to UBS, and a bunch of other banks and then we had the great recession and um I actually worked on Wall Street and I lived on Pine Street, which is just behind Wall Street. So, I was literally watching people walk out every single day with boxes and I would see clients **[00:04:00]** leaving with their box and they'd be like, "Hey Josh, how you doing?" I'm like, "I'm sorry to see it. Um, just remember me wherever you land, give me a call." But the banking industry, which we were selling translation services to, so that was like equity content. So, for traders, Mhm. you get all of these um equity research reports and data insights coming from all around the world in all different languages and they needed it back within 40 minutes. So we would build these like automated workflows um where you could have multiple translators work with an editor and integrate through a SOAP compliant API which we still work with today. Um, but the content management systems are a lot more advanced than they used to be. Uhhuh. And we would suck all that information out. And that was the sales pitch. But I quickly realized that the banks didn't have any money to spend with us. U so I was looking around who's going to like help me pay my rent. And that was the United States government because George Bush was in power and putting in trillions of dollars into the US economy to overcome the recession and handle the war. So we pivoted the entire business to the public sector and healthcare sectors which we knew would continue to do well and and that proved to be a great business decision and opportunity. So we the cool thing about that business is you don't have to go and find opportunities. They just get advertised. But you do have to build out teams because you're now in a very regulated space. It's got FAR regulation. You've got to get on the GSAs, you know, which is another kind of network of government buyers. So we did that and uh we grew the business every 18 months. we were growing it and we had all the AI **[00:06:00]** disruption because machine translation came along which was uh a neuralbased machine learning platform and it took about 70% of all of our words and basically made them close to free so we had to overcome that and we did and then I went into the interpreting space and the difference is interpreting is the spoken translation is the written word outside of the industry, we just everyone calls it translation. Um, but the interpreting space was spoken and that was like an Uber model before Uber even existed. So, we would send people in person. Then we added the video interpreting and the over the phone interpreting and that blew up and now we were becoming a multilingual communications business with you know with voice and uh and visuals and also then written uh and we just grew from there. We we took on military contracts. We were out in Afghanistan and Iraq and Syria with humans. We're providing Medicare and medic uh and Medicaid uh over the phone interpreting as we still do today from this office that I'm sat in now in Miami. And we're in like 50% of the police stations in the Netherlands. Uh about a third of all people arrested need language support. So, and we do every single court in the UK is contracted uh with The Big Word. um a good chunk of the uh police stations there and health care systems and I believe every single police station sorry about 50% of police stations. So that's what we do. It's missionritical work. It's fully tech enabled. I it's got full AI tools built into our technology but we build AI self-driving workflows and we've actually grown because of those. We've **[00:08:00]** taken our operations cost which used to represent about 20% of all of our revenue down to five 5%. So 20 cent in the dollar was just admin like for operations. Now that's down to 5 cent in the dollar and we done that through uh workflow automation and and really good tools and that's allowed us to go out and win. So that's how I got to where I am. Somewhere in there, my father retired and uh I begrudgingly took over the business alando and that was about seven years ago and uh um I was terrified that I was going to screw up 40 years of his hard work but the business essentially has tripled in size since. Wow. I hadn't considered all these required language services, but that's this makes complete sense especially like you're saying all the arrests uh especially in countries that have a lot of migration and it's how are they going to support that especially in smaller communities. So this this this is a whole new market that I had no idea existed. Yeah. I mean, it doesn't matter what side of the political aisle you're on. If you uh want to get rid of people who you deem not American or British or Dutch or wherever, I mean, you're all for the interviews because you need to get rid of these people. If you want to bring them in because you want their talent and their capability and workforce, you're for the interviews because you want to interview them and bring them in and make sure that they're coming in safely. uh and responsibil uh responsibly. Uh you know, so we typically do well regardless of who's in power. Um we live in this digital world where it's borderless. So you can't close a digital border and that drives revenue to us as well. You **[00:10:00]** can technically commit a crime in a country that you've never been in just by the internet. You can breach copyright laws. So think about, you know, all the noncriminal uh things that go on. And we in the westernized world have systems where you have a right to understand what you've been held accountable for or if you're in a hospital, you have a right to understand what the doctor's saying to you. So we provide those services and they're missionritical. 99% of all content that's translated online is done by machine. We focus on that 1%. But here's the cool thing. Every 15 months, the entire content doubles. So think about that for a second. Go back well before Christ. So thousands and thousands of years. If you take all the content, it right up until this point that we've ever created. In about 15 months time, it's going to double again. and another 15 months time it's after that is going to double again. So that 1% is growing. So our industry despite it being completely and radically changed by AI since 2013 when neural machine learning came out which is what we now call AI. Um the big difference is we've got LLMs which is generative large language models but we got just as good outcomes before in smaller corpora. You know even then our industry still grows around 6% revenue a year. Wow. A lot of people don't understand because their only exposure to AI is today LLM. that's their that's their first dip into AI and so they don't quite grasp just out of regular common **[00:12:00]** ignorance not not intelligence or or other it's just you haven't heard of it there has been AI in very high use for a long time like you were saying machine learning neural networks etc what is the main difference between what you would do maybe eight years ago versus what you've been doing for the last four years with with LLMs. Yeah. I mean, first of all, I I would just say it's a whole branding exercise. We've been using AI mainstream for decades. My grandfather was a tailor. He was uh 93 when he passed away, but he retired at 60 because a machine took over what he did. If the machine isn't artificial, I don't know what it is. It was intelligent. it could make a soup, a whole soup. And you know, instead of having him do it, a machine did it. So, we've seen these AI revolutions on a consistent basis. The internet, the dotcom boom and boss, that was an AI revolution. Uh, your thermostat in your home is artificially intelligent. You tell it what temperature you want the air to be, and it then goes and tells. And it's artificial, right? It's not a human. It then goes and tells your boiler how long to stay on or your your or it will take your AC and reverse the cycle if you live in Miami like I do. So, you know, we've been using it all the whole time. And The Big Word since 1998 has been using rules-based AI, which is exactly the same as LLM AI. the with one big exception and and that is and this is more technical we had to tune it with content that we created or found whereas now in the world of large **[00:14:00]** language model AI it reads the internet it watches every single YouTube it's going to listen to this podcast and it's going to improve itself that way it has a massive issue and an Achila's heel that now the majority of content is created by AI and therefore it can't really get that much smarter because it's just tuning itself on itself and that that's a that is a major issue. So you're getting all these companies crop up like The Big Word who is saying we own billions and billions of words and minutes of interpreting and translation that are being done by humans and we can tune it to get a better result. So, while you may get really good translation, if you go on to ChachiBT or or Gemini or Anthropic, um, and and you're trying to speak to your mother-in-law who like mine doesn't speak a word of English, and I'm trying to get her recipes for Jacn and, you know, and, uh, all the things that she creates. She she does a lot of Yemenite and Tunisian cooking. um is great for that, but you try and have a medical conversation or a or a legal argument, then it's not so good. And think about it. Let's say, God forbid, I'm under you were incorrectly accused of murder and you're you you're on trial and you have a a witness who doesn't speak English. Is 95% accurate good enough? No. you know, you in that witness has to be 100% sure to at least a reasonable doubt level and therefore it's not good enough. And in the police station, those witness statements need to be accurate. And you know, cuz that's what attorneys do. They look at whether the witness is changing their stories, whether the accused is changing their stories. And it's not **[00:16:00]** just in criminal law. What about in family law where children need to be heard? It's not their fault that they were brought to the US. It's not their fault they don't speak English. They're just children. And you know, we need to understand, are they being beaten? Are they being well cared for? So, we provide all of those services. And then the average immigrant that comes into the United States and the UK is a huge boom for the economy. I mean, I'm an immigrant myself. You I'm from England. I came in here and I paid many, many millions of dollars worth of tax. I mean, I like to joke that uh I'm not quite as good return on investment as Elon Musk, but I'm pretty good, you know. Uh we we both came in here, him being an African American, me being a British American, and we created jobs. I created thousands of jobs here in the United States. So again, I'm not pro or against immigration, but The Big Word exists to grease the wheels of the nation. And I like to joke that we are the company you've never heard of. Now where we can to answer your question now more directly, we are now able to provide all of the superficial language services that makes it easier on society. So we're using those LM LLM, the uh chat bots if you like. Mhm. To do the triage. So you go into a hospital. Now, if you're going to have a colonoscopy, you will have an interpreter in the pre and post op setting with the doctor, but when you go into the into the hospital in the US, and I don't know where you're based, Armando, but if you go in Texas, so you know this, you go into a hospital, you have a lot of forms to fill out, and they're usually in two languages. If you want, English or Spanish, what happens if you speak Haitian Creole? What happens if you uh if you **[00:18:00]** speak really good English but you don't understand the legal speak? I I know a lot of people who speak English and don't understand the forms in English. So they you you know there is no real option there or no easy option. So we provide that that AI service where 95% to 97%'s good enough. M um you know when you go for your colonoscopy though and they say you know we found a polip we don't think it's dangerous it was small at worst it's probably precancerous but we got to test it anyway and do a biopsy and then you get the results read to you well that you probably need in you know your current language uh and I don't know if you speak uh Spanish Armando or uh do I was guessing from your last names. You probably do, but do you know what the word for gauze is or polip is in Spanish, right? Yes. The the medical tooling you may or may not have heard, especially if you either lived outside the US for a while or interacted with people um before then, you might have never heard those words and you're like, "What is what that what's Gasa, right?" Like, exactly. Yeah, I know. I I live in Miami and I have 10 words of Spanish which I recycle and speak, you know, and use like 25,000 times a day, you know, much gracias amigga tamo, you know, what, you know, I know what I need to say, but again, I I can't really speak Spanish and and you know, when um uh my boat breaks down and uh I need to call STO and someone rocks up to my bow, I'm not complaining that they don't speak great English. I'm ready to hug them and thank them that they're there to save me. Uh, but I have the big words app on my phone and I use that AI. I **[00:20:00]** speak to them and it's great and they tell me, you know, uh, turn off your engines or put your boat in neutral or wrap this rope around here and it's good enough for that. So, it has its place. It's really, really good. But when I'm doing, you know, a legal contract and we're expanding in Japan as we are right now, yeah, I want to see I want to know exactly what I'm signing. Yes, that makes perfect sense because if you've interacted enough with LLMs, you know this to be true that sometimes even when you're speaking to it in the language that it's supposed to be, sometimes the interpretation is different. there's not quite there and uh sometimes especially if we're not from the AI space or the machine learning space or even the software space we might not understand that this is all trained on data like you were saying before and so if there's not enough training material for that specific subject matter it's it's going to have to either hallucinate something or make a best guess assumption kind of thing and that's where you can really get into that's what generative AI is I Don't think people understand it is a guess. It's a really highly educated guess. So, you know, for example, you can get the transcripts of this podcast and it's not going to know every word we say because it's not going to be clear. You're going to hear kind of interference or whatever in the AI and so it's going to guess what you say and it's really great. That's good enough. you know, we're we're not trying to solve war and peace in this podcast. We're trying to get people some information that they can learn from or be inspired from. Um, but you know, if we're on a witness stand, you can't have a witness where the translation of what they're saying is guessed, you know, by its very **[00:22:00]** nature. So, yeah. So we've had I I was actually reading Armando um this morning the in the last 5 years the revenue per head in the top 500 publicly listed companies globally the global 500 um is up 31%. Really that's huge efficiency. Now, if we rewind five years and I told you Armando the amount of revenue someone's going to do per head is going to go up 30, you know, 31%. You would probably be betting on the most ginormous recession uh of all times. Mhm. But yet we've had the lowest levels of unemployment or another way of looking highest levels of employment since. Why? because the macroeconomics changes and you know so how have we become 31% more efficient it's not that people are working harder in fact they're not working any harder they're not doing any more hours this is AI AI's been implemented for a very long time in our business in fact since World War II the average efficiency is around about 1 and a.5% that's an average efficiency gain per year 1 and a half% unnormalized for uh inflation. So we've had this, you know, this is far less an AI revolution and form far more an AI evolution. That makes a lot of sense. The point that you were talking about earlier about the AI tuning itself on itself I think is a very important subject because while every 15 months the content is doubling like you said this is before AI I expect right before before AI creating a bunch of content to **[00:24:00]** itself which probably will make that curve change and you can you can yeah it could be doubling every month in in a few years time But I I like the analogy with with music because if you have I was listening to this this uh interview the other day with someone very very well-versed in studio production and music and all that. They were talking about the AIS that create music like so and and all those tools. And so he was saying that his kids when he whenever he turned on something was AI generate, they were like, "Dad, why are you listening to AI music?" Like, and he was like, "How do you know that I'm listening to AI music?" And so the kids could tell there was something in the in the sound that they could tell. And it's there's there's certain harmonics and there's certain artifacts there and and maybe the kids now just just hear it more directly. And so the other guy was explaining, well, what happens is that all these companies created the training material for these agents based on all like Spotify and YouTube and iTunes and stuff like that. And all that has already been compressed. So it's not the original master files, much less analog recordings, of course. They're just digital recordings, but they're already compressed to be like MP3s or very narrowed down. And and if if whoever's listening doesn't know how how music compression works, you have this large file which has all the frequencies of the music and it's very full, if you will, but then you want to fit it into 3 megabytes or 5 megabytes so that it can stream faster. And so you remove a lot of it. You filter out a lot of information that we might not hear to compress this this file. So if you imagine that you're taking off like 70% of the material to make that song and now all these AIs are trained on just that already filtered down or watered **[00:26:00]** down version then the output has to be a watered down version of and listen it's okay that you know I can get into my car I think in the next year or two and I can say this is what music I want to hear this is what subjects I want the music to talk about. Um, I want to be inspired uh to sell more and the music it will create music for that and my and I'm pretty cheap with my cars. So, you know, I don't have the upgraded speakers. I'm not going to hear all the different sounds that that you or your friend might hear and it's absolutely fine. But it's fine for me. It's not necessarily fine for everyone else. And I think it's a good analogy because businesses will become more niche because of AI is kind of filling the center space. So businesses have to find how are they going to significantly improve the AI products through orchestration tools or through wraparounds in order to find their niche for the people who want, you know, who've got really expensive speakers and are going to hear, you know, more of the notes, more of the sounds. And it's funny because I listen to a lot of the podcasts that I'm on in uh my car and you know and it kind of sounds a little tiny and then sometimes I'll be in the gym and I'll put it in my headphones and my wife bought me some beautiful headphones uh on Amazon like oh my god this sounds so much better. How did I ever listen to that uh you know what in the car? So again, businesses have to think about this and individuals have to as well. So you know as an individual, what is my edge? I can create an AI version of myself. A lot of **[00:28:00]** what CEOs do is make bets. They make decisions. I've got a friend and he says, "I just have to make four good bets, a year. That's it." And he he's a a real estate investor. Well, I I as a CEO of a company of 15,000 people in my workforce uh and government's counting, I have to make more than four big bets a year that that work out. I have to make probably a hundred. But that's not a lot of work, especially with the research tools that I've got and the data tools and the BI tools that exist today. So, can I build an AI version of myself to make bets? No. I can make uh an AI version that could probably create some mediocre results, but then it's not really going to be competitive. It's not going to drive the business forward. Why? Because I have human intelligence. I go out, I speak to the clients, I get a feeling from them. You know, sometimes they'll tell me that they're not buying on price, they're buying on value, and I'll explain why. I know they're buying on price. It's an instinct because what they're saying is all very much true, but I've got other data inputs. You know, I can see that they're trying to communicate with me what they want, but they also it's government and governments are under a huge amounts of pressure to spend less for more. And I know the fact that a government RFP is going to be bid bid out using an RFP system which gets marked on points and 30% of all the available points are available to the cheapest price bidder. The other 70% is quality scores. But I know that there's four or five companies that can write really high quality bits. So the difference maker will be on the cheapest price. So I'm making these **[00:30:00]** decisions based on real human inputs. And then the person who's going to be marking those bits is just human. And as much as they want to be objective, it's subjective. So if I go there and I listen to them and I feel their pain and I listen to them and I'm empathetic, I can give myself an edge in the writing of the RFP. And on the other side, I can influence them. I can say to them, "Have you considered this? Have you considered that?" And I was just in a meeting where someone's bragging that AI is writing all their RFP responses. Now my win rate is over three times this, you know. So again, you got to kind of look at yourself as a business and you've got to look at your business as a business. You're going to say AI is great. It's a revolution or an evolution. Doesn't really matter. It's here. It's changing the world. What are you going to do to compete with the machines? What are you going to do to give access to the machines? And if you can come up with those answers, I think you've got a really great future ahead of me. That's amazing because it's not the it's not only, oh, I'm going to replace everything with AI like a lot of people say for the headlines. Oh, it's taking all the employees and all that. Well, yeah, it's going to take some jobs, of course, because but it's shifting. It's it's moving what you did before to maybe someone now does something else because now the AI takes part of that job but like you're saying the highlevel jobs the decision making roles that is still better made by a human which granted with the proper AI can get more data at the right time in front of their eyes and help make the decision easier but you still have that intuition and that that personal link to people that **[00:32:00]** can influence. Yeah. I mean, think about your plumber. It's 3:00 in the morning. You hear gushing water. You run down into your basement and there's water all over the floor. The lights are out because they're fused. You know, do you go on AI and start figuring out, you know, how to fix your pipe or do you call the plumber? No. You're going to call the plumber. for a start, even if you figure out how to do it, you don't have all the tools necessarily in the parts. So plumbers are safe. But a plumber still is using AI. They can employ AI to do their sales and marketing, to take their calls and their messages, to organize their calendar, uh to manage their their parts and their spare parts in their van. So you know this is the harmonic harmonious environment that we uh see and it happens in all businesses across the spectrum. I just recalled something that happened recently. My dad has been a business owner all most of his life and and he runs a car shop and a good friend of him told him, "Hey, my my car it's an expensive SUV. um has an issue. There was this noise. I took it to the dealership and it's going to be $15,000 to fix because they have to change the suspension. I don't know. It was going to be very expensive. And and then my dad said, "Okay, that sounds off." Because of his intuition, right? So, he goes to the dealership, looks, races the car, looks at it, and then he makes an educated guess based on years of experience. He says, "No, I let's take out these there's these little rods. There's a couple of rods, little thing." Takes them both off, both sides of the wheel, and notices that one is like a millimeter or two crooked, just a little bit. And he said, "Put that one **[00:34:00]** in the other side. Let's just check." And the issue appeared to go away. So then he calls me. He says, "Oh, we can't get these here because it was in Mexico. Order them, please." So I order them. They're 15 bucks. the the thing 50. So he has get two of them just in case, right? So I get two, send them over, they put them in and it got fixed perfectly. No need to change the suspension. No need to change and nothing was broken. It was just that the thing got a little bit twisted under tension and and that's not something that the the AI would have likely deduced just by looking at it. What what you're describing is knowledge versus understanding. AI has the knowledge. You could go in and you can say this is what's happening in my car. You can upload some pictures and it will look at the theory behind that and it will search all the chat rooms and everything else and it will apply knowledge. It doesn't have then the ability to go from, you know, that knowledge to true understanding. Your dad could go in and say, "Lift it up. Let's try this. Let's move it around." And, you know, AI can't do that. So, he can't guess that for you. So, we have a a real opportunity and a real challenge all at the same time. We have to get really good at applying understanding to what's now commoditized knowledge. And the whole education system is going to have to change because my kids I've got two girls and they one of them was studying the other day and she had to learn what DNA stood for and she had to learn how to spell it. And I said, "What a waste of time. I can't believe I'm paying $35,000 a year for this this education. And she looked at me and she's like, "You can't say **[00:36:00]** that, Dad. This is important. I want to become a doctor." So I said, "But you don't need this. You need to understand what the importance of DNA is. How can it use? What is the um what is the research on it at the moment? And how can we edit it and change it?" And none of her questions in her tests were on that. They were all on what is the what is DNA from a very technical perspective. So it's the same with law. I am smarter in legal knowledge than my attorney because my attorney refuses to use AI or at least he tells me that. uh you know uh but I I can go and find any law. Uh funnily enough, and this is a true story, my father spent over $100,000 on tax advice and then went on AI to test it. Found out that it was wrong. No. And then he sent a bill for $100,000 back to his tax advisor for correcting them. And they paid him because they were so scared of being sued. and and can you imagine the media interest uh around that? So knowledge was there. what we now have to focus on is how do you apply that knowledge and the education systems changing but I'm really happy because even though I if you're just listening to the audio of this uh you know me and Armando may look great but we're not that young and I think for the last 5 to 10 years maybe even longer younger people have had the advantage in the workplace it's shifting just in time for us Armando although you're, I'm sure, a lot younger than I am. Um, the older you are, the more experience you're likely to have. The more experience you're likely to have, the better you're going to be able to apply this understanding. **[00:38:00]** So, I think that you're going to start to see a return in the workplace. They're really valuing experience and understanding because just knowledge is is near free at this point. That's a fantastic point. I'm glad you brought it up. I've been tossing around that idea of the shift in power that that the older generation that typically you would see, oh, now they start to taper off and they go and retire and go off to Florida and just go away. No, they've got I've always said they have so much knowledge, so much experience, so much they can bring to the table that there's no way that in a valid amount of time, you're going to be able to have that arbitrage that that just the sheer amount of like like you said, it compounds. It's not just the knowledge. It's the experience. It's the lessons learned. It's it's the ability to in a split second make a choice based on a million different factors. And if we can in a way bring which I expect like from what you're saying it's also is definitely happening at some point bring those people back into not necessarily the work force. They don't have to go and work and hit things with a hammer. It's more of a knowledge transfer or experience transfer where we're having those people around is going to give huge upside. Yeah, I think it is an experience transfer. I give you a good example I like to give. I am addicted to YouTube channels on flying planes. I don't know why. I'm not a pilot myself. Um, I just enjoyed the process, the workflows, the complexity of it to the point where I think that I can fly a plane pretty well. **[00:40:00]** I've never actually flown a plane. Now, imagine that I study all the technical stuff. I pass all my exams. I, you know, spend the next two or three years watching pilots in the Dreamliners fly. Next time you get a Dreamliner, would you like me to be the pilot? No. Not necessarily. Exactly. You wait for a choice. Exactly. You know why? Because I've never done it before. I might have all the knowledge in the world. In the same way you wouldn't want me to be your attorney. I can go I'm really good at writing legal prompts and scanning documents in a way that the average Joe doesn't. You know, they'll say, you know, they'll upload something into chat UBT saying, is there anything on here I should be concerned about? Yeah. I will write really good prompts such as if you were going to try and break this contract, tell me where you would look, where is the weak parts, and then I'll put a new subprompt in there to dig down deeper. And then I'll say, please now rewrite it to overcome that. So, I'm really good as a chat GBT lawyer. you know, but if you were getting divorced uh or sued for whatever reason, would you want me to be your attorney? No, because I've never been in a courtroom. I've never had to do a negotiation with another attorney. And the truth is that Open AI can pass the bar much better than the average trained attorney. You know, that's a fact. It was PR that they did and AI can beat you know even the best chess champion in the world. It can understand the logic behind all the moves in a different way. We have to apply AI in our businesses in the same way and we have to apply the hi where the AI fails. So in my business **[00:42:00]** when my automation can't find an interpreter to deliver in real time uh it will go through to the call center and the job of the call center is to call up all the it works like Uber right so if there's not enough taxi drivers on the Uber app then you're not going to get a taxi no matter how good the technology is same with me if there's not enough say French Creole uh interpreters logged in ready to take a video call then you the wait time will get longer and eventually you'll give up. So when that happens, it goes through to my call center and I have hundreds of people in my call center whose job it is to go and then call up these these individuals and say will you log in? Will you start to take jobs? And sometimes the individuals be like uh I want base rate plus you know uh or or I want to apply a travel rate to it that because this is a facetoface booking and trains are really expensive or gas just got really expensive. So that's what they do. But my call center can only make one call at a time. So or one individual in the call center can make one call at a time. One AI Amy calls up who's our AI agent. Uh she can actually make 30 calls at a time per server and we can have as many servers as we want. So the chance that AI is going to get someone an AI is fully integrated in the back end. She knows what we can afford to pay and what gross margin targets are. She knows exact distance from you to a job. And I mean, you can even ask a to sing a song about the job and she'll make one up and you know and sing the song. And we do that more as a gimmick to show our clients. But it exists. What AI Amy can't do is beg. She can't do it. So when I go in a call **[00:44:00]** center sometimes I you know I I listen into a call and I you know we do this thing called barging where you can plug into a call and this and they're just outside my door here. I'm right in I'm in a booth right now but they're out they're just outside my door and I'll do these call bargings and I'll hear uh Virginia say look we really need to service this client at 98 plus percent. We've got a quarterly business review coming up. If you take this job, I guarantee you next time I can give you an uplift, I'll give you a call. I'll intercept it from the automation and give you a call and give you the uplift. Please take it. And nine times out of 10, the person's like, "All right, go on then." That's what the humans can do. And if it wasn't for the Virginiaas of his world, we wouldn't have as high service rates and uh and we wouldn't have as happy clients. So you need to be able to use AI in your call centers or in anywhere in your business, but you also need to know where the humans are required. So we figured out that the minimum number of humans that we would need to run our business without the AI is roughly a 400% increase. So um and and that's why The Big Word is so prolific in government contracting because we've been bringing in technology and investing in it for so long. In fact, we have only once in the last 20 years taken any money out of the business. 100% of the profits have been reinvested into building technology and that was because we understood the power of it. So, you know, I I I try and give takeaways for for podcast guests, whether I'm the guest on the podcast or uh or or or the host of my own one. And that is this is not the time to be uh **[00:46:00]** greedy. This is a time where you must leave that money in your business. Your competitors are worried. They're reading all the same stuff as I'm reading about the economy. They're getting rid of their marketing teams. They're get getting rid of the sales teams. They're cutting down their capex R&D uh work. This is the time where the best of the best double down. When your competition are selling less, increase the size of your sales team. Invest in the products because you won't get an easier time than now. And don't believe this is the the last one because I'm going to have to uh go in a minute and do uh some work for The Big Word. But would you believe if you were going to buy gold today um and you went to a gold dealer, you're not going to take their word for it on the price of gold in the next 6 to 12 months. Why? Because you know they're not ever going to be negative on gold. Even a negative is going to be sold as a positive. Mhm. And I think all humans understand that, you know, I'm not going to take advice from someone who's got a huge financial benefit in selling me the product they're advising me on. Except when it comes to AI because we have now made these people like gods. So Elon Musk, you know, he'll come and say everyone needs to think about the future because we're not going to have any jobs and there's going to be universal income. Well, guess what happens? People then invest money in his stock because they think he knows something that no one else knows. And his stock trades at a massive uh uh premium on all the other companies doing the exact same thing. And he's figured that out. So, he keeps saying those things whether he believes them or not. **[00:48:00]** And then the whole world is freaking out over this thing. Open AAI, same thing. In fact, uh, he was fired from the board, the CEO and they had to reinstate him because he was so good at speaking up the valuations of these businesses. Now, if they were right, why isn't AI profitable? And I'll give you the reason why. Because people don't want to pay for it. It's not actually creating as much value as the perception. So, open AI not profitable. um the co-pilot part of uh Microsoft running at 75% loss. It's insane the amount the billions of dollars and you know a lot of people say without knowing Josh but it's going to become profitable. It's a scale game. Actually not true in the world of manufacturing things get more expensive over time not cheaper. Think about cars in 1970 versus in 2020. Um, there are some exceptions, LCD television screens, but software nearly always gets vastly more expensive because and just look at AI. The chips are mass manufactured now. We have incredible technology to print these uh uh chips that are required. They're up 50% yearonear. The electricity to power this AI is up massively. So, you know, AI is going to get more and more expensive and yet it can only be sold at a loss. Open AI, I was reading, I don't know if it's true, have signed $1.5 trillion, sorry, $150 billion worth of supply chain deals with the likes of Nvidia and so on. Why do you think their biggest investors are other AI companies and doing these circular deals because they got to prop up their prices? So, I'm here to tell you AI is **[00:50:00]** happening. It's been happening since the industrial revolution. It's an evolution and not a revolution. And you've got to be part of it and you've got to be investing in it. You'd be like my grandfather who lost his job because he was absolutely sure that the machine couldn't make the suit as good as him and for as good price as him. But on the other hand, don't buy into the sensationalism. Add value. Take your time. Really consider what you need to do for your business. And I guarantee you if you do that, you can be here in 40, 50, 60 years time. As long as you're still alive, but you and even if you're not, your business will still be around. Josh, this was amazing. Thank you for all the very, very helpful advice. I'm sure our listeners will really enjoy this show. If people want to learn more about The Big Word, where can they find more information? So, I have uh nowhere near as good as yours on Monday. I have my own YouTube podcast. It's called Exec. So, if you look at Exec Joshua Gould, please uh like, subscribe. It makes my day. And make sure you do the same for Aando as well. We Thanks. We're not charging for this, I don't think. So, you know, that's the best way you can pay us back. Otherwise, you can hit me up on LinkedIn. You know what to do. Put Josh Gould, The Big Word, in the search bar, and I'll love to hear from you. All right. Thanks everyone. And if you haven't liked and subscribed, like just said, please do on both accounts, and we'll really appreciate it. Thanks so much, and we'll see you in the next one.