ADmire!
ADmire!
Media Entrepreneur John Pasmore of Latimer.ai
In this episode, host Larry D. Woodard interviews John Pasmore, founder and ceo of Latimer.ai an inclusive large language model designed to address bias in AI by being trained on a diverse dataset that includes experiences, cultures, and histories of Black and Brown communities.
Thanks for listening. Don't forget to subscribe.
It was not long ago that AI was just a popular buzzword. I think we can safely say that time is past. AI is in use everywhere. Foodly in education, finance, national security, criminal justice, transportation. Everywhere. It's not an overstatement that the use of AI to integrate and analyze information and to put big data to practical use of transforming the world we live in. With that, however, come risks and dangers. First and foremost, personal privacy. As the number of data points on each individual grow and become more available, lots of human oversight and control exposed to the threat of malicious behavior. And a major danger and reason for my guest today, the risk of bias, built-in discrimination, and unfairness in the data sets and algorithms. My guest today is one of the most engaging, focused minds I know. He is one of those guys where if you talk to people who knew him when he was 12 or 22 or 32, they will all tell you that he was operating from an internal topic, that he was focused on getting to a particular destination that the rest of us could not yet see. His background is in media, early stage investing, capital portfolio management, and tech, all setting him up for what he is doing now. My guest today is John Pasmore, founder and CEO of Latimer AI, a large language model, trained with diverse histories and an inclusive voice. John, welcome to the podcast.
SPEAKER_00:Thank you so much, Larry.
SPEAKER_01:So when I have a great guest, I'm almost always guilty of rushing forward to get to the good part. So experience over the last four seasons has taught me that the favorite part for my listeners is hearing about the path, the series of life events that conspired to get you to this point. So if you don't mind, we'll start at the beginning. Tell us your origin story, where you're from, what it was like growing up, and what things happened early in your life that started you on the path to where you are now.
SPEAKER_00:All right. Well, I won't take all the time by going all the way back, or at least not all the way back in in detail. I will say this that I've always, you know, I I come from a um immigrant, you know, my mom's from Trudad, my dad's from Jamaica, you know, she was a nurse, very blue-collar. He owned a gas station growing up. Um, and you know, so so everybody was working, you know, so at times people were working two jobs. It was not unusual. And, you know, it just kind of I guess gets ingrained in you as a child or me as a child that hey, you know, you're gonna be working. Um, and so you know, I started work probably a little too early. Like, yeah, I I think I worked full-time in high school, which I definitely would not recommend to to young people. Focus on your your studies, that's a better use of your time. Um but I I guess I was I was anxious to get to what I consider uh the real world. Um started really in banking and finance. I was a stockbroker early early on. Um was fascinated by by all of that and you know um how how money was made and and who was making it and all that kind of stuff. Um but I was also really um interested in in the culture and how the culture was communicated. Uh I worked really early on before Chase was Chase, it was it was something called Chemical Bank, and then they bought a a bank uh down in Texas called Texas Commerce Bank. And uh and I moved down to Houston, Texas from New York. I grew up here in New York, um and uh didn't love banking, started my first media company there that started really as kind of an entertainment guide or a nightlife guide that we grew into a full-fledged city magazine over some years. Uh great um, great experience, great experience about fundraising, great experience about creating content, about working with creative people, you know, so many, so many lessons on so many uh different levels. Ended up selling that business and then wrote a business plan. And for whatever reason, I thought that Russell Simmons, the founder of Jeff Jam, certainly needed a media company in his uh suite of efforts, and came to New York and hired his brother really as an art editor for a new media company called One World. And then once Danny joined, Danny Simmons joined, uh, we had we had access to Russell and over some period of time and convinced him to become a partner in that business where you know we went on to create uh uh a TV show with Warner Brothers, Russell Simmons One World Music Beat. We created a custom published version of what we were doing with her. So a very you know successful young media company during a period, let's call it the late 90s, early 2000s, where there was a lot going on in New York City, you know, Vibe had launched, the Source had launched, and there was many, many other young um media businesses, but there were also young, you know, entrepreneurs that were, you know, whether you're on the West Coast with Dr. Dre or you're on the East Coast with guys like uh Russell Simmons and many, many others creating these, you know, these kind of pillars of the culture.
SPEAKER_01:That's um so so you're in media and and you're aware of, because of your banking experience, sort of finance. Um so what was the um the move that you made well appears to be when you look at your um your C V from media and uh even international media uh to uh sort of uh being more interested in not so much interested, that's the wrong word, but understanding the pivotal role that financing um and and sort of mentorship and ownership uh played in the whole um thing.
SPEAKER_00:Well, I think it goes back maybe to the fact you know that I'd started in banking, you know, I'd I'd I'd gotten a stock, you know, my stockbroker's license. I kind of knew how people raise capital. And then, you know, working with with Russell, I was and you know, seeing, let's say, at that time like Spike Lee is out trying to raise money for a film and he's asking Magic Johnson. I was like, well, the purpose of all of these funds and institutional capital does not seem to be working uh for even very successful or what I considered very successful black and brown entrepreneurs like Russell or Spike, it would seem to me that anybody would want to fund them, but that that wasn't the case. So it was my first kind of real realization that uh institutional capital, which is where you know you're kind of supposed to go. Hey, I have a great idea, I'm an entrepreneur, I can build this thing, uh, wasn't always available to the talent uh in our community. And uh, you know, some of that was because we didn't always bring the skills to, you know, have super detailed plans and financial projections and all the things that maybe institutional capital wanted to see. But also, you know, a lot of it is just that you know, there's a bias. There's a bias then, the bias still exists. Um, so you know, I I had raised money from a magazine from a media company down in in Texas, I'd raised money for a media company here in New York, and I just wanted to use that that experience and share it as much as possible with young people. And you know, I did some mentoring with tech stars and uh constantly, you know, still today mentor individuals where I can. Um because you know, these these are things that can be solved. It's still very difficult, but you know, you can put yourself in a better position to fundraise.
SPEAKER_01:So um at a certain point when you're working with these people, when you're partnering, when you're running companies for people and you're getting partnerships, um, did your thinking become oh, I want to be the guy at the helm uh and that I want the company that I work in to be a tech company?
SPEAKER_00:Yeah, I think you know, we saw that media really wasn't um transitioning, uh, you know, whether it was the source, vibe, us, you know, it was it was only really the digital, you know, the the companies that were really born as digital entities that were navigating this this kind of new tech frontier that um, you know, hey, you just plug in your website and you can reach millions of people as opposed to chopping down trees and making a magazine and physically trying to get those magazines distributed across the country. Um you just had to, you know, have the technical savvy to build something that's attractive and engaging, and you could have kind of the same financial result, if not much greater, because you'd have greater reach. We also saw that on the the revenue side that as an advertiser the the benefits of digital where you had you knew exactly the demographics. Uh you could either do targeting and we're only targeting this demographic, or you kind of knew exactly how many people saw your ads versus uh traditional media, which was you know, even Nielsen on TV, it's not it's not accurate. It's it's kind of a guess uh uh of a share of all the people maybe watching television at that time. So it's it's highly different than digital, where we could say, oh, you know, we we served your ad 1,122 times. You know, you knew exactly how many times the server was hit. Um and I I like that idea. Um, you know, we we'd transitioned one world. I went to to work with Trace TV out of Paris, um, saw another uh really compelling uh travel business called Voice TV, and that was really my first foray into digital where you know we kind of raised$10 million from a BC and a and a strategic partner, Mark Travel out of Wisconsin, and served as president uh and then became the CEO of that company. Certainly understood the benefits of, you know, as a president, you do have a CEO that you're kind of speaking to. As a CEO, you know, you always have somebody that's a board that you're speaking to, but as a CEO, you have a lot of uh leeway as long as you're driving, you know, you have your your metrics that you're trying to run the business to. And as long as your board is happy that you're driving toward towards those uh those solutions, then you do have a lot of freedom to to run the business as you see fit. Um that's you know, honestly, for me, it's it's just kind of more fun to have that freedom and run the business. So um, you know, stayed in the digital realm uh with several businesses, and most recently Latimer, where you know, I I can build the business and find partners who believe that we can we can build a business that we all envision, to ask a Meta or a Google, hey, can you correct our history because what we're seeing here is wrong. And you know, as a as a father of a of a teenager, the last thing that you want to leave as a legacy is hey, you have these four or five you know super powerful um AIs, and maybe or maybe not, they understand you or or people that look like you and your history, but they're not really focused on that. And you know, we saw, I still give this demonstration where I asked Chat GPT, who are the most important artists in our culture? And you know, without if you ask it that cold, it's gonna say Leonardo da Vinci, Michelangelo, uh Pablo Picasso, it's gonna give you basically all these European artists because in its mind it thinks when you say our culture, you're thinking Western European culture. And that's great, maybe for certain things, but certainly that's leaving out a very wide swath of the world, and it's leaving leaving out the contributions that we've made. Contributions like Lewis Latimer, who worked with Alexander Gravel, he worked with Thomas Edison, was a genius in his own right, but has largely been left out of the history that we've written about these, you know, Alexander Grab Bell is kind of this big, you know, tech hero as is Thomas Edison, but we're not sure because Lewis Latimer was a black man, he, you know, in many cases wrote the patents, he was doing the science, but we're you know, it's very unclear. You know, we gave 100% of the credit to Thomas Edison, but we're not necessarily, that's not necessarily where 100% of the credit was due. So we chose Louis Latimer as a as a as a name because he represents somebody who was a genius in his own right. Um we wanna we want to emphasize to young people that, yeah, you can go, you know, you can pursue a career in sports or entertainment, but you can also simply stay in school, use your head, and um, you know, get the same kind of acclaim and you know, wealth uh if that's your goal by building companies, using your head. And so that's kind of, you know, that's that's why we chose the name. That's why we do a lot of on-campus events to really kind of demonstrate that to young people that there is this path, that there is this new technology. Look at it not only as a tool to do your homework, but as a as an opportunity to create new businesses on top of it. Right.
SPEAKER_01:And and so I love the idea of uh of Latimer, just because of what you're you're saying, uh that's why it's important. Uh you've told us what it is, you've told us why the name. So uh I'm gonna ask you to give us uh another perspective. If you're in the middle of your journey and media was a solution uh you know and allowed you to move towards self-actualization 20 years ago, how do you view the future now? And if building a popular large language model is part of it, what else is on your mind that might be necessary for the future that you see for your son?
SPEAKER_00:Well, I think, you know, I think we see a future where automation um is coming to the forefront, right? So you you see, you know, uh as much as I don't agree with a lot of what um we're seeing out of uh the Tesla CEO, we are seeing uh things like robotics, right? Um and we know, you know, it's it's actually still fairly early days in the in the in the ability for you know a robot to be autonomous. Um and we know that it takes an awful lot of computer processing, you know, even for you know uh some something that's robotic to multitask. So, you know, robots are generally very specific. Oh, this robot can work in a warehouse, this robot can drive down the road, this robot can maybe slice up an onion, but you would never really have one robot that could do all three of those things because they're such different um types of tasks. But um we can see that that's somewhere on the on the horizon. Um I think that one of the things that we'd like to be, um, you know, we have a lot of blocking and tackling to do before we ever get there, but you know, robots also need to be able to speak to people um and interact. You know, we we're kind of seeing the the realization of so much that's been sci-fi. And I think you know, one of one of our goals again is well, when when you know some autonomous machine interacts with a person and has, you know, is is using some sort of visual clue like who am I speaking to? We already see issues in you know how um law enforcement uses images and and often uses images incorrectly or draws incorrect conclusions based on an image. Okay, this is a person of color and you know perceives a different threat level or sees or can't discern the difference between the person that it's looking at and a and a picture that it has, and therefore says, oh, this person that I'm looking at should probably be arrested. But and that turns out to be an error because maybe the engineers that created you know the the optical capabilities really didn't test it with black and brown people. And and those those types of issues, you know, we see even that issue in the blood oxygen um meters that we've used or that were used during during COVID wisely is with a little bit of melanin in your skin, was it was wildly inaccurate. So apparently they didn't really do such a great job testing against all different types of skin skin tones. And obviously that can have real-world impacts if uh, you know, if if some doctor is taking the cue of what it thinks is an objective uh piece of equipment. And, you know, so going forward, we don't want to be in those scenarios and we want to be the solution to as many of them as possible because again, I think that they're they're all commercial opportunities. I don't think that they're always that there's like these things are done out of malice. They're kind of done maybe more out of neglect. But again, so for us, there's a commercial opportunity because if you can correct that and you know go to market and say, hey, well, we have a correction to make whatever you're doing better, um, we think that that um you know our potential partners and all of that will will respond to that. And that's kind of what we found with Latimer. That all, you know, a lot of educational institutions are like, oh, this is better. That's great. And you know, a lot of educational institutions also have additional data that's maybe not on the internet, hasn't been scanned, so it's not part of what any large language model knows. And we want that data as well. We definitely want to make sure that we have as much um accurate data as possible in some of our early relationships, you know, like talking about Morehouse going live with their trial. Um, you know, these these institutions have data buried in their libraries, stories or manuscripts or writing that was bequeathed to them by you know authors that are important, uh, but maybe they're not scanned, or maybe they're only scanned for the students. So no LLM has access. So we want to also be the bridge for those.
SPEAKER_01:So that all sounds to me objectively practical. Uh my next question is potentially loaded, uh, you're forewarned. Um I don't necessarily mean for it to be. It's just the the times we live in. Uh but I think that your perspective might be a good one for our listeners. Uh mostly business leaders listen to my podcast. Uh we have an administration in office that's working hard to dismantle many of the things put into place to begin to cure the deep legacy of bias in our country. There is no discussion about why it was needed in the first place, just a single-minded focus of removing anything that's not to consider the role bias has played in keeping our country from being inclusive and diverse. What role does Latimer AI play, and how do you operate in an environment where you may be swimming against the current?
SPEAKER_00:Yeah, I mean, we're still figuring out what that what that means. Latimer's um, you know, we've always said, and it says on the on our on our homepage, AI for everyone, um, with the idea that everyone should want the most accurate artificial intelligence possible. So that's kind of what we believe. And it's not so much that um, you know, we're we're doing one thing that's going to denigrate somebody else. Um, you know, we're just trying to make the the record and the data as accurate and as complete as possible, noting that it's not complete now. And again, maybe not out of malice. You know, we've certainly found that when you're out looking, you know, certainly, okay, so you've ridden, you've you've read the entire internet, but let's say the Schaumburg Library isn't doesn't have their you know materials on the internet. So now when you leave out, you know, these massive cultural institutions, sure the AI knows a lot, but you're leaving out an awful lot of context that if it had it, it should even be better. You know, there's been studies that with diverse data that the responses from AI are more accurate because now it doesn't have to hallucinate because it actually has the data to to make a proper response. So we're hoping that we, you know, number one of them actually hoping that I never really have to kind of confront and and explain why, you know, AI for everyone or or why anybody would want the most accurate AI. Uh we haven't had to yet, uh directly, but I'm sure indirectly, you know, what you're talking about probably affects, you know, let's say a VC who's looking at, you know, hey, I can invest in all these things. Um, you know, here's here's you know uh a startup that's focused on legal or accounting, and here's one that's maybe focused on rates, which is kind of a hot topic at this moment. So you can see how a VC might feel like, okay, this one, this Latimer venture has this extra layer of risk. And um, you know, we're trying to do our best with getting uh getting customers and revenue to show that you know that risk can be mitigated. But you know, I can see how people might say, you know, there's there's definitely the perception of uh extra risk there.
SPEAKER_01:Yeah, and and and it's actually really interesting. I mean, I grew up um, you know, living in Europe for uh for a good part of uh my childhood. And if you go to school in Italy, you know, and um they're very proud to tell you that Marconi invented the radio and and all these things. But if you go to Germany or you go even to Russia, it's it's a different person. Right? And it's not even that they're claiming that the person which they are uh invented the thing. They're just emphasizing what role they had in it. You know, so if you go to Ecolay or you go to Germany, you know, um uh everything the history is slightly different from the perspective of the person who's telling it. You know, and uh so you know, so naturally that would be the case here. But uh I like the way you talk about you know, you building this business tool and um and the whole idea of of AI for everyone. So in your endeavor, how can the business community help? You know, what are your obstacles? What are your opportunities?
SPEAKER_00:Um the business community can help by subscribing and and then using is the best way. I mean, we work with a big healthcare company out of uh DC called ThinkHare, over a billion dollars in revenue last year. And you know, they were they're using uh Latimer to write call center scripts. So we're pulling data on customers before they're called and creating a customized script that um using Latimer versus, let's say, ChatGPT is maybe more empathetic to uh the diverse people that they're calling. It just talks to people different. You can look at uh you can look at these AI platforms as almost synthetic beings at this point. They all have a perspective, they all have a way that they speak, um, and that will become more pronounced as we go. So any support, you know, any trials uh with with businesses is is certainly certainly welcome. We aren't generally focused on higher ed at the moment. Um, you know, it's you know, we don't have uh billions of dollars to pursue every single uh vertical, so we've focused down uh on education uh where there's a lot of um there's a lot of overlap too because as as we as we build what we want for education, we also see that cities and others are going to have this big retraining effort as uh they have traditional workforces, they would they would like to keep them employed, but they may need to be retrained to use AI in ways that they you know didn't come out of school learning. So there's a huge, huge portion of what we do with higher ed that could be applied in different scenarios from education to retraining, which we think is a big, big opportunity going forward.
SPEAKER_01:So uh how can listeners learn more about Latimer AI?
SPEAKER_00:Um go to Latimer.ai. Uh certainly, you know, I'm always open and connecting with people on on LinkedIn. Uh, you know, it's just John Passmore's, I think it's Jay Passmore on LinkedIn, or just Google my name and I'm sure you'll you'll find it. Um happy to connect and talk more. I appreciate all these conversations, always end up uh being being valuable and surfacing people that we wanted to have conversations with. And uh each conversation I think helps move the business forward.
SPEAKER_01:Cool. Any final comments or observation for our listeners?
SPEAKER_00:Um only that you know, regardless, we're we're seeing moms embrace AI as a way to maybe level the playing field with uh tutoring. Um it's it's a wonderful tool and it's very, very easy to use because you just have to talk to it. So whether it's latimer.ai or whether it's chat GBT, I just encourage everybody to download it or go to the websites and use it and just experiment with it. It's very simple. We all have a free tier where it doesn't cost anything, so you just can kind of begin begin a conversation about any topic. We've all read the entire internet. So whether it's physics, whether it's a recipe, whether it's philosophy, um, you know, you can have at it and experiment with AI.
SPEAKER_01:Very cool. Thanks, Jeff. It was a pleasure having you on the show.
SPEAKER_00:Thanks, Larry. Appreciate it, man.
SPEAKER_01:This has been in my host, Larry Witter. Encourage you to subscribe to Apple. Remember, you can get it in my Apple Spotify and anywhere you get your podcast.