Hour 4: Artificial intelligence can benefit everyone

Thank you, Chris. Great job. As always, you're joined now by Andrew Schwartz Professor in the E.J. Uso College of Business at LSU. He's been... He's been researching technology-related issues for over 20 years, and he recently has been focusing his attention on AI. Morning. Andrew, how are you? Good, me today. Good. Real good. Let's talk about this piece that Mark Zuckerberg wrote for the Wall Street Journal. I presume you're familiar with it? Very much so. Yeah. Tell me what some of the points that he made, what you agree with, which you disagree. For people that have not read it, I guess we should kind of recap it. Go ahead. Absolutely. So what Mark Zuckerberg's perspective is that a lot of the models that we're using for artificial, the information should be publicly available so that we can understand and figure out how these models are working and have their transparency with artificial intelligence. Do you think the public could understand it? I think we need some education around artificial intelligence and how it fundamentally works so that we can understand it better overall, and so that we can make better, wiser decisions, you know, consumers of these information, these platforms. You've been on with us several times, but every time we talk about it, new people listen. Maybe they didn't hear the time that you were on. So let's take a step back and talk about while everybody has heard of artificial intelligence, they might not know what it actually is. So let's pick it up there. Sure. Let's start conceptually. So artificial intelligence is simply a technological system that attempts to mirror human intelligence within a technology platform. And most of us use this as if you on Facebook and use meta's AI, like when you search for someone on Facebook or Google Gemini, when you do a Google search, or if you have chat CBT, these are all different types of artificial intelligence. So... Technically, what's happening with the computer to create this artificial intelligence, if you can kind of simplify it and dumb it down for us? Absolutely. So what happens is, let's say I take a document, and I scan that document in, and then I expose it to artificial intelligence. AI is what AS doing is taking a look at that document, and it's breaking it down into a series of words. Then once it understands the words that are in the document, it tries to understand the relationship between the words. So let's say, for example, let's say that you have a document that talks about meteorology, and it says the sky is blue. Well, the AI's going to take that in and then learn, basically, that the sky is blue. Based on what? Based on references that it has before? been fed about what the color blue is, about what the sky is, or what? I'm trying to figure out how that connection gets made. It's based purely upon what you teach the system. So basically whatever documents that you use to quote-unquote train the system is what it's going to learn. So if you were to feed an AI system document that says the sky is purple, it's going to learn the sky is purple. So it's learning based upon what we're training it. Okay, so the question is who's training it, right?

Exactly right. And depending on how you train it and what documents you use to train it, you'll get different results. So what Mark Zuckerberg's position is, we need to understand documents you're using in order to make those, in order to train your systems. And then once you train it, what's happening on the inside is a series of mathematical calculations. And his perspective also is that we need transparency into how those calculations are taking place. What would that look like? That's a big question. That's an open question. And the other open question beyond that is, so all of these platforms are basically, it's proprietary. That's how they're able to produce the AI. So the moment you expose that and make it transparent, you're also exposing all those trade secrets and all the information that they use to create their systems themselves. So at that point, from a competitive perspective, if you're forced to expose your trade secrets, which you continue to do what you're doing, of course not. So there's some downsides to all that exposure of transparency. So how involved is Mark Zuckerberg with AI? Okay, so here in lies the question where you have to begin to become skeptical of the source. So we've just talked here about Google and Meta and ChatGBT. The truth is that meta is not a leader right now in AI, and I think we can all agree to that. So for him, if it becomes more transparent from a purely competitive and business perspective, It enables him to take the playing field and actually make it more equal. Wait, you're going too fast. Go back to Meta and Mark Zuckerberg, the linkage there for people that don't know. Sure. So Mark Zuckerberg is the CEO of META, and META creates Facebook and Instagram and WhatsApp, as well as their own AI. And their AI is called Lama. So if you're in Facebook and it says search meta AI, behind the scene, their model is called Lama. Well, Lama is not being used by nearly as many people as Chachibit or Gemini. So for him, if the more transparency, the more that he has a chance to potentially get more users and more adopters and... a better bottom line for him as well. So it goes to consider the source. Exactly right. Now, what he's saying may or may not be accurate, but we have to distill that to understand that better that where this is coming from. Now, you and I have talked over the years that I do believe that I do transparency and regulation is necessary because I think we do need to have some degree of transparency. Where that line lies, however, is a very open question. Because you're talking about fundamentally you could change history here, right? Because if the wrong thing gets fed into AI...

Well, then future generations won't know, and they'll believe that what the AI tells them is true whether it is or not. Accurate? You're 100% accurate, and there's a broader concern as we move forward. We're hitting a point of saturation where we've begun to almost at a point where we can't feed any more information into AI because we've consumed so much of it. So we're going to hit a point that scientists are projecting. Go back. What does that mean, Andrew? You can't feed any more in it because we've consumed so much of it. because we've trained it with all available information at this point. We're getting to a point where it's approaching this. So meaning we've brought every post from Reddit or Facebook, we've brought books in, we brought all this information in. I think about when you interact with ChatGBT, and you say, tell me about this book, it knows that book, which means it was trained on that book. So if we start talking about the vastness of human knowledge at some point, well conceptually will hit a point where with it becomes almost uh... impossible to continue to feed it more human information well at that point engineers are going to potentially start to think about what if i take the output from a i and feed it back into a i what if i started to train it on itself what were there resulting The scary part is none of us know. We really don't know what happens because we never hit that approach to that point. And so the concern once we hit transparency of all those weights is you and I have talked about foreign bad actors. What if a foreign bad actor takes the information about all those different calculations and all those documents and then uses that to train that a new model using AI-produced information? We simply don't know what the ramifications of that are going to be. And no one knows that. But it is not beyond belief to say that a hundred years from now, AI could create a history that never existed and mankind would be basing its decisions and everything else on made up history by AI, right? It is entirely conceivable. And I would challenge all of your listeners out there, go out to your favorite platform and ask it about, ask your favorite AI platform to tell you about the time that Einstein went to the moon. We know Einstein never went to the moon, but AI may not know that. And the other thing I'm thinking of is, as you talked about books, tell me about this book. Well, that depends on who put the information in about the book or do they put the text of the book in because when you go back to English literature, maybe a poem or maybe even a novel, what is the author getting at here, right? They ask you that. What is the meaning of the poem? What is it? And my answers are always different than everybody else in the class. So it goes back to that kind of scenario almost, right, as far as what does it mean to whomever it is that put that in there? Absolutely. And think about the source itself. Think about it if you're an author. Would you want as an author for your information to be in chat, UBT, without your authorization? So what about intellectual property? Let's say that, let's pretend Tommy that someone to take every single episode of every single time you've ever been on the radio and put it on AI.

wouldn't you want some type of compensation for that? Yeah, you would think, yeah. And we're not at a point, and we're not at that point right now, to be honest. Let me take a break. We'll pick it up and come back. Y'all, look, there are no dumb questions when it comes to this. We're talking about AI with our friend Andrew Schwartz, who is a professor of IT and the E.J. Uso College of Business at LSU. Been research and technology-related issues for over 20 years and has recently been focusing his attention on AI, which can be... Very scary. Maybe not. We'll talk about some ways where it might not be so scary and where it can be positive when we come back. Meanwhile, your questions are welcome at 504-260-1870, and I'll pass them along when we return here on WWO. Having an interesting conversation about AI, artificial intelligence, as we always do with Andrew Schwartz, Professor of IT and the E.J. Uso College of Business at LSU. Professor, somebody texted in. I think this is a good question about... If it's fed conflicting information, who decides what's right and what's not as it relates to this data base that AI uses? There has to be a referee at some point, right? That's a good question. The answer is actually it's not a referee. So just like we can hold conflicting views of a phenomenon. AI can as well. So it's actually kind of modeling how we think about information as well. If you think about it, if you have a product, you have a love-hate relationship with it, the same thing can be true for AI. It can have both pros and cons of a position. Let's go with Wikipedia, for example. I don't use Wikipedia as a source because I know at one point people could go and put whatever they wanted on there. I don't know who's riding hurt on that or not. A, as far as you know, Wikipedia are a reliable source or not? That's an interesting question, right? The philosophy of Wikipedia is the wisdom of the crowd. So just in case, for those of you don't know Wikipedia, because believe it or not, my kids have never heard of this process. Are you kidding? It's basically an online encyclopedia, but if you find an error in it, any person could change it. Well, it's interesting about that is one of the sources of information that Shatchabit used was Wikipedia. But how are you defining an error that you can change it? If you, is it, you know, I understand what I'm saying. If I see something I don't agree with, I can call it an error and change it, but it's not really an error. It's just my opinion that's going in there as a fact. That's why I don't ever use it as a source. And this is where we're getting more meta when it comes to truth and what is true and what is not true. And this is why you also see so much of a question of skepticism of information that can be found online. Because one of the broader concerns we have is not only whether or not that information is true to begin with, but if AI does generate information and you do rely on that information and you make decisions on that information. then that's a potential problem, right? And more broadly, what about the role of misinformation and using AI to create misinformation? Who's going to be held accountable if that holds to be true? So is Wikipedia, however you say it, in a way, artificial intelligence in microcosm, because it started from what everybody put in there, and I don't know if there was one...

source that started this or not, but it's just an accumulation of all kinds of different things? It was an accumulation, and what's interesting, you brought up Wikipedia and manipulation. There's many stories over the years of this happening. For example, Microsoft was actually publicly found to be responsible because they paid an employee to go onto IBM's Wikipedia site. to change information to drive down the stock price. If that was happening in the era of Wikipedia, imagine what could potentially happen in an era of AI when AI is creating misinformation and potentially changing how we view phenomenon or information. If we go outside of our country, for example, at other countries where information is limited, you know, it's going to vastly change how they view history as they get exposed to information, whether or not be right or not. As we joke last time, North Korea just won the World Cup for the fifth time. We don't know what we don't know, but when it comes to stock prices, when it comes to non-disclosure agreements, et cetera, there's no telling what's going on now, right? Oh, 100%. We see this close to home right now, right? And in our state, with all the NDAs that have been about the metadata center, where we're starting to learn more about the tax breaks and some of the information that's come out with that as well. So I typed in just to see to play around during the break. When did Einstein go to Disney World? And it told me that Einstein never went to Disney World because he died before it opened. How did it come about that? that fact. How did he get that answer for me? You said somewhere along the way, one of the information sources was actually about Einstein, and so it then put together and stitched together two pieces of information to arrive at the conclusion. Then I put... And typed in, just to be topical, when will Earth lose gravity? Because there was this Internet hoax about Earth losing gravity for seven minutes yesterday, right? And it came out and said Earth will not lose gravity, viral rumor, claiming the planet would experience a seven wounds gravity blackout. I don't know what the wounds OONS means due to the League program. Is that a typo in there that was supposed to be seven seconds or what? Again, it depends on where they drew that information from. And if you question it, it was really interesting as what we're starting to see is the tide is starting to turn on this conversation, because what you're starting to see with chat CBT and some of the platforms is they're actually starting to cite where they got the information from. And for a lot of us, analysts, we think this is a really good move in the right direction. That would seem like the best way. But then again, it goes to really what is absolute truth, right? Exactly right. And where does truth come from? We do know, of course, from a lot of research that we tend to live in echo chambers and we tend to consume the information that validates our biases to begin with. So there's the potential that as AI learns more about you, it actually could amplify those biases that you already have in existence. He, in his piece here, he said in the first sentence, a second sentence, in the next few years people will be able to use superintelligence beyond human capacity. Talk about super intelligent AI and how far away we are from that.

Absolutely. It really depends, again, who you listen to, how far we are from it. But the idea of super intelligence that is at some point, if AI is able to consume all the world's information, it therefore has the ability to have intelligence of beyond what humans have. because of, but that's what I don't get. It might not be real intelligence, right? It's all what's fed into it. Exactly right. And that's why we always have to use the word artificial as the key part here. Because the intelligence that the system is using isn't real. It wasn't based on the experience of itself. It was based on the experience at what you've been taught of it. So how do you ride hurt on this? I don't understand how you can keep it from developing into something that you don't want it to be because there is no referee there's nobody you know I used to joke around Andrew saying that you know when you're a kid you think oh well that can't happen because the adults are watching or somebody's watching and then you realize as an adult it can easily happen because nobody's watching anything right is it the same kind of thing here Yeah, it is. And so my wife and I joke, we're always looking for the adult-year adult in the room, right? Yeah. And so it's the same thing with here with AI. And I think that's why a lot of us are calling for regulation at the federal level, because these are concerns that we all have. We're worried about data privacy. I'm worried about how these models are coming to be, and worried about how they're going to be used for nefarious purposes and otherwise. So there's a lot of need for a high-level regulation around this industry. So when it comes to the regulators, that's presuming you can trust them, right? And they're not, haven't signed some kind of NDA or are in cahoots with somebody to have some kind of profit coming here when they're supposed to be regulating it for the greater good, because there's already a lot of chicanery going on just in everyday life. Forget artificial intelligence as it relates to politics, profit taking, etc. Right. And I think that's the point of like the New York Times piece on what's happening here in Louisiana with META, which is what are we not seeing about some of these deals that are being cut and what are the implications of them as well. And again, this is not a nefarious comment. I'm not implying or suggesting anything far from it. But when we talk about regulation, we need to make sure that regulation, that the regulators. and those being regulated are fair actors in the process. And who's watching the watchers, right? And do they understand it? Because it's not as simple as selling a stock short. We're talking about really advanced calculations, algorithms here, et cetera. And does anybody really understand one person, at least, understand the capabilities of what this can be or do? Absolutely. And I think the message I think to listeners and to us as users these platforms is knowledge is key. And we really have to learn more about this. And I'm actually going to break a little bit of news here for you on air here, which is this morning we're going to announce free for all Louisiana citizens. LSU is going to be offering a free AI course, a fundamental course, where you can learn about AI. It's going to be released. They're going to send out a press release this morning, but as of August 26th,

We'll have a free course forever in Louisiana to learn about AI. That would be online? That would be online, absolutely. So we want this conversation to take place so that we can open these lines of dialogue up to understand it so that we're not scared of that we understand of it. Is the public woefully undereducated when it comes to AI? And I don't mean that in a negative way. I just mean it's... Go ahead. I don't, I 100% agree with you, and that's not a slight at all. And when I say that, I think that even us that are quote-unquote experts are a little naive to how this is going to unfold. I really don't think any of us really know where this is going to go. And that's why we need a level of engagement in the public to understand and watch over these platforms. Somebody texted in, I don't quite know what they mean by this, but is there any way this collapses on itself like a house of cards? Yeah, absolutely. And there's a big question right now, especially on the stock side of it. Are we seeing another inflation of the market where it's irrational exuberance all over again? And I can tell you the valuations are higher than they've been in a long time for AI-based companies. We're seeing a lot of investment in the valley around these kind of platforms as well. So could the industry collapse on itself? No. I think AI is here to stay. Could the technology itself collapse on itself? There's going to be some questions, I think, about how we consume information. More broadly speaking, I think we need to think about how AI may be going on in the background in ways in which we may not even know about it. Yeah, not to go too far down the rabbit hole here, but you could make claims about AI that aren't true that would inflate a price of a stock, and then when they turn out not to be true, you still could have made a whole lot of money on that, right? Absolutely, right? And it's more and beyond stock. Think about competitive dynamics and competitive moves, not at the corporate level, but at the country level, a country signaling they have an advanced model that can do more than any other country. We could be in the midst of an arms race from an AI perspective where each country tries to outdo each other in this world of AI. Which is none of it based in truth. which we don't even, and we won't know because of no transparency. And I think the concern at our country's levels, what if we start to regulate this, but other countries don't? Are we going to be anti-competitive as we develop new products? The truth is right now, though, if we look globally right now, I think America's leading the world right now in terms of the adoption of these platforms and the use of these platforms. So you could easily see a nefarious actor stepping up to try to get rid of us in that perspective. So as you mentioned North Korea winning its fifth world cup, There could be North Korea having some type of nuclear capability or that they really don't have, that they want people to think they have, which could lead to counterreaction based on something that didn't exist in the first place. That's exactly right. Now, this is not new. We've been doing this misinformation kind of campaigns for a long time. The difference is the scale. Now we're talking about something a lot bigger and a lot more global, because we can see more than we ever could before. Well, I'll tell you what, since we've been talking, I understand it a lot more than I did, and it's kind of frightening in a way, but I hope that somebody's watching something.

We all are, that's for sure. But at the end of the day, I'm an optimist because I see such potential for good. So if we can get the regulation figured out, I can really just see opportunities for us to tackle the bigger problems that we're confronting. In Louisiana, we have a lot of success stories. We've seen huge economic development for AI. We have companies working on autonomous drones and autonomous of vehicles. We have Waymo coming into New Orleans, so we have a lot of positive stories that are coming out from an AI perspective. And for the state, what that means is that we're going to be able to create a new economic base for us when it comes to AI. And that's what these data centers are so important for us. Thank you. Appreciate your time, as always. Andrew Schwartz, professor of IT and E.J. Uso College of Business at LSU will take a break and lighten up. Talk about where we're going to be at this weekend with Way Yet Magazine's Josh Danzig right now, 938. Time for traffic, WWL. 944-16 till 10 Thursday morning. Time to check in with Josh Danzig, owner of Wayette Magazine to see what's going on this weekend and find out where we're going to be at. How are you doing, Josh Wayette? Where are you at? Good morning. How are you? I'm all right, sir. Tell me about the magazine about the website. Yeah, where you have magazine. We have the Saints preview issue is out on the streets now, 650 locations across Greater New Orleans. You can find it at coffee shops, restaurants, bookstores, retail outlets. Then, of course, our website where you're at.com, which is constantly being updated with the minute info on entertainment, where to go. who to see, what to eat, and all the great things happening in the Big Easy. So where are we going to be this weekend? Where are we going to be this weekend? Lots going on. I think the big thing this weekend, we don't have a big red dress run, of course, or anything. That was last weekend, which was a huge, huge event. But you got a thanks game coming up on Saturday. A lot of people are going to the first preseason game. That's a great thing to bring the kids to. And I'm not sure, I guess there's some question as to whether the starters are going to be playing Tyler Shuck and so forth. But it's a great reason to get into the dome. We know it's a nice, cool, comfortable place to spend on a Beanie Saturday afternoon. And a great thing to do with the family. If you're not interested in going to the dome. You know, Josh, let me jump in for a second, because that really is a great time to take. the kids to a football game, a preseason game, because it's not as intense as a regular season game, may not be as loud if the kids don't like the loud noise, and if they want to get the experience, people a lot of times poo-pooh pre-season games, but that can be an opportunity if you look at it from a different way.

100%. And, yeah, you know, not as crowded. The tickets are usually much more affordable. And, you know, I knew when I took my kids, when they were younger, they were sometimes good for a quarter, maybe two quarters. And then by halftime or so, they were ready to head to a playground or the zoo or something more kid-friendly. So, yeah, it's a great thing to consider taking the kids too. They're playing the Jacksonville Jaguars who look to be competitive this year. you know, nice, nice reason to visit the dome. If you want to see the game, but maybe not go in person, they are having a watch party over at the Brong Iron on the Greenway, which I don't know if you've been to, but that's a really popular indoor outdoor beer garden. They've got tons of TVs, air conditioning indoors, and then lots of fans on the outdoors. So, yeah, that's a great thing to consider this weekend. Something else that's kind of new, and I know you're a fan of the paranormal to a degree, but the vampire apothecary, which is a great spot in the French Quarter, they're doing a new thing on Sunday evenings called Enchanted Evenings. This is an immersive Gothic storytelling. It's a speakeasy with ritual and lore all about New Orleans legends and... Vampire Apothecary is a vampire-themed bar. It's a really a great place to go to if you've never been. It's a great place to bring guests to. So on Sunday evenings, they're going to be doing a multi-room kind of storytelling experience. They'll have some signature cocktails. They'll have some season storytellers that lasts about 90 minutes. A great thing to do, kind of a mini vacation event. You can get tickets on their website at the vampire. Carrie, that's definitely something to consider. Fun things, the New Orleans Public Library, they're doing on Saturday something called Saturday. Well, you broke up a little bit. What is it? What day? Oh, yeah. So this Saturday, I'm sorry, the New Orleans Public Library at the main entrance, the main library, Sands Early Learning Center. They're doing something called Catterday. This is in conjunction with the SPCA. I don't know if you're a cat person or a dog person, but it's a meet and greet with cats. And it is kid-friendly. It's created around kids. Every child will also leave with an animal-themed coloring book. And that starts at 2 o'clock. Just a good thing to play with some cats and enjoy the library, which is a great facility. and the like. You might just take one home with you. You might wind up with a new pet. And you might just take one home. Exactly. It's in conjunction with the SPCA, which is a fantastic group they're going to be doing.

presentations to raise awareness for cat adoption. And like you said, maybe take home a kitten or a cat, which is a super thing. And then another neat event that I wanted to throw out is something taking place at the Desire Oyster Bar, Urban Street, that's inside the Royal Sanesto Hotel. So it's called Sip and Shuck. They're doing this every Sunday through August. So do you, Tommy, know how to Shuck an Oyster? Absolutely. Okay, well, not everybody does, and at this event, Sip and Shuck, they're going to have their head chefs, Kentrell and Chef Dwayne, teaching you how to Shuck an oyster. And you're going to have your own band-aids or not? Yeah, right. I'm sure they'll have some on site. And as you know, for people who've never shucked an oyster before, it can be a little bit tricky. But once you get it down, it's like riding the bike. So they're going to be providing a dozen Louisiana oysters that you're going to actually be able to eat after you shuck them. Really? Then, yeah. Shuck your own oysters? Up your own oysters, eat your own oysters. That sounds like fun. And then you're also going to have a nice glass of wine or a cocktail or a beer to accompany that. And again, they're going to be doing that every Sunday through the month of August from 3 to 4 p.m. And it's just a neat thing to do. That is a great idea for people that have never shucked an oyster. You can shuck your own oysters, but they show you how to do it first, and I presume they give you everything you need, like the knife and so forth, or lead if they use it, to open the oysters. That's right. That's right. They provide everything. They give you hands-on training. And then when you're done shucking, you get to eat them and enjoy them and then learn the new skill, which will come in handy in other parties and events in the future. And I'm guessing make as many Shucks jokes as you want, right? All Shucks. Tyler Shuck, whatever you want. That's right. We'll let you run with those. Dickie Brennan's Steakhouse going on, something there? Yeah. So every year, Dickie Brennan does something called the Steakhouse Summer Picnic. So if you love steak and beef, this is an event for you. So this is a showcase of Louisiana beef in conjunction with local farmers. And it's taking place on Saturday at the Steakhouse. You can get tickets online. It's a summer tradition. They've been doing this for a number of years. They're going to have live music, and then they're going to have food from six of their chefs from their different restaurants, which are Dickie Brennan Steakhouse, Bourbon House, Tableau, Pascals Manali, Auduban Clubhouse and the Commissary. Each one will be featuring a different dish. You're going to find things like sliders. You're going to find things like meatballs. There will be carving stations and then lots of side dishes.

And it's a great party. You've got cocktails, you have drinks, you have wine, and just a great way to go out and learn about Louisiana grass-fed beef and the local ranchers behind it. Tell me about sip and dip at the zoo. Sip and dip. You would love this one, Tommy. As most people know, Audubon Zoo has the Great Cool Zoo, and that's a super thing to take the kids to this weekend. It's a great, lazy river, all kinds of fun little water games. So they've been doing something called Sip and Dip, and that is a 21-and-over event. It's going to be tomorrow evenings from 7 to 10 p.m. And this is basically a water party, a pool party for grownups. Sounds awesome. You bring your bathing suit and you just get ready to have some fun. You bring your friends. They're going to have cocktails. They're going to have music. You hop in a floaty and you go around that lazy river. So every week they have a different music theme. This Friday will be EDM night, which EDM is electric dance music. So they'll have Dr. Romance. And it'll be a little bit of a more. scale or I'd say high energy music scene. But they've also had things like 70s night. They've had things like country night. But you can get VIP cabanas and chairs. And what a great way to spend a Friday evening with friends and with your grownups and all that kind of great stuff. Take us out. Josh, tell us about where you had real quick. Yeah, Worryette magazine, pick up a copy. Visit us at where you at.com right on the homepage. We have the Wariat weekend picks. You can find that. And something else mentioned, a red dress run. If you go to Wariat.com and you click photos, you'll see our photos from last week's red dress run, including some really funny ones of Newell wearing a red dress so you can give him a little bit of a hard time with that. He's with Superintendent. and put Kirkpatrick and some great photos, so you've got to check that out as well. He's sitting right next to me, but he didn't hear you, but I'll pass it along. Thank you, Josh. Appreciate it. Quick break. We'll find out what Knowles got planned when we come back on WWL.