EP 288: Big Tech's Critics Have Gotten a Lot Wrong on AI

This is the Everyday AI show, the everyday podcast where we simplify AI and bring its power to your fingertips. Listen daily for practical advice to boost your career, business and everyday life. There's critics all the time talking about AI on all sides saying, oh, you know, in the future, we won't need humans, we'll have AI or people. People on the other end of the spectrum saying, hey, this artificial intelligence thing, it's brand new. We shouldn't pay attention to it. So today I'm excited to have an industry veteran with decades of experience in the technology field to come in and help us dispel some of these big critics who are just getting a lot of this wrong about AI. All right, I'm excited for today's show. What's going on, y'?

All?

My name is Jordan Wilson. I'm the host of Everyday AI and this is for you. We, we are your daily guide in the form of a live stream, podcast and free daily newsletter helping you all learn and leverage generative AI to grow your companies and to grow your careers. So before we get into that, I just have to put in the plug. If you haven't already, make sure to go to your everydayai.com each and every day. We recap our interview in our free daily newsletter, so make sure you go sign up for that. All right, before we get into today's topic, let's do as we do every day and go over what's going on in the world of AI news. So first, Elon Musk is bringing a lot of AI to Memphis. All right, so billionaire Elon Musk has announced plans to build his AI supercomputer at the former Electrolux site in Memphis through his startup xai. So Elon Musk startup XAI will be constructing an AI supercomputer at that former site in Memphis. The project has been dubbed the Gigafactory of Compute, and it marks a significant investment in AI technology and could have some pretty big positive economic facts on the city of Memphis. So there is no exact details on the cost, but it has been called a multi billion dollar investment. So we'll see how this will compare to the reported collaboration between Microsoft and OpenAI, the 100 billion dollar Stargate Compute Project. All right, next, actor Ashton Kutcher has access to OpenAI's Sora and said it will change Hollywood. So actor and entrepreneur Ashton Kutcher has praised the OpenAI's Text to Video program Sora, a generative AI video tool, for its impressive capabilities and said it has the potential to revolutionize the film industry. So Kutcher actually has a lot of involvement in the AI space. People don't know about that through his venture capital firm, Sound Venture. So he's not new there. So in the recent interview, Kushner praised OpenAI Sora for its ability to generate realistic videos, potentially reducing the need for expensive CGI and stunt performers. All right, last but not least in AI news, at least for our recap this morning, and this one's a big one. Nvidia has officially passed Apple in market value and has become the second most valuable public company in the world, at least for now. You know, that could change like 10 times today, but Nvidia's market value has risen rapidly due to bets on its potential success in the AI industry. Also, there is an expected stock split by Nvidia that's expected to generate even more demand for for company shares. Their growth is a reflection of overall optimism about AI in the business world and comes just a couple of days after. We've seen multiple days where they've increased market cap value by 5%. Multiple days and some a little bit of original reporting here that literally no one is talking about. So yes, Nvidia has become one of the first. You know, I think there's only three companies to hit the three trillion dollar market cap value, but they did just break the record for the fastest company ever to add a trillion dollars to its market cap. So we looked at this yesterday, whether it's going from one to two or two to three. But Apple, it took them 15 months to add a trillion. Amazon, 10 months. Microsoft, about two years. Google, 22 months. Nvidia 37 days. So just it's been a wild ride over the last about five or so weeks for Nvidia. So make sure we're going to have all of that in more in today's newsletter. Make sure you go sign up for it. All right, that's enough for AI News. I'm excited for today's guest and we're going to talk about what big text critics have gotten wrong on AI. So here we go. Please help me welcome to the show David Michella, who is a senior fellow at the Information Technology and Innovation Foundation. David, thank you so much for joining the show.

Hey, great to be here.

All right. Hey, for, for people who don't know and you have a very diverse and deep background in the technology fields, but tell us a little bit about what you do at the Information Technology and Innovation Foundation.

Well, the foundation is really there to try to make America a more innovative country. So we try to develop ideas and policies that advance Americans technology interests and obviously that's very relevant today. So we look at everything from AI to competing with China to sustainable energy, just all the things that really help to determine America's innovation future.

And also we'll, we'll be sharing this in the, in the newsletter today. But tell us a little bit about the, about the new book that you have.

Yeah, the book is called Technology Fears and Scapegoats. And it's basically a series of 40 essays taking on what we see as the myths that are used to tarnish the, the image of the tech industry, that it's polarizing the world and destroying privacy and increasing biases and wiping out the middle class and harmin children and all of these things that is routinely accused of. And it's just pushing back on those to show how the benefits in all of these areas tend to be underestimated and the downsides, which sometimes are real, tend to be exaggerated. So it's a balancing act to really try to tell a much more positive story about the impact of tech on America and the west so far.

And I'm, I'm extremely excited for today's conversation. And hey, as a reminder for those of us, those of you joining us live, get your questions in now. You know, we have a tech veteran here with many decades of experience and, and I'm excited to get into some of these common misconceptions. But David, let's, let's maybe start here. What are some of the things that big tax critics are the most wrong about when they're talking about AI?

Well, it's funny because someone's been around for a while. AI has long been the holy grail of the tech industry since people like Marvin minsky and John McCarthy in the 50s and 60s to all the smart people, they've been trying to make machines as smart as they can. Just as we get to the cusp of doing that, people start saying, well, maybe this is a lousy idea after all. And in our view, there are really three problems with that view. The first is that the fears tend to be exaggerated and the benefits underestimated. So they're telling an overwhelming more negative story than we think is true. The second is that, and really the west really needs AI, we may need all the AI we can get to remain competitive, but also just to take on the societal challenges of the future. This almost certainly will require sort of a redeployment of people and assets from much of what is going on. And we use the analogy of people shifting from farms to cities, or people shifting from offices to rebuilding society along many, many dimensions which we can get into. And the third that I think is really relevant is that the reality is that there are very few areas. There are a lot of things AI can't do, and even the things that it can do very, very well. And we're big fans of AI in every case that we can think of. The combination of a human and AI is superior to either the human alone or the machine alone. And this idea of looking at AI as sort of a cobot, something that works along side of you, is something we see all the time and seems to us by far the most likely pattern going forward. And that's a great story, we think.

Are you still running in circles trying to figure out how to actually grow your business with AI? Maybe your company has been tinkering with large language models for a year or more, but can't really get traction to find ROI on gen AI. Hey, this is Jordan Wilson, host of this very podcast. Companies like Adobe, Microsoft and Nvidia have partnered with us because they trust our expertise in educating the masses around generative AI to get ahead. And some of the most innovative companies in the country hire us to help with their AI strategy and to train hundreds of their employees on how to use Gen AI. So whether you're looking for ChatGPT training for thousands or just need help building your front end AI strategy, or you can partner with us too, just like some of the biggest companies in the world do. Go to your everyday AI.com partner to get in contact with our team. Or you can just click on the partner section of our website. We'll help you stop running in those AI circles and help get your team ahead and build a straight path to ROI on gen AI. You know. Yeah, it's, it's a good point you bring up there, David. You know, I, I think when people think AI, they think, oh, it's, it's brand new. With Chat GPT, it's like, you know, the first chatbot was eliza in the 1960s, right? You know, but one thing that I, I'm always divided on because I think some of the smartest people in the world have this take and then there's people that don't know much have this take, but there is this thought, you know, that hey, in the future, you know, us humans, we may not even need jobs, right? Like, you know, Elon Musk famously said two weeks ago that jobs would be optional. And you know, people like Bill Gates and Sam Altman have kind of opined on this. Do you think that those, you know, not saying those people are wrong, but do you think that that train of thought where people say, hey, humans aren't really going to be needed in most jobs, is that also where people are going wrong?

Yeah, I mean people have been predicting that technology would eliminate jobs since the Industrial revolution over and over again, since originally the so called Luddites and the textile mills and Revill. And there's never been any evidence that that's proved true. You have temporary dislocations of workers, but the economy itself expands and creates more work. And we don't see any reason why that's not true with AI at all. And as we say, is AI going to repair a road? Is it going to install plumbing? Is it going to grow here? Is it going to convert a high rise building into a vertical farm to grow food? Is it going to repair the environment or all the war torn and damaged cities around the world? Is it going to do all those things? No, it isn't going to do all those. Well, it may contribute to them, but humans are going to be needed to do that. Plus it's always been true that when new technologies come around, humans find other things to do. And you don't have to predict the industries of the future to believe that very likely that we'll have them, because we always have. And so the entire history and most of the arguments about the future seem to us as overwhelmingly positive. And when we listen to even tremendously capable people like Musk and others who predict sort of the downsides, when you actually hear their arguments and their downside examples, they're not that they don't really make sense. People talk about student plagiarism and they talk about deep fakes and these things sort of seem quite manageable to us. So the big fears seem really highly, highly speculative in contrast to the benefits which we see is quite real today.

Yeah, and I'm gonna, I'm gonna throw something up on the screen here for our live stream audience. You know, you said something there David, about hey, can AI repair a road? There's kind of this, you know, viral thing billboard that said, hey Chat, GPT finish this building. A building under construction. Right. But I guess on the flip side of, of that, David, you know, maybe to play devil's advocate and looking at future technology, right? But you know, obviously the AI systems, the large language models themselves are becoming, you know, more more powerful in theory. You have things like computer vision, you know, getting better and better, more affordable as well. And then you also throw in this, you know, speaking of Elon Musk, you throw in this, the humanoid robots, right? And then people start to connect those things and they say, okay, well, could AI replace para road? Well, if you have an autonomous vehicle and you have humanoid robots with computer vision, I, I don't know. Could they? So, you know, as we look at the future of technology, right, and I know it's hard to predict, no one has a crystal ball, but is it crazy to think that some of those even jobs could in theory be done by whether it's AI or, you know, humanoid robots, factories, etc? Could human jobs just maybe look completely different?

Well, they might, and everything you just said might happen, but there's so many steps to get there that worrying about today seems, etc. But to me, almost even a bigger part is that we actually need some of that to happen. The world is looking at an environment, particularly in the west, of aging and declining populations. And if you didn't have the mass migrations we have, they'll be dropping even, even more. And when you get into that mode, you do need to automate a lot of things. You do need a lot of robots. And there's sort of the marquee examples of that today are if you look in, in Korea, North Korea, South Korea rather, and Japan, where the populations are going to drop very, very sharply in the next 20, 30 years, and their use of automation and robotics is now how they plan on coping with the changes. And, and so, yeah, some of that will happen, some of it needs to, to happen. But, you know, the speed, the ability to do that is still pretty far out in the future.

That's such, such a good take. And it's something that maybe doesn't get talked about enough, you know, especially here, you know, in the U.S. you know, people have been talking about the impact of, you know, kind of baby boomers retiring. It's. I, I forget the name Silver Streak. Right, but. Interesting take there, David. But you know, I'm, I'm curious as someone that has been in, you know, kind of market research around technology companies for many decades, how, how would you explain the potential impact of, you know, not just AI, but generative AI? Because I think sometimes people, yeah, they say, oh, you know, they go to the Industrial revolution or they go to the Internet or they go to the cloud. Is it fair to maybe compare today's generative AI to current techno, like technological revolutions of decades past? Or would you say maybe that generative AI and where we're at today is maybe a unique enough case that you can't really compare it. It's not apples to apples.

It's a very fair comparison. The industrial revolution versus the information revolution. How can you not compare those two things? But the one thing we would say is that it's always been true and I think is true today, that the physical world of food and shelter and transportation and you know, and energy, that these things are fundamentally more important to the lives of people than information. The emails and texts and e commerce, social media, all of these things are really interesting and important, but they are of a lesser impact on most lives. One of the examples we use is that if you live in a hot climate, a lot of people wouldn't trade their air conditioner for the entire Internet. That there's a certain hierarchy of needs and those needs are really being seen now in the wars in Ukraine and what it's done to food and energy and you're seeing with China what it's done with supply chains, that these things are vital to people's day to day lives. And ChatGPT is fantastic. I use it a lot. It is incredible what it does. But if I didn't have it for a month, it would have almost no impact on my life. In fact, it would have no impact on my life or take any of those other things away. And so people, because it's the industry of our time, it's the growth, it's the jobs, it's the money, people tend to think that it's more important than the changes of the past. And we would argue they're actually fundamentally less important than the changes. You know, clean running water, refrigerators and ovens and electricity and lighting and cooling and appliances and all these things that define the last, you know, first half of the 20th century. Though to most people those things are not just more, but they're head and shoulders more important than anything going on in the entire Internet. And so we don't say that to demean our industry. I mean we all spent our careers in it. It's done phenomenal things, but it's easy to overestimate our importance when you, when you live in that bubble.

David, you bring up an interesting point there, right? So you said, hey, if, if Chat GPT goes down, it's not really going to affect your life. And interesting timing because about two, two days ago, you know, Chat GPT had some major down, you know, downtime perplexity did as well. Claude did it well, cloud all kind of at the same time and a lot of people online had the opposite, you know, of take as you did, they said, wow, it's hard for me to work right now. So they've become maybe too reliant on AI. Maybe their employers or department heads have implemented it too much. So might there be a time in the very near future where people and businesses become maybe a little too reliant on AI and kind of the expectations for productivity and KPIs increasingly go up because of AI? Is, could, could that be problematic if we're all relying on two or three systems?

I'm sure it could. In the same way that we're all dependent on the Internet to do most of the work that we do these days. If you're in the nature of the jobs that, that you and I have, that you can't really work, the Internet goes down. You can't really work if the power goes down either. And so, you know, those fundamental infrastructure services we rely on, and there almost certainly will be a time that if, you know, well, you know, you look chatgpt right now, but you know, in the future there's going to be multiple other. There already sort of is, but there'll be many other forms of AI and like any business system, they go down, people will have a hard time working. But, you know, the levels of reliability in the tech industry are pretty impressive. You look back over the years and how many times has Google or Amazon or Microsoft gone down for any significant length of time? And he. You got to give them a lot of credit for that. It's, it's pretty amazing. So, you know, the backups they have, their ability to replicate things around the world, it is pretty high levels. The biggest problem they would have would be the same problem that the economy has. They don't have the energy to run the systems and some sort of energy disruption could stop it all in a fundamental way. But reliability is something that every layer of infrastructure has to do. Electricity has to run, TV stations have to run. Everybody has to run. And we generally do a good job of that. That's overall, I don't think been a huge problem.

David, something else as we talk about what critics, you know, big tech critics get wrong maybe about AI, it seems like there's a lot of fear and maybe some of this is timing with the election here in the U.S. but there's a lot of fear or maybe attention or conversations happening specifically around AI deepfakes, right. And what that means for our society and our ability to decipher, you know, what's real and what's fake are the critics who say that this is, you know, kind of something you can't miss. Like, oh, deep fakes are going to, you know, take over mainstream media. Like, are they wrong on that, or are deep fakes really that dangerous?

I would say, you know, of all the things that people worry about with AI, you know, copyrights and bias and surveillance and automation and wiping out jobs, I actually think deep fakes are the most real right now. I mean, that's a serious problem. The technology is amazing. They can fake a lot of things that make them look extremely real. And that is an issue. But how big an issue? We're already deep into the 2024 election cycle, and has it been a major factor so far? I would say no. And people are already waking up to it for a long time. People had to adjust it. Well, this thing is a photograph. But there was this thing called, you know, Photoshop that could make things look real and, and people adjusted to that without a whole lot of issues. And deep fakes, they're better, they're. They're harder to detect. There's no question about that. But people aren't stupid. They say, all right, the first time, you know, a year from now, people see something, they go, well, is that real? And how do we determine whether that is real or not? And there'll be whole processes about that. It might even do a good thing of getting people to rely on less on just any random Internet stuff and more on allegedly trusted sources. And, you know, a lot of things could. Could happen along those lines. But, you know, the. And a lot of it's just the lawsuits. You look at Scarlett Johansson and others just suing, saying, you can't do this. They're going to have a strong case. And people doing that in America, I think, will have real problems. Foreign actors hiding behind, you know, proxy servers, whatever, you know, that's. That's going to be an issue, but there's still no evidence that that's an unmanageable issue. And upside, deepfakes is there's so many really funny ones. I mean, it's a tremendous source of entertainment and comedy and satire and. And such things when, when you know that they're sort of just joking. So, you know, it's a real one. I say I think it's the most current, real one. But if you look at their perspective, is it swaying voters between Trump and Biden? Not that I can see.

You know, speaking of, you know, kind of AI everywhere. And, you know, I love David, that, that analogy, you know, Photoshop and. Versus photography and you know, people eventually kind of caught on that you can't believe every single photo. Right. But one thing that I kind of think on, on the flip side of that is I think people had this, maybe this extended time or extended period to even just get used to the concept of what it meant for something to be photoshopped. And maybe that was before, you know, you had, you know, social media that things would take off and people wouldn't vet to see if it's real or not. So in the, in the day and age where you know, things can in theory go go viral in, in minutes and not everyone, right. So maybe I think sometimes I live in a bubble and I just assume, oh, everyone knows AI deep fakes, but you know, a lot of studies show that the overwhelming majority of even Americans aren't using any form of AI on a day to day basis. So I guess how can you balance like those things in terms of the impact of, of AI technology on something like deepfakes with hey, is the world really ready with how quickly kind of misinformation or disinformation can, can spread?

Yeah, first of all, Americans are using AI all the time. They just don't know that that they are that every time you go to Amazon and see how they summarize their reviews for each product, that's a tremendous example of just mass scaled AI to deliver tremendous value that could never really be done by humans. Same thing with TripAdvisor and all those sort of services that summarizing those things and making it easy so people are using it and that'll be predominantly the case that most of the uses you'll do, you won't even know is driven by AI. And that's generally a good thing. But your real point is about speed. And speed is really interesting. As you said in your opening remarks, look how fast Nvidia has reached the scale it has does it is faster than the past in terms of the acceptance of the current generation of AI products. But as I said in my OpenHearts, AI from a conceptual point of view, if you go back to neural networks, goes back to 1940s and so a lot of this stuff has taken a great deal of time to reach any sort of useful and critical mass and it's now reached that scale. And you know, maybe a useful analogy would be, you know, there was all kinds of mobile phones for 10 or 15, 20 years and most people didn't use them much. They were good, but all of a sudden the iPhone comes along and breaks a barrier and all of A sudden boom. Everybody has it. And you could say, well, now iPhones are accelerating at a great pace. But no, they just had a phase change that brought them to the masses. And that's really what happened. That's what happened with Chachi. PT is not a new innovation. It's not a new, really anything conceptually. It just has the scale of processing power and data to deliver that as a service through the cloud in a way that couldn't be done before, essentially the cloud era. And so speed is a tricky concept. And you could argue that AI is really fast, or you could actually argue that's actually really slow. And, and we just happen to be in a flourishing phase. And. But, you know, looking ahead, you can expect pretty rapid developments, just as you've seen with, you know, the, the Internet and mobile and all these things.

You know, I, I think it becomes easy at times, David, to react to, you know, critics who are talking about AI. Right. I feel it's always, you know, hindsight's 2020 to sit here and say, oh, yeah, they, they have this wrong. But maybe if we were to flip, flip the script and say, okay, you know, David, now, now you're the critic, right? Tell us about AI. So, you know, how do you see this unfolding? And again, not asking you to, you know, crystal ball this, but, you know, as someone that's been following trends in technology for many decades, how do you see this unfolding in the next couple of years? Because I think a lot of people feel this uneasiness and this uncertainty of what generative AI means, not just for their jobs, but for society and for our relationship with the real world. How might you see this all unfolding?

Yeah, to your last point about people, I always just tell people to relax. If you look at how much tech has already happened since the Internet started, you know, most people really like that. You know, the real consumer Internet started roughly 1995. So you got 30 years there of things that consumers really like, and policymakers and critics and media do a lot of tech bashing, but the average consumer doesn't do much of that. They say, wow, Amazon is great and iPhones are great, and Google Maps is fantastic. And they just use all these things, many of which are free. So if you liked all of that last 30 years, you're going to like the next five or 10. The ability to translate any language, which is the most fantastic thing that, you know, in school, if you, if you use chat for student, you want to try to get some ideas about something to write, it's the most fantastic school aid compared to any old things people have ever had. So students will absolutely love it and I think a lot of teachers will too. And so I think for the short term there's going to be a lot of tremendous benefits. In the long term, you know, I'm not really a long term for the site that's sort of the sci fi world, but there are concerns. You could have something like what you see happening in China, a surveillance state where everything is tracked and you have social credit systems that monitor your behavior and punish you for bad things, in theory will reward you for so called good things that could happen. But to me that only happens if we want that to happen. And this is sort of why a lot of the work we do at IGIF is about policy that you don't want to go down that path and that requires policies that have sort of American versions, Internet type versions of that, rather than, you know, not to pick on China, but Chinese type versions of that today. And you know, that's a dystopian future potentially. And China could always reform and then it doesn't have to be horrible. But if you look at the path they're on now, that's path we want to avoid. And a lot of our work is about how to do that over the longer term and maintain the freedoms and choices and benefits and minimize the downsides. And you know, that's been true for the Internet for there's always been fears about the Internet but you know, they're, they're deeper and more real as the technology gets more powerful. So you know, that's something that keeps us awake at night and thinking about.

Speaking of things to think about, David, you've, you've given us a lot of those. So in, in today's episode you've given us, I mean we've talked about the future of jobs in an AI world, you know, potential dangers of deep fakes where AI is currently trending. But you know, as, as we wrap up today's show, what's maybe the one biggest takeaway that you want people to leave with when it comes to even just what big tech critics are getting wrong about AI?

Well, one thing I would just expand that things that text critics are getting wrong about AI are the same things they're getting wrong about technology in general. And that America has historically been a pro innovation, pro technology, society and economy. But we're moving in a world where that's less and less true. And we're now becoming an economy that often seems more concerned with preventing potentially harmful change than potentially exciting benefits. And that that has the risk of really slowing innovation in our society. And you know, the bottom line is America's dominated the tech world for just about its entire history, but we now have a real competitor in China. And China is going full bore on all of these issues and they may use it in ways that we don't like, but the underlying technologies are largely the same. And whoever is going to lead in those technologies is going to be very powerful player in the world. And when we hear people talk about putting the brakes on AI and slowing innovation and all of these things, to us that sounds like we might seed the field to rivals, not just China, but India and potentially others. And to us, that's a future we want to avoid. That America has benefited tremendously by dominating the tech world for so long and so deeply. But you can now see paths where that might change. And having attitudes that want to prevent that is just really important. And, and that's really the message that we do through our ITIF work is to try to encourage America to become remain the innovation leader in these areas.

So good, y'. All. My, my, my forearms are burning from typing so many notes from so many great takes from David. So, David, thank you so much for your time today and for coming on the Everyday AI show. We really appreciate it.

Thank you.

All right, hey, as a reminder, everyone, yeah, we covered a lot and that's only a fraction of it. We're going to be breaking down today's episode and giving you everything else that you need to stay up in the world of AI. So make sure if this was helpful, please let someone know. Let your co worker know if you're listening on the podcast. Thanks for your support. Check out the show notes you could leave us rating on Spotify or Apple. And if you're listening here live on LinkedIn, please consider sharing this. If this was helpful, repost this and thank you for tuning in. Go to your everydayai.com Sign up for the free daily newsletter and we'll see you back for more Everyday AI. Thanks, y'. All. And that's a wrap for today's edition of Everyday AI. Thanks for joining us. If you enjoyed this episode, please subscribe and leave us a rating. It helps keep us going for a little more AI magic. Visit youreverydayai.com and sign up to our daily newsletter so you don't get left behind. Go break some barriers and we'll see you next time.