← Digital Disruption with Geoff Nielson
Digital Disruption with Geoff Nielson
One of Us Thinks AI is Overhyped. The Other Doesn't | Live Debate
00:27:38
Loading summary
Transcript48 lines
- [00:02]
Jeremy Roberts
OpenAI, which has never turned a profit, is scheduled to burn $218 billion by 2029. The business model is not sustainable.
- [00:15]
Jeff Nielsen
We can debate exactly when it's going to completely change the world, but it's a when it is not an if. AI has like one big bucky, some of the worst marketing in the world, like rooms full of people bigger than this are literally booing AI. And so it feels a bit strange to be a little bit of an underdog making the case for AI. And by the way, the case I'm going to make today about it not being overhyped, I'm not making a financial statement, I'm not saying everybody invest in AI today. I'm talking about the technology. That the technology is going to change the world and that it's already changing the world. So let's start by talking about the now and let's talk about your work and your world. So let's start with your work. To me, right now, what we're seeing in organizations is that AI is endemic and it's contagious. It's endemic and it's contagious. What that means is that people are adopting it at an extremely rapid pace and they're continuing to use it. When you look at employee productivity metrics, we're seeing that when we actually measure this quantity of work and quality of work are both going through the roof. We're looking at task completion increasing by close to 55% by a Microsoft survey, improving of quality of work by 20 to 40%. MIT and Stanford came up with that. And 60% of AI users are saying they've been creating something that they couldn't have done a year ago. And I can tell you from my lived experience, the just output in volume I'm seeing, even the last weeks and days of slides that are AI, of emails is through the roof. And for anyone who I work with, I see you, I see your AI content that you're sending me. The second piece here is the functional transformation AI is doing has already begun right in it. 70% of IT departments are already using AI. 55% of security and nearly half of all developers are using AI right now. And in the business, it's not far behind. 50% of customer service teams are using AI. Sales and marketing and HR are both at 40 plus percent. And the numbers are just great. Now when I look outside the business, it's what's already happening right now. It's already being used for frontier science, right? We're hearing about, you know, MRNA vaccines were Taught for Covid. It was a key piece in getting the vaccine out. And Bob Leak just yesterday said that AI is already being used for the chemotherapy plan for his wife. It's changing the world. There's another story I love that I read about that's a little bit less popular, but AI is being used now to improve weather system forecasting, and it's improved forecasting by about 20%. And this is saving lives because if you can more accurately predict extreme weather, you can tell people to evacuate, you can give them the right information in a timely way, and it has dramatic human impacts. The other piece here that you may have seen is autonomous vehicles, right? These are powered by AI. Right now, Waymo is already gonna be in 15 US cities by the end of the year. And by the way, I don't know if you saw this outside the hotel, but we've got autonomous vehicles this year for the first time in Las Vegas, being deployed all over the city. It's happening right now, and it's changing our lives. And already we're starting to see industries change, whole industries. The design industry has changed overnight. The legal industry has changed overnight. I could go on and on, but let's talk a little bit about what comes next. So in your work, the agentic enterprise is going to be happening to us, right? The rewiring is already starting. As I said, it's endemic and it's contagious. And the cool thing about agentic that you may have already heard at this conference is that there's a recipe for it, right? We've been walking members through it, IT leaders through it, and we can start to build it for you in four days. And we're seeing people tell us we're getting 100 grand of value in four days, 9.5 out of 10. The recipe exists. It's already beginning, and it'll grow from there. And then I'm most excited about true differentiators coming from AI. So if you think about what happened when we moved to the Internet, as Joel said, when the web came, when we moved to digital product offerings, we're already starting to see this, and it's only going to unlock more and more. And then in the world, AI is becoming a utility, right? This notion of being able to have intelligence on demand, that's unmetered. It's already started, and it's changing the way that we do everything. And then the other piece, you know, I could give you 100 examples for this. I'm going to try and keep it brief. Health care is the one that's personally most exciting to me. And if you were in the session I did earlier, we're already starting to look at how doctors can use this to become more efficient. And you know, we're seeing when we do trials, AI reduced burnout in doctors by 10 points. And if we can get better accuracy in the way we're diagnosing, studies have shown already that even catching an Additional by using AI, you can catch an additional 6 million diagnostic errors in the US per year. That prevents 200,000 deaths, right? 200,000 more disabilities on top of that, more personalization, broader access. This is a sea change. It's extremely exciting and I think there's no two ways about it. And with that, I will pass it over to my opponent.
- [05:16]
Moderator Sam
Thank you very much, Jeff, and looking forward to hearing from you now. Jeremy.
- [05:21]
Jeremy Roberts
Well, so when they asked me if I wanted to do this debate, they framed it this way. They said, would you like to tell your boss, Jeff's my boss. He wrong. In front of thousands of people. And I said, I can't think of a better way to spend a day at Live. So let's have this conversation.
- [05:36]
Infotech Announcer
If you don't work in enterprise technology, you can skip ahead or take a quick braincation. But if you do, Infotech is your secret weapon. Infotech helps IT teams get projects done faster, better, and at a lower cost. Whether you're rolling out new tech, improving processes, or just looking for cost savings, Infotech provides unlimited access to practical tools and expert guidance. You need to execute at a fraction of the cost of traditional consulting. No matter the project. From AI strategy to cybersecurity to vendor negotiation, Infotech has you covered. Check it out at the link below. And don't forget to like and subscribe.
- [06:13]
Jeremy Roberts
I agree broadly that AI is very exciting and very interesting, but I have a slightly different perspective. And I think I have a sneaky, sneaking suspicion that many people in the audience will agree with me. Don't worry, everyone. This is a safe space. So the emperor has no clothes. And one time I pitched actually a tearaway suit and I was gonna anyway, vetoed. Somebody said you can do that at a different conference. So I've got a few points on this that I think y' all need to hear. Number one, many people who work with AI do not like it. If you do a poll of Americans, many of you in the room are Americans. Some are not. 50, 50% of people have a negative opinion of it. And AI has huge awareness. Everybody and their grandma and their son and their dog is using. This is familiar with it and they do not like it. There are very few things you can get 50% of people to agree on across the entire country. That's a Pew report. You can look into it. People don't like it. So if it's so great, why do people hate it? That's number one. Number two, the business model is not sustainable. Hey, Jeff, can I borrow $300 billion from you? You can just give it to me backstage. Don't worry, I'll funnel it back to you via Rob, and then you can use it to buy something from me. Look how much value we just created. Thank you. Thank you. In this example, Rob's Nvidia OpenAI, Jeff's Oracle. I mean, come on. Come on. This isn't a sustainable business model. Hey, Jeff. Hey, Jeff. So I'm debating him, so I got to go after him. Have you heard of Tesla?
- [08:03]
Jeff Nielsen
Tell me more about Tesla.
- [08:04]
Jeremy Roberts
I'm going to tell you more. I wrote this down. I didn't want to get it wrong. Tesla first turned a profit in 2018 after burning $9.3 billion. That's a lot of money. Less than you're lending me, but it's a lot of money. Uber famous venture Capital Darling burned $18.2 billion before it eventually turned a profit. OpenAI, which has never turned a profit, is scheduled to burn $218 billion by 2029. That's a little underhyped. I think there's a little more hype that we can build into that. So I know you didn't want to talk about the financials, but this only works if we can actually pay for it. Many people who work with AI say they're saving little to no time. I showed a chart yesterday. If you ask executives if they're saving time, 20% of them say they're saving 12 hours a week. If you ask people who work at the working level, who don't have any reports if they're saving time with AI, 40% of them say they're saving nothing. And I know you quoted that Microsoft study about productivity, and it's funny because I went to a Toyota dealership and I took a quiz on what car I should buy, and they said a Toyota. Come on, Jeff. Thank you.
- [09:19]
Jeff Nielsen
Thank you.
- [09:19]
Jeremy Roberts
I worked hard on that. We focus grouped it. We hear a lot about AI automating the boring stuff. The boring stuff is how we train juniors. No juniors today, no seniors tomorrow. It can be an individually rational thing, but bad for society. I saw one of my colleagues talk about the diamond shaped organization because we Cut the bottom out. Is that actually a good thing? Does anybody actually want that? I don't know that I do. I think it's overhyped for now. It's not even cheaper. We're going to save money. Are you a mystery organization? This is reported, I think, in Politico or axios. They burned $500 million in AI credits in one month because they forgot to put guardrails up. I haven't seen any Mega Caps doing amazing things. So I think some of that money might have been wasted. It's not even cheaper. Imagine paying more money for what is often a worse output. And then, finally, this is maybe a more philosophical point here. Dehumanizing interaction isn't a net benefit. I want a show of hands. Who here likes getting a chatbot when they need support?
- [10:26]
Jeff Nielsen
No, I don't see.
- [10:27]
Jeremy Roberts
Do you see any hands, Jeff? Gotta look that way now. I don't see any hands. Nobody wants this. The people pushing this aren't pushing it for them, they're pushing it for you. If you get a robot that says, I really feel your pain, the next agent will be with you. That's not exciting. That's not a good thing. You ever heard of a Waldorf school? Lots of wealthy people send their kids to Waldorf schools that don't use technology at all. And guess what? Before I pass it on, because I know we've got to have a little bit of a discussion, the reality that I'm seeing on the ground is that there's a lot of people out there who are very excited about this technology in the abstract, but the excitement they have isn't concrete. And to me, that's the definition of being overhyped. So I'll leave it at that. And I think maybe I'll click to this next slide here and let's have a discussion. So, Rob.
- [11:18]
Moderator Sam
Absolutely, yes.
- [11:19]
Jeremy Roberts
Pass that back to you.
- [11:21]
Moderator Sam
All right, well, Jeff, you've got some time to disposition there. Jeremy has clearly won the hearts of the people by the interaction.
- [11:31]
Jeff Nielsen
Thank you.
- [11:32]
Moderator Sam
But hang on, hang on, hang on, hang on. Anytime there's a generational, transformational technology or change in our world, it will be met with maybe some level of skepticism, but there's some value, and you talked about some of the value. So I'm interested in hearing some of your points. Is this something that has been overhyped, or is there some value out there that we can't lose sight of too?
- [11:57]
Jeff Nielsen
Yeah, And I'd like to remind the room and Jeremy, my opponent, that we're not actually debating who's the best comedian?
- [12:02]
Moderator Sam
Right.
- [12:04]
Jeremy Roberts
Well, that was a good line. So you might have.
- [12:06]
Jeff Nielsen
If that was what was on the ballot, you know, I might be in real trouble. But I want to talk about AI and you know, even in the stats you presented, a lot of them say Things like, well, 50% of people aren't really liking it, or 50% of people aren't getting a lot of benefit, but you know what that means? 50% are. And that is transformational. Right. For a new technology that's still in its infancy, for people to be using it and getting real time savings out of it, I think that's extremely exciting. And as I said, we're seeing more of it, not less of it. I don't see anyone who's saying, you know what? I used to use AI but I don't use it anymore. And that, to me, is extremely telling. This is a train that more people are getting on than they're getting off of. And that, to me says we're going only in one direction here in terms of it not being cheaper. That's every technology, right. Who remembers when we were all going to move to the cloud because it was going to save us money, Right? Oh, what a sad time that was. But that doesn't mean that it's not the future. It doesn't mean that it's overhyped. It means that it unlocks new capabilities. We can do things with it that we never could before. And so I don't think that really bears a lot of fruit there. And honestly, all of this comes into the question of when Jeremy makes his point. I still want to ask you, are you going to adopt AI or not? Are you going to invest in it? Are you going to learn it? Are you going to tell your boss, hey, I went to this conference and there was this hilarious guy and he said, don't worry about AI. You don't have to adopt AI. And you say, boss, I'm listening to the funny man. No AI for this organization.
- [13:39]
Jeremy Roberts
So you think I'm funny. Thank you.
- [13:42]
Moderator Sam
That's going to.
- [13:43]
Jeff Nielsen
That can go in your performance eval that. That I will actually write for me, by the way. And you'll probably use AI to read. There will be a point about it being funny, and we'll see what I think about that. But I just want to be really clear. You will all go back, and if you're not working on an AI plan, you will. And by the way, I don't know that there's anyone in this room who's saying you Know what? No. AI is not for me, because you can't. Because we all recognize that, that it is the future and we can debate exactly when it's going to completely change the world, but it's when it is not an if.
- [14:17]
Jeremy Roberts
Okay, So, I mean, I love your points, Jeff, and not just because you're my boss. There's some good points there. I will say there's a couple of things. First of all, I love your cloud example because anybody in here been in it for a long time. Remember when the cloud sort of first became a thing? Yeah, I see some hands up. I mean, if we were having this the cloud is overhyped conversation, I think we're kind of on the same side. It was overhyped when it first came out. It was expensive, it was a difficult thing to migrate to. It wasn't right for every situation, but the cloud sort of eventually found its spot. If you came in and said the cloud is going to fundamentally transform the world and everyone's going to adopt it, yeah, I would agree that it is overhyped. I'm not saying that AI is inherently bad. I'm simply saying that the amount of excitement that it's ginned up in a very particular corner of the Internet, LinkedIn, I think, is dramatically overstating the real value. And like, I've got a good example here for you. Did you guys know that Alphabet, the parent company of Google, is actually diluting its shareholders? Anybody hear about this? For the first time since it went public, Alphabet has issued new equity, $85 billion in new equity because they need to pay for these massive data centers that I use to make memes of Jeff. For the first time, they can't. And I read a JPMorgan Chase article on the AI bubble a little while back and it said, don't worry, this is nothing like the dot com bubble because it's being funded by companies that can afford it out of their free cash flow. I guess not. So, like, if Google, one of the most profitable companies in the world, is dipping into a reserve that it has never had to dip into before it, if OpenAI is at a trillion dollar, almost trillion dollar valuation, if Anthropic is at a trillion dollar valuation and people are still coming to me and saying this AI thing is really interesting, what should I use it for? That, to me is the textbook definition of being overhyped. And so I think you got to vote for me if you think that AI is pretty cool. But we're still trying to figure out how to use it. Maybe we'll summarize some service desk tickets, optimize some deployments to AWS or whatever, but I don't think it's worth. How much market cap do you think AI has added to the S&P? $510 trillion? $12 trillion? Think about that number next time you make a meme using Nano Banana, which is a silly name, by the way.
- [16:49]
Jeff Nielsen
So I want to reiterate that we're talking about the hype of the technology, not the companies making the technology. You mentioned cloud. I'll come back to that in a minute. But talking about dot com, there were a ton of companies that went bust during the dot com boom and we very likely will see that again. But does companies going bust mean that we're not moving to the Internet, that we're not moving to digital? No, of course it doesn't. It is an innovative phase. There are going to be winners and losers. We're not betting on specific horses here. The reason that these companies are spending so much money on this is because of demand. It's because of people like Jeremy who demand to make memes of me as their boss. It's because we're all using it for work, right? Because we're doing things that are agentic. But the difference to me between a technology like cloud and a technology like AI is cloud is exciting for people in this room. AI is exciting from everyone to, you know, the kids who are making their own T shirt companies to the researchers who are using it to come up with new vaccines and treatments and change lives. Everyone is using AI. Everyone. And it's only getting better and only more people are using it. So these arguments, they just don't really hold water in my books.
- [18:02]
Jeremy Roberts
So everyone is using it. That's because Google says we're changing search to be AI. We just raised $85 billion to do it. You know what that did? It drove DuckDuckGo installs up 70% the next day. Because people see this and they don't actually like it. I mean, there's sort of the usability case sometimes. That Google summary can be pretty, pretty useful. At the top of the page. We can acknowledge that AI is actually pretty good at taking big pieces of text and making them slightly smaller pieces of text. $10 trillion in market cap. But is that worth the hollowing out of the entire Internet? Because there's now no incentive for people to create stuff because Google's not sharing links anymore and they're sucking up all the advertising revenue. I'm not worried about Terminator Jeff I'm worried about Wall E. You know, that's the direction I see us going with this. Right. So if AI can truly abstract away all of the really boring, drudgery type stuff that we do, it creates an amorphous mush of content that's out there that these big companies can sort of sell back to us. And to me, that's not a future that I'm excited for. Like, who here says, oh, all novels should be written by AI? Probably nobody. You know, I'm really looking forward to AI movies, you know, probably nobody. I'm really looking forward to AI, you know, doing the work that my kids would do in their first job at a consultancy like Deloitte or something. Probably nobody. And yet it's being sold as this incredibly valuable thing. And I'll add one more point before I move on. You know, I was listening to your points and it's well researched. I don't mean that in a glib way, but like, it reads a lot like a white paper about blockchain or the metaverse from 2021. And like, the march of technology is not inevitable. Just because we get better at doing something doesn't mean we should do that thing. VR is better than it's ever been. Still sucks. You know, Blockchain was never going to change the world, but we still talked about it like it would because there's a select group of people who stand next to what they perceive to be big piles of money and yell loudly about it. So maybe before I pass it back to you, Jeff, I wrote down a one liner here. I'm really proud of this one.
- [20:11]
Jeff Nielsen
One.
- [20:11]
Jeremy Roberts
You ready?
- [20:12]
Jeff Nielsen
I'm ready. Jeremy.
- [20:13]
Jeremy Roberts
AI is solving poverty one influencer and venture capitalist at a time.
- [20:25]
Jeff Nielsen
It's actually in some ways very nice for me that I feel like, Jeremy, you're actually making my point for me, which is that you're saying, I don't like where AI is taking us. Right, because by saying that, you're actually acknowledging that it is taking us there. I'm not making the argument that it's going to be all sunshine and roses and everything will be better and we'll live in a utopia created by AI. There's lots of folks that believe that. There's lots of folks that will tell you that coincidentally, they all sell AI.
- [20:53]
Moderator Sam
Right.
- [20:54]
Jeff Nielsen
That's not the argument that I'm making. What I'm making is that the change is happening. We don't know exactly what it's going to look like. We don't need people using AI for absolutely Everything. If there's use cases for AI that people don't like, that makes complete sense to me. We're still figuring it out, but we are using it. We're using it for a ton of things. It has a lot of good uses and we're getting there. And so when you say, I'm worried about this future, you're worried about it because it is happening. It's inevitable in some way at a transformational scale. So thank you for arguing my point.
- [21:27]
Jeremy Roberts
So not bad. I gotta give you that. I will say this. I got a couple of points here. I think the real test is if we go forward and we look at the future that I sort of laid out and that Jeff has. I'm not going to say you agreed with it, but, you know, we're talking about AI having a real impact on, like the Internet and search, for example. Will people look back if that comes true and say, wow, that sure lived up to the hype. Probably not. If we, you know, create the dead Internet. Right. And you've used other examples. You mentioned healthcare. I want to come back to that. Obviously, we've heard some pretty inspiring stories here, but at scale, I mean, this has actually been quite a tough business to penetrate. Anybody remember IBM's Watson? Yeah. Leading voice in AI. They won on Jeopardy. In, like 2011. Watson was going to take over the world. It was literally the go to name in AI. And then they pivoted into healthcare. That was how they were going to bring this to market. And they did pivot into healthcare, and it went about as well as you probably think it did. And IBM CEO was asked about why Watson isn't a household name in AI. And he said, because we went into healthcare. So even the people who are building these models are actually pretty skeptical about that particular use case. So I think the test that you guys need to apply when you're voting in a few minutes, it's very important to me that you vote. One person, one vote. The test that you need to apply is, do you think that the future that I painted and that Jeff painted is something that people will look back on and say, that was hyped and it lived up to the hype. And I just don't think, given the scale of the promise, given the potential negative consequences, given the lack of adoption of key tools, Copilot's only adopted by 20 million of the 450 million Microsoft 365 subscribers. That's a slam dunk. Just go sign up for an E7, you got yourself some AI. People aren't doing it. The adoption isn't actually at that level and we're in the VC funded stage. This is Uber in 2015 when rides were cheap. In five years, it's going to be Uber. Now when rides are shockingly expensive, are people going to be using it at that level? I don't know.
- [23:34]
Jeff Nielsen
If the debate we were having was is copilot overhyped, I would be in real trouble. With apologies to anyone from Microsoft in the room, but that's not what we're talking about. We're talking about AI. And when you look into the future, I want to ask you, is there going to be less AI, Is there going to be more AI, Is there going to be substantially more AI, or is there going to be so much AI that it's almost intended, inconceivable for us to even answer that question right now. And for me, it's pretty obvious that we're leaning closest to the last option I had there. It's going to be a level that we haven't seen. It's always difficult to look back at things and say, was it worth the hype? Was it not worth the hype? But it's penetrating every single thing that we're doing. And I do want to briefly touch on the hospital piece because I had a hospital CIO up here, a CTO actually, just about two hours ago, who's already using AI to free up substantial resources from doctors. Right. This is happening right now. If that's happening in hospitals right now, imagine what we'll be doing in five years, in 10 years, in 20 years. As we said. I don't think this is a matter of if. I know. It's a matter of when. And we can't even begin to scratch the surface of what is going to happen with this. It's like when we first got electricity. Being able to predict the Internet, we have no idea. But this is going to be a utility. This is going to be a foundation that we build on. We're already starting to see it. And you're hearing people all over the place say we can't even predict more than a year or two out because we don't know where this is going. That's the only future I can see at this point.
- [25:05]
Moderator Sam
Wow. Well, yeah, I want to make sure we leave a little bit of time. We're getting down to the last, really seconds here in terms of our time on stage, but can we get the poll back up? We're going to ask our audience, our attendees vote.
- [25:28]
Jeremy Roberts
Agree.
- [25:28]
Moderator Sam
We heard from two experts two recognized world class experts share their insights and the question to you is AI Overhyped? Agree or disagree?
- [25:43]
Jeremy Roberts
I'm not liking that. Down. Thank you everyone for voting. We have I voted stickers outside.
- [25:56]
Moderator Sam
All right, in 15 seconds or less your response to seeing the results.
- [26:04]
Jeremy Roberts
Well, it seems like most folks agree that AI is overhyped. And I will just say that doesn't mean that it's not good or useful in some contexts. But it does mean mean that you can't blindly apply AI. It's not like Flex Seal that you just slap on everything that's giving you a problem. You need to have a problem to solve. Solve it with AI if AI is the right tool.
- [26:23]
Jeff Nielsen
I just want to confess that we decided who would take each side with a coin flip before we did this. So as passionate as we were, this is performative for everybody. And I think we both fall somewhere in between.
- [26:35]
Moderator Sam
Awesome. All right, ladies and gentlemen, put your hands together. Let's give Jeff Nielsen and Jerry Roberts a round of applause. Great job.
- [26:45]
Jeff Nielsen
Congratulations.
- [26:46]
Moderator Sam
Well done.
- [26:50]
Infotech Announcer
But Digital Disruption is developed by Infotech Research Group, a leading advisor to technology leaders around the world. If that's not you, you don't need to care. So skip ahead and enjoy our content. But if you are a technology leader in Infotech helps IT teams get projects done faster, better, and at a lower cost. Infotech provides unlimited access to practical tools and expert guidance you need to execute at a fraction of the cost of traditional consulting, no matter the project. From AI strategy to cybersecurity to vendor negotiation, Infotech has you covered. Check it out at the link below. And don't forget to like and subscribe.
- [27:29]
Moderator Sam
Sam.