透过 AI 镜头看旅行：与 Booking.com CEO Glenn Fogel 对谈

There is no such thing as a moat. There is no such thing as somewhere you’re going to be protected against innovation. Today, we have a competitive advantage in areas, absolutely, but those can go away tomorrow.

The only way to win in the long term is to continue to develop new services and new ways to do things. How can we do your business better? What do you need? Where do you need more demand? Where are you hurting? It’s so much more complex.

If you think you’re just going to come in and do this business and make these very big claims, I think you should really understand what this business is before you decide to commit your capital.

So, Glenn, thank you so much for joining us today.

Well, thank you very much for having me.

You’ve had a really interesting career. You went to Wharton, you went to Harvard Law, and along the way you worked in, I think, the MIS program at Morgan Stanley. I’d love to hear a little about your early background and how that led you eventually to Booking.com and then Booking Holdings.

Sure, I can do that. I came out of Wharton with a degree in finance, and I ended up in the back office in MIS. My first job was putting tapes on a drive in a data center for IBM 3084s. That’s where I started. I was an operator, actually, operating a mainframe—very different. There was nothing in college that prepared you for that.

Yeah.

Then I became a developer, but I basically learned that this was not a career for me and that I should do something else. All the people I knew in undergrad and at Wharton were going off to investment banking and making all this money, and I thought I’d try that.

But you can’t go back once you’re going down one chute. You can’t just jump out of that one to become an investment banker. So I said, “I’ve got to be better at finance.” Harvard let me into the law school, so I thought I’d do that because that was another route.

I did that, and I ended up getting a job on Wall Street at a bank. I did that until 1995. Then the bank was bought by another bank, and they fired almost all the bankers, including me. But not everybody—because if they fired everybody, they’d say, “They fired everybody.” They didn’t fire everybody. There were actually some people who were picked to stay, but I was not one of them.

That was pretty bad. It was a really good lesson, though, having been fired and knowing what it’s like. That’s something I’ve kept with me throughout my career: how to do it right and how to do it wrong, and understanding what goes through the other person’s mind when you tell them, “I’m sorry, but there’s no longer room for you here. It’s no longer the place for you to be.”

I really learned that firsthand, being on that side of the table. So now I’m unemployed, and my father had just died not long before this happened. I lost my job, my grandmother died, and the dog died. It was just sad.

I was in my early 30s, a lawyer or something, or whatever, and I thought, “What do I want to do with my life now?” I said, “You know, I always wanted to write a book.” I started writing a book, and I got it done. Now I was going to try to get it published. It wasn’t really self-publishing, where you have to work on your own; I was trying to get an agent to pick it up.

I was introduced to a woman on a blind date through a friend. She was a lawyer, but I didn’t know. They said, “She used to work at Random House as an editor.” I said, “I’m very interested.” We had a date, and the book never got published. But I did end up marrying her, and we have 2 kids. It’s a wonderful life.

While I was trying to get an agent interested in the book, she eventually said, “You know, if this relationship is going to go forward, you should get a job.” I said, “Damn, what should I do?” I didn’t want to go back to banking.

A friend of mine from law school was a senior person at Morgan Stanley. I told him, “Amy says I have to get a job. Any thoughts?” He said, “We have this trading position here that you can do.” I said, “I’ve never traded anything in my life.” He said, “Don’t worry. You’ll be fine.” I said, “Okay.”

I ended up being head trader for a guy named Barton Biggs, kind of a Wall Street legend. I did that for a number of years, but I just didn’t like it. It wasn’t that exciting, at least not to me.

That was when the internet was really taking off—the first real explosion of the internet boom. That was the late ’90s. In 1999, I started trying to interview. I thought, “I’ve got some skills. I had been an IT person, and I know a little bit about corporate development because I was a banker.”

The only real company on the East Coast at the time with internet capabilities was Priceline. They had a job in corporate development, and I thought, “Perfect.” I got an offer, but I said I wanted to wait until I received my bonus for 1999, which was paid at the end of February 2000.

I got my bonus check and was ready to start, and that’s when, of course, the NASDAQ peaked. I had gone long on the internet a week before the NASDAQ peaked, and the stock collapsed. It proved that I shouldn’t be a trader, having just bet absolutely the wrong way.

Priceline had its difficulties after that. Our market cap went from where we were when we went public to $30 billion within a week or so. Back then, that was real money; it meant something. By the time I joined—a few months after the IPO—we were probably down about $15 billion. This was February or March of 2000.

Wow.

Within 9 months, our market cap was down to just a couple hundred million dollars. Our stock was trading at $1 a share. We were going to get delisted because the stock had gone below $1 a share, but we stuck with it. We did a reverse split to make sure we didn’t get delisted, so it went to $6 a share that year.

I stayed, and I’ll have been there 27 years—I’m in my 27th year.

Wow.

We went from that reverse-split $6 to, last summer, coming very close to $6,000.

Wow.

Over that little more than a quarter-century, it went up 1,000 times. The market cap was peaking around $180 billion. Remember, it was a few hundred million dollars.

Yeah, that’s amazing.

It’s been a good ride, but of course, you never stop. Every day is a new adventure. Every day, you’ve got to fight for a customer. I was quoted in the Financial Times after an interview with them saying, “You’ve got to fight for a customer every day.”

How do you think about the lessons from that internet era in terms of the current AI wave? We’re seeing this massive shift in market capitalization.

You think this is an explosion of new things, of all these new companies coming—everything just like in the late ’90s. The optimism about technology was that everything was going to be wonderful, but then you get a little bit of the backlash coming in. We’re getting that backlash now, too.

I guess it’s bigger than it was then. The numbers are much bigger, the issues at hand are much bigger, and the pluses and minuses are much bigger. So I do see a lot of parallels.

Yeah, because when I look at it right now, some companies clearly have enormous revenue bases. OpenAI and Anthropic are rumored to have $30 billion to $50 billion in revenue run rate each. In parallel, you see companies that are extremely highly valued—$10 billion and so on—that don’t necessarily even have revenue yet.

If you look at the internet era, I think it was something like 450 companies went public in ’99, and 450 went public in the first few months of 2000. Maybe 500 went public before that. So you had 1,500 companies, of which—what?—2 dozen are left at most. The other 1,480 are gone.

Those were IPOs, by the way. Those were the very strongest, or perceived to be the strongest, companies, right?

Right. I wouldn’t even want to guess what the ratio of successes to failures will be this time around versus that time, or versus any other time when there was an incredible boom.

Going back over the last 150 years, there have always been these kinds of speculative booms that create tremendous innovation, bringing in both money and people. I mean, California—the 49ers. Everybody was running off to the hills for gold, and it was going to rain gold. I’m sure there were a lot of companies selling axes and hammers.

Yeah, yeah. The Detroit auto boom—same thing.

Exactly. And how many of them survived? I don’t know, but this is not new. There will be a great deal of disappointment, and a lot of people are going to lose a lot of money. That’s just the nature of how our economy works when there are spectacular bubbles.

That will bring all of them down, but that doesn't mean that there aren't a lot of companies that are actually of real value and are going to—

Sarah Guo

How do you think, as a founder running a company or as a CEO, you should make the decision in terms of whether to keep going or whether to sell? It's kind of like, okay, Priceline is worth a couple hundred million dollars, and the decision was made: we'll keep going no matter what, and there may or may not have been options in terms of exits. I have no idea. But in today's era, there are quite a few options in terms of exits. Should people mainly be thinking about exiting right now, do you think? Do you think they should keep going?

Yeah, I think that's right. I don't think we can give a rule or general rule without knowing what the facts of that specific situation are. By the way, there were times earlier in the day when Priceline would have been happy if somebody had made an offer.

Sarah Guo

Yeah, yeah, yeah.

It really depends a lot on what the situation is, how confident management and the people who have put the money in that business are that there's going to be a future, and how concerned you are about it. What are you really trying to do? Are you trying to accomplish something? Is your goal to actually make something that matters, or are you just here to make money? There's nothing wrong with that. I'm not against that. You just have to understand what your motivation is and what you're trying to achieve.

On average—well, I don't know if you're an American male or healthy—but I always say a 78-year expected lifespan when you're born. Of course, the longer you live, the higher the expected lifespan. Whatever it is, how are you going to spend those years? What is important to you? What do you want to do? What's the meaning to it? I'll let the people actually involved in those situations decide. I would not give them any general advice.

Sarah Guo

That makes sense. Back to Booking, one of the categories that you all are obviously really crucial to is travel. There are a number of next-generation AI companies that have started experimenting with things like this. OpenAI had checkout in ChatGPT, and one of the use cases was travel, and then they canceled that feature. I think at the time, Booking went up 8% on the news.

What do you think didn't work there? How do you think people should think about travel through the lens of AI?

So, I think we should back up a little bit so we understand what's going on here. In any type of situation, you'll have people who are not that knowledgeable about an industry or about how things actually happen in the business. From the outside, it looked rather easy: “Oh, this is easy. AI will take care of travel, and all the travel companies won't be worthwhile, won't be worth anything at all.”

That was why companies like ourselves took a big hit as some of the new models were released that had a much better way of doing agentic commerce, as it was perceived at the time. And then, when people made an analysis—“Oh, we're not going to do that”—like when OpenAI said, “We're not planning to be a merchant of record. We're not even going to keep this in-app type of way of doing commerce. We're not doing that,” then people said, “Oh, well, I was wrong. I'm not going to worry about it as much.” That's the other side.

The truth is, the way we look at AI is as an incredibly beneficial tool and a way for us to be able to do our mission easier, cheaper, and better for our customers. In all businesses, what is the purpose of a business? A business is to do something of value to its customers.

We have 2 kinds of customers: travelers and partners. We are in the middle of that. We're a marketplace. How can we do it better? AI, particularly AI using large language models and other things like that, can help make it a much more valuable method for travelers to get the information they need, do what they want to do, and be beneficial to our partners. That's it.

Now, the idea of ChatGPT, you no longer having one method—I wouldn't read too much into that one way or the other.

Sarah Guo

Yeah, that makes sense. It's interesting because I'm in the middle of Silicon Valley, where people are very AGI-pilled, right? People strongly believe that AI will drive all sorts of things. In some cases it will, in some cases it'll take longer, and in some cases it won't.

It reminds me a little bit of crypto, where crypto was going to solve everything, and it didn't, but it was very important for certain aspects of the financial system. I think stablecoins and other things are increasingly valuable there.

On the AI side, what a lot of people are really moving toward is more agentic work. That could be specific companies like Decagon having agents to do customer support, but it's also the larger platforms like OpenAI, Anthropic, Google, et cetera, providing increasingly self-driven systems—Codex, Claude Cowork, or some of the things Gemini is doing today.

One of the arguments people are making is that the nature of UI is going to change, and you're going to have agents doing transactions on your behalf, sourcing things like trips, figuring out your travel itinerary for you, or buying or purchasing the actual different aspects of travel. A, do you think that's a correct vision of the world? And B, do you think—or how do you view that interacting with Booking and what you all provide as a service?

All right. So, again, we want to reduce this to understanding what the customer wants. Many people would travel to find a great restaurant. I know that. Who wouldn't travel to find a great restaurant?

Trying to put together a complicated trip with family, let's say, multiple destinations, and different things you want to do—it's complex and it's a pain. You start planning and then you stop because it's just too much of a pain. Everybody would like somebody else to do it. Many people would like somebody else to look for them.

In fact, that's why you'll find very wealthy people have travel concierges—people who are actually human beings who really understand the needs of that customer, what they really like, and have them do a lot of the hard work for them. People aren't quite that wealthy, so they'll have their partner or their spouse trying to do it for them.

I'll be perfectly honest, so I'm exposing myself here, but I'll say it, okay? My wife and I sometimes argue, “Okay, who's going to have to do all the travel planning for this trip?” Because it can be frustrating, et cetera.

Now, with AI, the beauty is it's going to make it so much easier, and it's doing it right now. We are doing it right now. Let's use that generic term and call it an agent. I can't wait until we—Booking Holdings and our companies—are offering these personalized agents that know everything about you, everything you want, and are able to do so much more for you than any human travel agent could ever do.

A machine never forgets anything. The machine has an infinite number of permutations, and it rapidly looks through and chooses what is the best thing. It can go down and then back up: that doesn't work. Why? This doesn't fit that one. And it can come back with—

Now, people will always want some agency, so they can make the decision themselves or at least confirm they want it. Most people, for the most part, when it's a complicated thing, don't want to double-check the ticket, the flight, or whatever. That's different from, say, a businessperson who says, “I've got to go from New York to Chicago.” You go and your human assistant does it for you. That's like an agent doing it for you. It knows what you need and all that. That's great. I'm always talking about the agency.

So, we have Booking Holdings and all our companies. We are doing that right now. In fact, you go to Penny, which is Priceline's gigantic AI assistant.

I just did it the other night. I put in a very complex need for travel with the family. It was my wife and me: we want to go up in the front of the bus. I want the young adults, who are adults but aren't paying, in the back of the bus. They're adults, so I got 2 cabins now.

One person has to go back to a different city. We're going to Europe. We're going to a city where we're not actually doing the trip from that city. We're going to land—how do they get from one to the other? Should we have the hotel where we land and then travel the next day to the other city? Or should we go that night? How are we going to do it? What restaurant? All the things.

I did it on Priceline's Penny, and it was incredible. I also had to add other things. I told it, “By the way, I got a lot of frequent-flyer miles. Should I be using my miles, or should I use cash, and for which ones?” It was just beautiful how it went back and forth and asked me questions like, “How many miles do you have with each airline?”

I gave it the information. Then we're going through the flight part: how much did it cost? By the way, it ended up as I expected. I'm using my miles for the upfront part for my wife and me, and we're paying cash for the kids on our flight. We're going to go to the hotel in the city where we landed. The next day we're going to get a shuttle. It was wonderful. Wonderful.

And that's what we want even more. Here's a real core thing: when things go wrong—and things go wrong in travel. Many times, things are nobody's fault: weather, mechanics. It happens, okay? You want to have that 1 point of contact that can fix everything, because travel is like dominoes. One thing falls over and they all start falling over.

That's the beauty of AI: being able to figure out, being able to look ahead, what can we do? My goal is to have a system that we're actually able to use to predict well enough what the problem may be before it happens.

And suggest changing, fixing. I have so many examples of this, but I see the future coming.

Sarah Guo

I guess at a generic level, your team on a call that we had prior said that Penny adoption—which, again, is this agentic tool that you built for Priceline—has doubled every month for the past few months. And it's generated a lift in conversion, plus faster search, lower path to booking, lower cancellation, and higher customer success. So, it seems like it's working in really interesting ways.

Are there a common set of use cases that you think are most common for Penny? Are there specific things that it doesn't do well that you just need the underlying models to get better for? I'm a little bit curious about that.

Actually, we have some areas where we're going to be coming out with some new things on it, but here's something really important. When you ask me the question as plainly as you say, it's all great, but you tell me it's not really doing much here—you know, the numbers don't really show up much yet because it's still really, really small in terms of the absolute number. We need to talk scale here. You know, last year we did $186 billion worth of travel. That's a lot of travel, okay? We did over a billion room nights. So, the actual numbers aren't that large.

Part of it is that we're not pushing it really fully right now. That's one of the issues that I always want to think about: What's the cost of this? That's a way to easily understand what the cost of running it is. How many tokens are we consuming? Where are we? How many times are they coming back and forth? Tell me, how much was the cost of booking that trip for that person, and what is our ROI going to be?

The next thing is, what's the return going to be in the long term? What's the lifetime value? Do they come back? Is loyalty up? By how much, and will it hold? These are things that we don't know yet. We're going to have to work on and develop them until we know.

And, by the way, the whole thing of token economics now—which model should we be using for which purpose and when? Obviously, you can get tokens a lot cheaper for certain parts, or different models may or may not be cheaper. That's something that we have to look at very closely, too.

It is fascinating that we can do things that I'm so thrilled we do. For example, customer service right now: When we're using AI for customer service, it's great. It happens much faster. Instead of having to staff with humans at peak times, somebody has to wait for somebody to pick up. We've all been in that line and that queue, waiting for somebody to pick up when you need it.

But now, with AI, the computer can pick it up. It's not a problem, and it can solve the problem even better or faster than you ever will have to in the future, when you finally talk to a human and the human says, “I'm sorry. You'll have to be on hold again while I get somebody else who can solve that problem.” By then, you want to throttle somebody. AI will solve that problem.

But here's the question again about that: Sometimes, people want to talk to humans. You have to balance that, because what you don't want to do is end up saying, “Yeah, you can do it all with AI,” when that's not actually what the customer wants. In the end, it's always what's best for the customer.

Sarah Guo

Yeah, that makes sense. And I think you said on your Q1 call that customer-service costs are already down about 10%—

Well, we—

Sarah Guo

—for reservations and the booking experience, about 10%.

Well, let's just go—let's go to this thing: Our cost per customer-service contact is down.

Sarah Guo

Good.

That's great. Customer satisfaction is up. That's even better. But we have to make sure that we're able to always recognize that some customers want a human being, and some customers are happy as could be, straight up.

Sarah Guo

Yeah, that makes sense. I think you also mentioned that you're investing something like $550 million of cost savings into AI and the platform on that same vertical. Where are you investing it, or where are you putting the brunt of that, both capital and effort?

The amount we're investing is actually higher. We talked about approximately $700 million being invested this year, but it's in many, many different areas. Developing more AI—there's definitely money going into that, call it tech enablement—but there are different projects and different areas. I would not put that all into, “Oh, you're investing in technology and AI.” That's not correct. There are lots of areas that are being invested in, and we talked about that on the call.

What's important is the idea that you've got savings, money, and cash flow. How much should you be putting and reinvesting in your company? How much should you be looking at, perhaps, acquisitions, and how much should you be handing back to shareholders? That's always a balance, trying to figure out what's the right ratio.

The first thing is, do we believe investing in the company is going to give a positive ROI that's sufficient to justify doing that? After that, are there acquisitions? If you can't do either of those, then get the money back to the shareholders, because they can then invest it better than you can. That's what I've always believed in.

Sarah Guo

Mm-hmm, that makes a lot of sense. And I think you folks did something like a record $4 billion in Q1 in buybacks and other sorts of returns to investors.

I really am very proud of the fact that, over the last, let's say, dozen years or so, we've bought back approximately 40% of the outstanding shares. That's good. And we offer a nice dividend.

In the fourth quarter, we bought back $3.6 billion worth of stock. We gave out approximately $300 million in dividends. In addition, we were also paying the taxes for the equity grants that vested during the year. We do it by withholding the shares. That's part of the vesting. That's another $300 million or so.

A lot of money is going back to the shareholders if we don't think that we can use it properly ourselves. From my investment-banking background, maybe back then or maybe in trading, I remember companies that just built up huge amounts of cash and weren't doing anything with it or giving it back to the shareholders. I'm like, that doesn't seem like the right thing.

Sarah Guo

I think one last thing that you mentioned earlier that I thought was really important is just the scale of your business. I think it's at an enormous scale, and that's kind of underappreciated as an asset.

As an example, you closed 2025 with what I believe is 8.6 million alternative-accommodation listings. That's people listing homes or rooms or other things for rental; it could be a variety of different types of spots. But that creates a really interesting, I think, durable asset.

I feel like people sometimes overstate how AI is going to transform certain businesses. It's obviously going to transform everything, but there are also things that are very hard to build and that are very durable in the long run. If you have a user that's going to go and book an alternative accommodation with you because you have all the listings, right? You have the marketplace built.

Are there other aspects of your business that you view as especially durable going into this era?

On that one, it is a good point. I think you're right, but I think it's underappreciated by some people—probably, I'd say, Americans. In the alternative-accommodation area, obviously, a big player is Airbnb. So, when we invested in it very early, I mean, congratulations to you both on your part.

A lot of people don't recognize that globally, when you look at our number of listings and you look at Airbnb, it's not that different. Even more so, when you look at our total amount of transactions worldwide, you see that we are approximately three-quarters the size of Airbnb. And that's just our alternative-accommodation area; the much bigger hotel business is on top of that.

Over the last 5 years, we've grown faster than Airbnb in the alternative-accommodation area. On a like-for-like basis, we've grown faster over the last 5 years. It's a great product, a great thing, and it's going very well.

Your question, though, is whether that gives us an advantage, so to speak, against somebody who comes in and is creating a nice, AI-intensive type of system and hopes they're going to get the connectivity to these players. I'll say there is no such thing as a moat. There is no such thing as something where you're going to be protected against innovation.

That's what I'm trying to get across to the team. We've got 25,000 employees, and I try to get this across to everybody: Every day, we've got to be fighting. Today, we have a competitive advantage in areas, absolutely, but those can go away tomorrow.

The only way to win long-term is to continually develop new services and new ways to do things, and come up with new, gigantic travel efficiencies. I want this universality that will make it so much better. That's the only way.

Working on the other side, by the way, is also very important: with the partners, where, as you said, we're helping them. Again, a lot of people don't understand the complexity involved. It's not just getting the inventory and loading it into some database.

Anybody can do that. That's nothing. We have thousands of people dealing with hotels and other property managers. How can we do your business better? What do you need? Where do you need more demand? Where are you hurting? What can we do to make your system better?

It's so much more complex than I believe many other people who look at this industry from afar understand. Secondly, something a lot of people really don't understand is the regulatory framework around the world. Dealing with travel is very highly regulated. It's not a bank, okay, got it. It's not an airplane or an airline, for that matter, and it's not, you know, drug discovery, but it is very regulated.

It's complex, and if you want to be a merchant of record in travel, you have to adhere to a whole bunch of rules. Around the world, much more than in the U.S., those regulations are increasing. I don't know if they're necessarily increasing exponentially, but let's say they're increasing. That, too, is something that, if you're big and at scale, you can afford to deal with.

If you think you're just going to come in, do this business, and knock away these very big players, I think you should really understand what the business is before you decide to commit your capital.

Sarah Guo

So, I guess if you reflect on life, or you reflect on things looking forward—because you've had an incredible run, right? And the run is by no means over. You still have so much stuff you're working on and doing.

It's not like we're just at the start. No, seriously, this is the most exciting time ever, ever, because of the ability to build these new things. I tell everybody that.

Sarah Guo

Yeah, I agree. I think this is a transformative moment in terms of this technology in the world, in society, and everything else. It's so exciting to be in the middle of all this. Obviously, you all are playing a really prominent role in one aspect of that, or a key aspect.

You joined Priceline, to your point, when it was in the hundreds of millions. The stock is now a $130 billion-plus company. It was $180 billion earlier in the year. I'm sure it'll go back there over time, given all the things we're working on. Fingers crossed.

If we do what we're supposed to do.

Sarah Guo

How do you think about what you hope to accomplish more broadly in life? What is the right measure of a person, of an outcome, or of the next few years? I'm curious because we chatted very briefly earlier, and I felt like you're somebody who's thought deeply about more than just, “How will I drive bookings for us?” Although obviously you think about that quite a bit. What is the right measure in general that you're measuring yourself against, or when you think about everything over the next couple of years?

I am very blessed. I've been very lucky in my life. I'm in a position where I can pretty much do what I'd like to do. Somebody could ask me, “Why don't you continue doing what you're doing?” I say, “Because I think I'm part of this and doing something good.”

We're not curing cancer. I know that. But I think travel is a very important thing for a lot of people. It really adds to their lives. Our life mission is to make it easier for everybody to experience the world, and I believe that does improve everybody. It improves the world by getting people to travel more and experience other cultures, other people, and so forth.

If we can do it right and make it easier, that's great. I want to be part of that. I do believe that's adding something. Everybody needs to understand, I believe, why you're doing what you're doing. You only get one life. You get one life.

Some people don't have choices at all. That's the only job they have. They do it to be able to support their family. I get that. I believe that. I know that. But for people who have a little bit of ability to choose, I think they should choose wisely. Choose wisely, because you will not get that time back.

Some people have different ideas about what they believe their life should be. Whatever it is they want to pursue, that's great. My biggest fear, too, is people who take paths that, in the end, leave them middle-aged or later and a little bit wistful about it. They think, “Gee, what did I do? I wonder if I had done that. I'd probably be realizing now…”

I hope not too many of my law school classmates feel this way, but I fear that too many of them chose to go to law school because that was just a path. Then they became lawyers because that's normally what you do when you come out of law school. I didn't, but many people did. Then it paid really well and was in their comfort zone, and later in life they're thinking, “Gee, what did I do?” That's what I think everybody should really think hard about and make sure they choose wisely.

Sarah Guo

One thing we've talked about quite a bit is the impact of AI on jobs. There's this claim of a jobs apocalypse and all these other things going on. I'd love to hear your views on that and how you think about it.

It's really interesting. If we take it in the general sense of technology and job replacement, that's something that's happened forever. We look at the way agrarian societies moved more toward urbanization as technology advanced, the Industrial Revolution, and so forth. We can go through anything like that, and we know that happened.

The issue that's really interesting, though, is the speed of the change. If you look at where we are right now, I've seen it happen throughout my time at this company. At the beginning of the company, when we acquired Booking.com, they were doing hotel reservations in over 40 languages. All the content, all the scripts, everything was in 40 languages, and there was customer service in 40 languages.

There were a lot of people involved in that because, at the time, all the translations were being done by human beings. There was no machine translation at all. Now we have machine translation. All those jobs are gone. Nobody has to go and do that anymore. Those jobs just disappeared.

What happened to those people? Where did they go? What jobs are really taking their place? What are we doing, in fact? That's an example where we see this happening in real time.

Now we have the issue not only of AI, but also the fear of, “Will I get a job coming out of university?” All of the jobs seem to be gone. What was necessary was an analyst in the financial department of a bank or a corporation, and all those jobs are now basically being done through AI. The jobs aren't there. What's going on? How's it going to affect us?

We know, on the other side, that new jobs are going to be created. We all know that, too, just by looking at this problem: The speed of job disappearance and new-job creation are probably not happening at the same rate. The second thing is, what about the people who are not able to make that change?

I think about a 50-something-year-old truck driver. That person was making a very nice living and felt very good about themselves because they were driving an 18-wheeler across the U.S. and felt responsible for helping contribute. It's a good job. Now, all of a sudden, that's completely automated.

Sarah Guo

Mm-hmm.

They're out of a job. How are we going to retrain that person? What is that person going to do? That person is going to feel very bad. We've seen in society how these types of large dislocations have caused problems in the past.

I'm concerned that not enough thought is being given to how we're going to deal with these changes if they happen too quickly. Society as a whole has always benefited from technology and the creation of new possibilities to do more things, but we have to learn how we're going to deal with the flip side—the costs that come with it.

Sarah Guo

Do you have a specific guideline or proposal for what we should be doing there?

I'll tell you one thing that we do here at our company. We're always trying to upskill people. I'd say every day I'm talking with my team about how we can do the best training. How can we get people ready for the future? How do we get them to become AI-literate? That's probably a phrase we've got to learn, along with how to use AI to do AI.

That's really important because even if we end up not being able to replace, retrain, or put someone in another role, at least they're better skilled for a job somewhere else. I feel a real obligation for that. That's our point. I think everybody should be thinking that way, too.

It's good for our company. It's positive in our lives for somebody to be able to use new tools in a better way and be more productive. It's great. It also helps them with their career.

Sometimes I can see somebody, in the short term, saying, “I don't want to spend the money for that.” That's not the right way to think about it. We could have governments coming in with certain types of programs and trying to come up with ways to help, but retraining by governments over the last 50 years really hasn't worked out so well.

So, I'm not sure that's the right way to go, either. But I do believe this is something that I would really like to have more conversation about—how to do it—because I am concerned that if we end up in a situation where people start rejecting technology because of fear, that will end up being bad for us as a society. And by the way, the parts of the world that are not going to have that problem—they will be disadvantaged, I assure you—but not in China. They are not having that same feeling of, “Oh, AI is bad, and we shouldn't do it.” That is not what's happening there. So, I really think we've got to talk honestly and openly so we have the right conversation and do not end up on the bad side of people coming out and rejecting what actually is going to be good for society.

Sarah Guo

That makes sense. It's interesting, too, because I know at least one group that has rerun consumer surveys on AI because there's this claim right now that people are very negative on AI. It turns out it depends on what question you ask. If you ask people, “Do you love using ChatGPT, Gemini, and Bard and all that?” they're like, “We love it, and we'll pay more for it, and it's wonderful, and it's helping our lives in all these different ways. It's helping our kids with school,” or whatever it is.

Then, if you ask them, “Are you worried that AI will come and destroy your life, take your job, and ruin your career?” of course people are like, “I don't like it.” So, I think the questions are being asked a certain way on purpose, and I think, to your point, we need to be level-headed and say, “Okay, what's the real implication for different areas of the economy? How do we make sure that people benefit overall? And how do we make sure that people can participate?” That's very different from taking a pure doomer view or taking a pure negative view on what's coming. I appreciate your perspective on that.

You're so right about it. Part of the problem is that in a democracy—which we have, and which I'm in favor of—you'll have people who are saying certain things, not because that's what they believe, but because they believe it will get them a vote. That's also problematic. I favor democracy, but I'll be in favor of people being a little more honest about what they're saying.

Sarah Guo

Yeah, yeah, yeah. And what's their overall purpose? Yeah, 100%. Thank you so much for joining today. I really appreciate you sharing your various views across all these topics. It's been really great chatting with you.

Well, thank you, and congratulations on all the things that you've accomplished. It's pretty impressive.