Jason Calacanis and Brad Gerstner on AI in Travel, Work, and Daily Life — Full Video

Jason Calacanis and Brad Gerstner are looking forward to the days that generative AI becomes reliable and has a real effect on how the world operates. And they believe that could be the case in some areas in as soon as a year.

Jason Calacanis, a tech entrepreneur and investor, is the author of the book “Angel” and hosts two podcasts: This Week in Startups and All-In. Brad Gerstner, is the founder and CEO of Altimeter, a tech investment firm based in Silicon Valley. He says he has participated in more than 100 IPOs. They spoke with Skift CEO Rafat Ali onstage last week at the 2023 Skift Global Forum in New York City.

Watch the full video of the discussion below.

Interview Transcript

Ali: You are in for a treat because nobody’s left in this room and we’re just going to shoot the breeze. We’re going to talk about all kinds of stuff.

Calacanis: Let’s do it.

Ali: The reason I was so excited for both of them, I emailed them, I don’t know, four months ago together and my email was, “It would be my dream for both of you to come to the Global Forum.” And literally within five minutes he said yes. And he said, “Jason, what are you going to do? You want to come or not come?” And he said, “Jason, of course I’m going to come.” The back history is that I’ve known Jason for 25 years now. I was one of his employees.

Gerstner: I’m sorry.

Calacanis: Yeah, very sorry about that.

Calacanis: We had a good time.

Ali: In the Jason 1.0 phase.

Calacanis: Silicon Valley reporter.

Ali: Silicon Valley reporter phase.

Calacanis: You were how old? 23?

Ali: 24, 25, yeah.

Calacanis: And I was maybe two or three years older than you. I was probably 27 or 28.

Ali: And this a story for another time. But Jason literally saved my life, which is a story that at some point I will tell the world.

Calacanis: Maybe for your autobiography in 20 or 30 years.

Ali: I’m sitting here alive because of him. So it’s actually a true story. But Jason is also an investor in Skift, which has nothing to do with why he’s here, but he just happens to be a huge supporter of Skift as well.

Calacanis: It’s amazing to see what you’ve built, by the way. I mean, I read it and it’s just amazing.

Ali: Appreciate it. And also, he’s at this point, probably one of the most famous podcasters in the world. How many of you know the All-In podcast?

Calacanis: Holy cow. Wow, that’s incredible.

Ali: So most of the people here are in the room for Gerstner, and there’s groupies that really want to meet Jason.

Gerstner: That makes sense.

Calacanis: But we’re also best, we’re besties. We do a lot of business together and travel together. So we have a lot of adventures that we can talk about.

Ali: Literally, I got so many emails, “Can you please introduce me to Jason?” And so there’s going to be a line of people after that really want to meet you.

Calacanis: Sounds great.

Gerstner: Serious.

Calacanis: Hopefully some founders of companies.

Gerstner: Let’s get to talking. 

Ali: And Brad, let me just also set up for Brad. So in 2011, I think maybe at the first Phocuswright conference or something, Brad is the star and everybody’s trying to go to him and I couldn’t reach him. We hadn’t started a Skift and nobody gave a shit. And we started Skift in 2012, and I’m going to say three, four, five years later, Brad, I ran into him or maybe he came to our conference or something happened and he said something to the effect that, “Rafat, you built a great thing.”

And so I immediately called Jason, my co-founder, and Dennis and said, “I think we’ve arrived, because Brad said that we’ve done something good.” So anyway, this is a context as a fanboy to start the Q&A. Both of you have a very spectral view of the world, as I’m sure all of you who listen, and — I watch on YouTube, the All-In podcast — they talk about all kinds of topics. Brad, from your perspective, from a macro perspective, where we’re sitting, everybody’s trying to figure out is the boom phase for travel over while the world obviously has slowed down? It seems like every CEO says we’re building capacity, we’re starting new routes, we’re starting new hotels. That continues to happen. From your macro perspective, you’re in Altimeter Capital which invests in a lot of these companies. Give us a sense.

Gerstner: What I would say is the human desire to travel, I’m just going to show you, look at this slide. That’s this morning, planes in the air arriving and departing from New York. The human desire to move, to have experiences, is only increasing. What stands in the way of that is of course human productivity, what we have to do to survive, and our capacity to access it. And so I think if I go back to 1999 when I got involved in online travel, Rich Barton said to me, “We’re going to take the green screen, we’re going to turn it around and we’re going to give power to the people.” We’re going to let them for the first time see the flights, and the times, and the prices and it just removed friction to the experience.

The secular trend around travel is steep as ever, but the cyclical trend around what we’ve all gone through, we just survived a global pandemic. Those cycles don’t go away. And so the cycle that we’re in right now, the Fed has radically increased interest rates, 8% mortgages, 10% credit cards, or 20% credit cards, 10% car loans. Student loan debt now has to be repaid. Consumers have burned through their stimulus savings, we talk about a lot. And so if you look at our own travel demand forecast, travel demand has come back, but now it’s flattening out and it’s starting to turn down a bit again. So we’ll see where this goes. But the revenge travel coming out of Covid, what it proved to us is humans hate lockdowns. We want to be here, we want to come to New York. I mean Uber’s on fire in New York. All these businesses are doing incredibly well, but make no mistake about it, consumers are under economic pressure. That was the intention of the Fed.

Ali: In terms of how, and you’ve obviously have been on boards and understand the OTA world very well. How should OTAs think about the next coming phase of travel? You and I talked a little about that.

Gerstner: I’m ping ponging this back to Jay.

Calacanis: Well, no you answer that question. I’m interested in hearing your answer.

Gerstner: Listen, and I’ll tee this up maybe in a way that you and I can riff on this because we basically live parallel lives in internet 1.0, 2.0 and we’re thinking a lot about where the world’s going now. But Larry Page once said to me about Google that, he’s like, “This isn’t where we want to be. This is just where we are.” He’s like, “Effectively what we created was the world’s largest card catalog.” Right? Ten blue links is like pulling out an endless drawer of documents that allow us then to go get one step closer to whatever we’re trying to do. But what are we really trying to do? We’re trying to figure out the perfect hotel room, the perfect flight, the perfect car, the perfect sunset view, the perfect restaurant. And Google’s pretty unfulfilling in that regard, but it was a hell of a lot better than the world before we had access to anything.

And so it felt automagical. But here we are 20 years later and we still experience the world in roughly the same way. And as much as I love all the OTAs, they’re effectively search engines in the underbelly of Google. So you have a search engine that’s a card catalog that leads you to another card catalog, and a page filled with advertising that’s really tough. And I really look forward to the experience when I just say to my general purpose AI who already knows everything about me, knows where I like to stay in New York, knows the restaurant, knows the view, knows the table knows — I just say, “Hey, next Thursday, book me the standard.” And it happens and it’s automagical.

Ali: It’s not there yet though, today, but it will.

Calacanis: What do you think, Jason?

Ali: So Jason, the history of Jason also is you’re early in everything, you’re early in newsletters.

Calacanis: Email newsletters. We did it in ’96.

Ali: You were early in zines in the ‘90s.

Calacanis: Yeah, in zines.

Ali: You were early in web blogs. In blogs.

Calacanis: Blogs. Yep.

Ali: You were early in podcasting as well.

Calacanis: Yeah, one of the first.

Ali: You were early in Uber as well.

Calacanis: Third investor. Yeah. It’s a very simple technique. You mentioned five things that I got to early, but you didn’t mention in between each of those were 10 things I tried that didn’t work. So what happens is, and this is-

Ali: But you’re still early even in the things that didn’t work.

Calacanis: Yeah.

Ali: My question is what’s the value of being early in the AI world?

Calacanis: If you’re early, and early in our industry as technologists is somewhere between six and 12 months. I think people really overestimate what early is, and then they underestimate what can be accomplished in decades. How many people used ChatGPT today? Raise your hand if you used ChatGPT today. Okay, that’s phenomenal. It’s like a third or a half of the audience. Pretty amazing. We’re in essentially year zero, the first year of consumers really using AI, not a developer writing AI, that’s existed for decades, but consumers using it. And yeah, and it’s cluegy and it doesn’t work exactly, but it’s going to be the most disruptive trend of our lifetimes. It will be more disruptive than the internet. It’ll be more disruptive than mobile, certainly cloud. And it’ll be even more disruptive of the PC revolution.

If you put all of those together, those were the prequels to what’s happening right now. The main event. And the main event is going to be massive job destruction, massive efficiency, and then solving the world’s most complex problems. And it does sound crazy when I say it, but I’m seeing it happen right now. The companies we invest in are…

Ali: And you invest in like 150 companies a year.

Calacanis: So I have an investment firm, Launch. We have 19 full-time people. We have 20,000 people apply for funding a year. We meet with 3,000 of them, and then we invest in 150 of them. And we deploy tens of millions of dollars a year doing this and every 50 investments or so historically, maybe every 40 to 60, somewhere in there, I’ll hit a unicorn and it pays for all the mistakes. And so it’s a really fun job if you have the stomach for it.

But what’s happening right now that I see across all the companies is entire job categories are being made 30% more efficient, 20% more efficient, 50% more efficient in year one. And I think in year two it speeds up. And so what’s getting hit right now? Event planners, I think travel agents, which were hit before, but some of that planning kind of stuff or planning an event like this, research, what’s called an SDR, a sales development rep. A lot of repeatable stuff. And there’s a really interesting way to predict what will happen, right? Because you said, I’m early to stuff, I hang out with a lot of smart people and I take notes. And Sarah Tavel from Benchmark, she had written a blog post of how she was looking at business process outsourcing in India, what jobs were being sent to India to be done there to get the arbitrage of labor.

And so anything that’s being business-process-outsourced will be done by AI. And in fact, the business process outsourcing companies are using AI, and then pretending that somebody spent 50 hours this week on it when somebody really spent five and they just charge you for 50. So we will see entire categories of work go away, whether it’s people driving cars. Just like we saw, whenever labor has its moment and fights for more, they have to be very careful to not fight for too much. Because then what I’ve seen is a lot of organizations will start adapting technology faster. So the perfect example that was here in New York, anybody remember about 15 years ago fast food workers went on strike in New York, they were upset. And now you go to McDonald’s and do you order from a cashier or do you order from a kiosk?

Ali: From a kiosk.

Calacanis: Exactly. And so that technology I was investing in at that time, and voila, that job went away. Now you have a bunch of auto work and I’m very pro-labor and people should fight for their best salary, but this time will be different. This change is more profound and it’s happening at a hundred times the speed. So take that one example of cashiers, boom. And then the person making the french fries, person making the burger, the accountant for that McDonald’s, all of that is just going to get abstracted away and done by AI. And we’ll have one person doing the work of 10, just like we saw in agriculture.

Ali: And so travel industry, and you know this very well, is one of the world’s largest employers of people. People move into middle class as a result of travel. The frontline workers in travel are probably the largest creators of middle class. That’s how they move into it. As you’re thinking about investing as well as looking over the world from a tech perspective, how do you reconcile? Where would the jobs come from and how do we retrain here in U.S.?

Gerstner: Well, I mean there are a couple really big embedded questions there. Is this good or bad for the travel companies? Is this good or bad for the online travel companies? And then what does this mean for the social contract in America? And I think those are all going to be the profound questions. The last one in particular that we’re going to struggle with for the next decade. I totally agree with Jason, this will be the single largest labor displacement in the history of capitalism. What we saw happen to blue collar in places like Tesla. You walk into a Tesla factory and there aren’t a lot of humans in it. 

Calacanis: And he’s working on Optimus. I don’t know if you saw the Optimus Tesla robot?

Gerstner: And so to Jason’s point, engineers are now 20, 30, 40% more productive. Meta CAGRed their engineering headcount by 40% for the five years up to 2022. And now Mark Zuckerberg is saying they’re going to CAGR it at 1 to 2%, right? 

Ali: So explain that for people who want to understand.

Gerstner: It just means the company’s still growing really fast, but you’re going to hire a lot less engineers. They’re expensive, they want a lot of stock-based compensation and other things. But because each engineer is now 1x, 2x, 10x better as a result of copilots like GitHub, et cetera, you don’t need as many to accomplish the same task. But the thing that’s most interesting for me in this room is to think about how that flows through all of these things. And the place I focus most on is in online travel. And as a consumer, what I’ve always wanted is that automagical experience that I explained earlier. And I think there is an open question right now. Does the unique data and talent and skills of Booking.com, and Expedia, and Kayak, of TripAdvisor, et cetera, does that entitle them to an advantage in this new thing?

Or as Clayton Christensen and others would talk about the classic innovator’s dilemma. Does your economic business model and the talents and treasure you have inhibit your ability to reinvent yourself? So I would argue, and I did, I have on the pod, I think that Google, Google right now is a monopoly with monopoly profits. So whatever the thing is in the future that the world’s going to become, I would argue it’s unlikely that it will be as good for them as it is today. It’s very difficult for monopolies to survive disruption. See AT&T.

And so when we get to the other side, I think even if Google executes incredibly well, there’s going to be more competition. Half the room raised their hand on ChatGPT. A year ago if he had asked that question, 100% of you would’ve used Google for that exact same task. To me, my criticism of Sundar was he lost the verb. He didn’t say, “How many of you have used Bard?” He didn’t say, “How many of you have used Cohere, or Anthropic, or Claude?” He said, “How many of you have used ChatGPT?” That is the verb today for AI, and that’s extraordinarily valuable. So I think this new thing is sufficiently disruptive. I thought that mobile was going to disrupt search. It kind of did, but didn’t really.

Calacanis: I mean, what mobile did was it increased the amount of time we all spend online because all those periods of time between when we were at computers became available. And so there is going to be a really amazing opportunity in travel, which is there are other trends that have occurred while AI is happening. So while this massive efficiency happens, that means because technology is deflationary, people’s costs will go down. And then you have generations change where Gen X’ers, we wanted to be successful. We bought into trying to be in charge and working hard. And then millennials got a little softer. And then you have these Gen Zs who are like, “You know what? It’s all bullshit. I want to retire early. I want to have experiences.” And what is the UAE, the auto workers, what are they fighting for? A four-day work week? That’s the fight. What is the fight today for American workers? “I don’t want to go to an office.”

This is where we are, people. We’re fighting to work less, and to not commute, and to live more. Who benefits from that? Y’all benefit from that. People are going to be taking longer vacations. Americans are going to start looking like Europeans. The concept of an American working in August will go away. We will all be taking August off. Congratulations. You can let your boss know when you get back to work tomorrow that you get August off. But that’s what’s happening. And I see it with young people. Any young person you would offer this deal to, here’s your salary, it’s going to be $100,000. Or you can make 60 or 80 for working three or four days a week. Guaranteed nobody takes five days a week, guaranteed for anybody under 30. They all would want more freedom, more flexibility. This is why Uber has 6 million drivers compared to the 1 or 2 million people who work at McDonald’s or Starbucks.

Having watched that movie, the press, which is filled with a lot of dishonest people — I don’t want to sound like Trump here — but they have a massive agenda and they really hate technology because of what we did to their business. We destroyed it. Google, Facebook took all the ads. Craigslist took all the listings. They’re kind of bitter about it. And if you just look at how the world is changing generationally, it’s going to be extraordinary how much more free time people have. And I’ve watched the trend as a New Yorker living in California. Everything starts in California, then it goes to New York, makes its way to London, and eventually-

Ali: You have changed, Jason. You’re saying it starts in California?

Calacanis: Well, I mean, nut milks, yoga, four-day work weeks, working at home. This is all crazy loony shit that started in California that when I lived here in the ‘90s, if you asked for nut milk, they would be like, “Are you fucking crazy?” Or if you took your clothes off and started smoking fentanyl in the street, people would be calling 911. You’d get tackled. And then I live in the Bay Area. You go to San Francisco, people are drinking… The default at Blue Bottle is a nut milk. If you ask for a latte, they give you oat milk. Not milk.

Ali: Not milk.

Calacanis: Has that happened here yet?

Ali: No.

Gerstner: Thank God.

Calacanis: See, that’s where you got to draw the line. And so I do think there’s this massive generational change.

Ali: So are you optimistic about the young people?

Calacanis: Oh, I’m super optimistic about the world right now because I think as scary as the word disruption is, humans have always found something else to do. The fact that we now have a population of people who make a living TikTokking, Instagramming, podcasting.

Ali: You just interviewed Mr. Beast at your conference.

Calacanis: Mr. Beast. Yeah. I mean, back to the Uber example, what people missed about Uber was it gave people control of their destiny. The idea of going to a 10-hour shift, or working three three-hour shifts when you wanted to, if you made half the amount of money that was twice as valuable to a person who wanted to drop their daughter off at school, pick them up, have lunch with them and take them to soccer practice. And so that’s the trend in the world, and I think it’s going to be more free time and it’s going to be amazing for travel. People… Also, the nomad thing is very real. I watch a lot of sub-Reddits.

Gerstner: You’re like a filibuster. You need to get down on the question. 

Calacanis: But anyway, sub-Reddits are a great place to look for this. There’s two sub-Reddits that are really great. One of them is retiring early. There’s an entire movement of people who want to make a shit ton of money and then retire by 40. Then there’s another group that are overworked, over employed, and what they do is they’re developers and et cetera.

Ali: They’re doing multiple jobs at the same time?

Calacanis: Yeah, they say they quit Google and then they go work at Twitter or wherever, but they don’t actually quit.

Gerstner: Google doesn’t know the difference.

Calacanis: They trade strategies of how to set their schedule up and have two laptops and be on two Zoom calls at the same time. “Oh, my internet went out, sorry.” And then they answer the other one.

Ali: Wow.

Calacanis: Where are you going to go now?

Ali: So here’s where I’m going to go.

Gerstner: I’m about ready to tag him as the world’s greatest moderator if you don’t pick it up.

Ali: Well, he’s not already trademarked it. Second, I’m fine being the second to him. You talk a lot about social inequality. So both of you I think talked to Ray Dalio ten days ago maybe. Was it? And he opened the talk at your All-In Summit saying that the fall of civilizations happens when wealth inequality becomes the highest. You have a point of view on it?

Gerstner: Yeah. Well, Jason just described a world that’s a little post work. If you read Alice Huxley in Brave New World, that was the world they described. A world where labor and capital would become so productive that our productive use would no longer be required and it would start 80-hour weeks to survive, and then those would become 60, and then 40, and then 30. And I do think we’re rapidly approaching a future where the social contract is going to have to get renegotiated because the consequence of the industrial revolution, particularly post-World War II and the consequence of the information age has been massive wealth concentration that Ray Dalio correctly points out. He does this — I encourage you to go on YouTube or watch the All-In Summit pod with him. Like the fall of almost every civilization started with massive wealth inequality and led to internal revolution in the country.

And we have that occurring in spades in the United States right now. Lots of discussions about what to do about it. People have talked about universal basic income, which I think is a terrible idea because I think it deprives people of a principal source of purpose, which is work. But an idea that we’ve talked about a bunch on the pod Invest America, hopefully we’ll have bipartisan legislation introduced this fall to create a private investment account for every child born in America, seeded with $1,000 from the federal government. Think of this as a 401K from birth. So now we have 3.7 million children born every year in America. It’s less than one 10th of 1% of the federal budget to create this universal infrastructure of ownership. Compounding is the eighth wonder of the world. And a child born 30 years ago who had $5,000 in their first two years of birth, it’d be worth $287,000 today.

It’s hard to hate America. It’s hard to feel like you’re left out of the system, right? 70% of people in this country don’t own anything, don’t have investment accounts, don’t have savings accounts. I would argue like Ray Dalio and I think all of us on stage, that’s untenable. And it’s about ready, we just told you all the people are going to get displaced, about ready to get a lot worse. So I think as a country, we need to get ahead of it. One of the things that I’m excited about that comes out this podcast is we not only talk about all of these things, but what are the actions that we’re taking? They’ve been interviewing the presidential candidates, which I think is a great service, the actions that we’re going to take in order to bend the curve on this. But that’s one we’re working on. There’re going to be lots of others that are needed.

Ali: Let me talk a little bit about the podcast. So for people who are obviously fans of the podcast, they want to know this question, what clicked? Why did it catch fire?

Calacanis: What do you think it is? 

Ali: I just happen to know both of you, and I happen to know Chamath a little bit. And it’s the chemistry between all four of you. Super successful people who can talk about the world intelligently in a broad sweep of things. And your moderation. It started at a time during Covid. 

Gerstner: Somebody just put Brad for president. 

Calacanis: Keep going, keep going. No, I think you nailed it.

Ali: And I know you, your views on current media, and you won’t say this, but I will say this, that you are still at heart a journalist.

Calacanis: I mean I do random acts of journalism now. When you and I came up we were taught journalism is you collect the facts, you check the facts, you present the facts, and then y’all make your own decision. And then over time, Fox News and Rupert Murdoch had this really great idea, “Hey, split the audience, pick a side and they’ll give you more money.” And that really worked. And then the left-ish side and the moderate side were like, “Yeah, that’s terrible.” And then Trump got elected and people were like, “Well, that’s an existential crisis. I can’t handle this. We have to pick the other side.” And so then the New York Times just said, “Subscribe to the New York Times and the Washington Post so that we can investigate Trump.” Literally that was their advertising and that’s when you knew it was over. I think Trump is a horrible human being. I think he’s a criminal. I think he’s an existential threat. He’s a horrible human being. But I still think the journalists should have played it down the middle and not been advocates for one side or the other. And it’s advocacy journalism today, right? And I think something’s broken. But when things break, this is one of the great things about living in a free capitalistic society is that you get a chance to rebuild them.

How many people here have listened to a podcast in the last 30 days? Raise your hand nice and high. If you listened to any podcast, it’s everybody, right? How many people think that Fox and MSNBC have picked a side? Okay, so we’re all in agreement here. How many people want journalists to pick a side? Raise your hand if you want journalists to pick a side. Nobody here wants that. But let’s be honest about journalists. You and I grew up with them. These are elitist intelligent people and they at some point decided that it was their job to save society or whatever. And there’s a huge incentive. Now they’ve learned the worst lesson, which is the Rupert Murdoch lesson. When you pick a side, you make more money.

And the New York Times has learned it. Oh, subscriptions. So it’s like the best business model and Washington Post learned it. And journalism’s a disaster. But what that means is now that we’re 30 years into the open internet and it’s really easy to create, and all those tools were made over the last 20 years in broadband, and now you have a phone that is basically a CNN studio in your pocket. You have a studio in your pocket. It’s crazy. We could just do a podcast from out here and the whole world can watch it. It’s nuts when you think about that just 20 years after we were trying to make broadband work. 25 years, maybe. What that means is the people who were in the stories, the subjects of the stories are now going direct to their customers or direct to the audience. And what you correctly point out about, I think the All-In podcast is it’s four people who are in the arena, so to speak. It’s four people who are capital allocators and company builders, and we’re all 50 years oldish. So you got whatever that is, 30 times whatever. And then Brad’s the fifth best so you add him over a hundred years of people actually building businesses. And there’s a very interesting thing happens when you become, call it what it is, post money. You’re not motivated by money, let’s say. Or that’s not your primary motivation and you’re not worried about running out of it is there’s like if you drew an X, Y axis, there’s a certain tipping point where you just don’t give a fuck. You can’t cancel me. I’ve already made the money. I was telling my daughter, we like to talk about woke stuff and I was like-

Ali: You have three daughters?

Calacanis: I have three daughters, but the 13-year old is amazing, just keeps talking about me getting canceled and I’m like, “I’m going to get canceled today so that I can ski with you more.” And she goes, “Okay, go ahead and do it.” And I said, “Okay, trans people aren’t good.” And I start typing on my phone and she’s like, “Don’t do it dad.” And I’m like, “They’re great.”

And then we just play this game of how to get canceled and stuff like that. So when they tried to cancel Chamath for saying, “Listen, I don’t care about the Wiegers, I care about my kids, they’re children of color, I care about what’s happening in America. I care about my kids getting pulled over by a cop and getting shot or that is what’s important to me, not what’s happening around the world. I think you should focus here.” They tried to cancel him.

The White House wrote a note about it. The Warriors wrote a note about it, “Hey, we don’t agree with Chamath,” kind of thing. And then this whole very lame cancel culture and the Overton window being shut and nobody can discuss everything. Even when I brought up trans people for a second, everyone’s like, “Oh, shit.” But why can’t we talk about that? We can talk about any subject. And so that’s, I think what we’re getting back to is we can talk about any subject, we can solve any problem. You don’t need to be scared. People are generally good is what I’ve learned in my years on the planet. People generally have good intentions and a good debate is good society.

Ali: And you guys have a lot of ideological differences among all of you.

Calacanis: Of course.

Ali: And then what I hear in the podcast all the time is you’re going to go then play poker after.

Calacanis: Yeah, it’s crazy. But you could be friends with people that are different than you.

Ali: How does that work?

Calacanis: I don’t know. I mean I grew up and it was just, people were different than each other. And I remember when I was a kid watching like, “Okay, this person likes Bill Clinton, this person likes George Bush, this person likes Reagan. This person likes whoever was running against Reagan.”

Gerstner: Jay Cal said this before. So our Thursdays, we have dinner at 6:00. We show up, we play a little poker, have dinner, have a lot of discussion, a lot of vulnerability, and we have a continuous argument throughout the week on text in our group chat, and then we play poker after. And Jason said, our love language is arguing. It really is a brotherhood. We trust one another. We bring our best to the table. Like our analysis, our arguments. We really are committed to excellence because we’re all deeply passionate analysts and anthropologists about the world. That really is the glue that puts it together. But then when we get around the poker table, we want to mentally abuse everybody around the table and take all of their money. And that is just the competition among friends.

Ali: Who wins most of the time?

Calacanis: That’s a good question. Phil Hellmuth, world’s greatest poker player. Yeah, he’s the most consistently winning player in the game. For some reason we play with the world’s greatest poker player and we’re all amateurs, but we’re good amateurs. It kind of swings around. Nobody really wins or loses too much. It’s a friendly game in that way. And if you were there trying to win all the time, like you might do in a casino where if you’re up 10K, you would just leave immediately. If you were to do that, you just wouldn’t be invited back.

Gerstner: It says zero, but I think we should go another 10 or 15.

Gerstner: There’s a lot of filibustering. We got shit to cover. Let’s go.

Ali: The unit rule. So let’s go to the Q&A. This actually is more a question for Brad. AI seems to take out intermediaries. Will AI usher a new wave of book direct with the brand? If so, how do brands prepare for this windfall?

Gerstner: Oh good. I’ve got a slide for this. Ultimately, I think when you think about all the plumbing, all the APIs, whether it’s metasearch or whether it was OTAs or whatever, ultimately it’s what we’re trying to do is just solve the consumer impulse to get me this incredible thing that I’m looking for as easy as possible. We want convenience, we want price, we want it to anticipate and solve our problem for us. WhatsApp is a dominant platform for payments and communication in India. So this is a product that they launched last week, which is in India, a lot of merchant transactions just occur organically on WhatsApp. You buy your food and it’s just people chatting and they use the payment platform or they’ll go off platform to do payments. But you have a billion people who are doing this in India. Well now they have inline direct bookings on Air India as an example. So just go ahead and choose it. And if you look at this, I would say, “Oh God, that kind of still feels Web 1.0, 2.0 just on a phone.” But why would Zuckerberg be doing this? Well, today they launched or announced their ChatGPT. The Meta agent that’s going to live within WhatsApp. You’re not going to type anything in here. You’re just going to say to your Meta AI, “Book me the Air India flight next Tuesday at 9 a.m.,” and it’s going to do it. It has all your payment information, has a direct API to Air India, has everything it needs to know. And so we’ve been talking about this for 20 years. When mobile came, desktop search volumes went negative in 2012. We’re like, “Oh my god, this is the end of Google.” But it wasn’t the end of Google. Why? Because they just pushed down all the organic advertisement, filled the page with a bunch of junk advertising and we had nowhere else to go. There was no ChatGPT. Now there is. Now there are alternatives.

Calacanis: I think Google’s going to actually do great. I think they’re going to execute well. We have a little bit of a difference about this.

Gerstner: No, big.

Calacanis: And the reason I think that’s going to happen is they just integrated Google Flights, Gmail, Docs into the Google search. Anyway, it’s a little complicated. I actually told Sundar this week, I was talking to Sundar and I told him, it’s too hard to turn it on. You have to go to your Gmail account, you can’t use your Google Docs. Anyway, long story short, it doesn’t work perfectly, but you’ll be able to say things like, “I want to leave from the East Coast and go to Dubai. What’s my best city to leave from if I want a business class ticket?” Because we’ve all played that game like, “Well, if I went to DC and left from DC, do I save two grand? I could see my mom when I’m there or whatever.” And you’ll be able to ask very interesting questions. “Hey, I want to go surfing, or I want to go snowboarding. I want to do it in deep powder. What are my choices and can I get a business class ticket there and I want to do stuff for under $10,000.” What you would describe to a travel agent. Those people, remember those people?

Ali: Travel agent. They exist. They exist. They exist.

Calacanis: Those people, they would come to you with an envelope with a bunch of printed cards. Those people. That’s coming back. That’s what the interface is going to be and it’s going to work. Right now, it doesn’t work. It’s cluegy, it breaks, it hallucinates, yada, yada. 12 months from now, I guarantee you you’ll be able to say, “Book me three Michelin star restaurants when I’m in Hong Kong and book me a flight business class. I want a sleeper seat with the best wifi and I want to go for a week and give me, I don’t care what days I leave. So figure that out.” In one year that will be done for you. That entire agenda will be done for you off three sites, Open Table, Google Flights and whatever.

Ali: Do you think voice becomes the interface or chat?

Gerstner: Chat’s going to.

Calacanis: If you’re on the subway, I suggest chatting because you’ll look crazy, but if you’re at home the whole, okay Siri, thing and hey, Alexa is actually going to work next year.

Gerstner: No, but that is in my mind, whether it is voice or typing it — the interface is a dialogue. This is not one interaction. The first interaction is, book me this next week. Then your agent, your fiduciary, your friend, your personal… Use a great personal assistant as a metaphor, if everybody had the world’s best personal assistant, you don’t run a search with your personal assistant, you have an interaction.

And it becomes more efficient over time, but every single thing requires some refinement. That’s why the metaphor, why the interface has to fundamentally, whether it’s voice or you’re typing it has to be a chat type interface. That’s why Zuckerberg understands that with WhatsApp, you have two billion daily active users who are already chatting. You know what the… Lex Friedman, the great podcaster and I, we were having an argument one Saturday morning about this very topic on WhatsApp and we went back and forth for an hour and he said, “Oh, I don’t think it’s going to be chat.” I said, “Lex, look at the knowledge exchange that we just had for an hour via text. That doesn’t occur in a search exchange.” And so again, I stipulate Google has all the primitives. I just think it’s going from a monopolist with monopoly power, to something that’s going to have to compete with all these others that also have primitives that are going to be very good. And either way, the consumer wins.

Calacanis: Interesting thing about the chat is reinforcement learning, which is very underappreciated right now because it’s not implemented in most AI experiences. Most AI experiences don’t take into account— If you’re on YouTube, you don’t tell YouTube like, “I’m not interested in that video.” There is a way to do it. You hit the three buttons, you say, I’m not interested. Nobody does it. What’s happening now is you’ll be like, “Hey, play me a song I like.” And then you’re like, “I don’t like that song.” And it’ll be like, okay, and it’s going to learn something. And just like when it gives you your restaurant reviews, “Oh, these are the three Michelin star ones.” “French? No thank you. Can I get some Japanese in there?” Then it’s like, okay, next time it’s going to start with Japanese, which you do see in Yelp now or if you open Uber Eats, it’s starting to know what you like, very slowly.

Gerstner: Memory.

Calacanis: Yeah. And that dialogue is reinforcement learning, being completely underappreciated right now. Anybody here use autopilot on a Tesla? Raise your hand if you use autopilot. Okay, there’s like maybe 20 people in the room. Now when you disengage your autopilot, you disengage it because you want to drive yourself or it gets disengaged because something happened. You were not paying attention. It asks you what just happened, and then you hold the steering wheel volume button and you say, what happened? That goes to somebody who then watches the video of you and there’s some line that wasn’t painted properly and it didn’t go in the right lane. Because that’s what happened to me and I keep explaining to it, “Hey, this lane is not painted properly.” And then it all of a sudden a week later knows that edge case and fixes that edge case.

Gerstner: I saw somebody ask a question here I thought was an interesting one. Somebody said, “Is the government going to prevent use of AI in order to protect jobs?” We’ve had a bunch of conversations most recently as last week with Congressman Jay Obernolte that’s overseeing the house policy on AI regulations. It’s very clear that the U.S. has good balance. The great power struggle in the world is not going to be fought with aircraft carriers anymore. It’s really on the field of AI. We’re not going to ankle AI in this country. And what we say and what I’ve heard Jason say, and I totally agree with everybody in this room, all my analysts are using Claude every single day more than they use Google to turn themselves into 2 or 3x analysts.

I remember when Google came out, but some people are like, “Oh, I’m not going to use that. I have my way of doing it.” This tool is going to be super empowered and the people who become very knowledgeable about it, whether they are designers and media, it’s going to displace a lot of designers who resist it, and the designers that embrace it and figure out how to help people do it will be the winners.

Calacanis: So for every job that will go away in customer service, that’s a person who can teach people how to kiteboard, or take them on a tour, or cook them a meal. And so humans are very clever. Everybody keeps saying this industrial thing that occurs is going to kill a bunch of jobs. It will retire jobs, but it won’t kill human productivity or the desire to be useful in the world. Airbnb is the perfect example. There are a large number of people whose job it is to host Airbnbs, and did we see all the hotels go away because of Airbnb? No. We saw people take more days off, more travel off, take longer vacations, take vacations to places where maybe they couldn’t afford it or new hubs become new destinations. And so for the travel company I would invest in, it’s Uber and Airbnb. 

Gerstner: I mean I would say one, there’s a question, name one travel company you’d invest in today. I think you have to redefine what we think of as travel, like sleeper companies. The sleeper companies in e-commerce and I would argue experiences are Instagram and TikTok. They are generating massive demand in the world today, but they’re doing it in a very long tail. I mean, you mentioned Mr. Beast, right? A solopreneur that literally has built a couple hundred million dollar empire. Around a personality, around experiences, et cetera. But it is in every category. There’s travel category, beauty category.

Calacanis: It’s a good question by Edward. “Please address how smaller firms across verticals with no data science teams will be able to take advantage of data and AI.” The tools are going to allow you to do that. There is a data interpreter in ChatGPT4 if you pay for it and you can just upload any CSV file, any data file that’s floating around your organization, say, “Tell me about this.” And it all of a sudden just starts telling you about it and then you ask it a follow-up question, what trends do you see in the data? And so it’s going to start figuring out just random stuff for you. The concept of a data scientist being a distinct job in the world would be like saying that you have a typing pool in your company. Remember typing pool? Anybody here remember the typing pool? Anybody remember the mail room? Anybody go to an office? There was a photocopy room. It’s where they made copies of the paper. 

Ali: That’s what I did for you for a long time.

Calacanis: Typing was like a pool. The New York Times reporters used to drop off their stories on legal pads to be typed. Lawyers used to go down to the typing pool. That’s the data scientist equivalent. You’re all going to be data scientists. We are all going to be able to talk to this AI assistant and say, “Tell me which neighborhoods are going to be the next places where people want to put Airbnbs or which neighborhoods in Tokyo don’t have Airbnbs but should.” And it’s going to be like, “Well, okay, there’s a couple of great restaurants here in a great park, but there’s not a lot of Airbnbs, would seem like a good place to put one. It’s right above a cafe. In fact, I think this building would be the best building.” It’s going to be that smart.

Gerstner: Let me make two plugs for my good friend Jay Cal.

Calacanis: Oh, here we go.

Gerstner: I like to bust his chops a lot. Mondays on This Week In Startups — great AI episodes where they actually try a bunch of products. You can watch it and see the trial. Second is Jason, like he said, takes 20,000 applications a year, invests in 150 companies a year, puts on Founder University.

Calacanis: You have an idea and you’re building it: [email protected]. We’re seeing a lot of people who want to build stuff in travel, and it almost universally has something to do with the itinerary or the experience on the ground. And so I think people are working backwards from what is the feeling, not the place? What is the feeling, not the place? I want to feel culture. I want to, in Brad’s case, dress up and dance until sunrise to tribal dance music.

Gerstner: Burning Man reference.

Calacanis: Burning Man reference, so he’s going to want to do that.

Ali: Not this year.

Gerstner: I went this year.

Ali: You went this year?

Gerstner: Yes I did. It was a great year.

Calacanis: Burning Man is the perfect example of experiential. People want to feel something and they’re going to have more time to feel something. Travel is the greatest vertical to be in right now. You just have to be super creative and think about the feeling that people are trying to have and work backwards from the feeling and most importantly, the Instagram photo.

Gerstner: Yeah.

Ali: Okay. Travel is the greatest vertical to be in. Let’s end here because Brian is going to kick me out. We are over time. Thank you very much for being here.

Calacanis: Thank you for having us.

Gerstner: Thanks for having us.

Calacanis: Give it up for Rafat. I mean, amazing.