Election Security, Skift's Rafat Ali, and Anthropic Unleashes PC AI | E2030

27:43If they were to steal a ballot and then steal somebody's ballot and submit it, what would the potential penalty be? It depends entirely on that. That's going to be illegal under both state and federal law. So it's going to depend on whether you're prosecuted by a local DA or a federal prosecutor. It could be a misdemeanor. It could be a felony. But, you know, if your argument is going to be that, well, if it's against the law, obviously people aren't going to do it. Well, if that were true, our prisons and jails across the country would be empty. Well, motivation does matter. I mean, and the penalty does matter.

28:22So that's what I'm trying to get at is, you know, if the value of stealing a vote is obvious, right? Like my candidate has a greater chance of winning or somebody might pay you for your vote. I guess that's a possibility as well. But, you know, the cost of that in the database has been severe. Some people have gone to jail for many years. Yeah. Some go to jail, but others don't. I'll give you an example of this. About two years ago, a former Democratic congressman in Pennsylvania, Mike Ozzie Myers, you may remember him. He was a Democratic congressman who got caught in the abs scam scandal more than 20 years ago, went to federal prison.

29:04When he got out, he became a political consultant. Well, two years ago, he was indicted by the feds for bribery, 12 counts. Why? He was bribing poll managers in Philadelphia to stuff fraudulent ballots into the ballot box at the end of the day after the polls had closed on behalf of his clients. And he went to jail for this. On the other hand, just a couple of years ago, a U.S. attorney in North Carolina prosecuted about two dozen aliens for illegally registering and voting in elections there. Now, that's normally a felony under federal law. All of them just got slaps on the wrist. They basically got a very small fine, no jail time, and were let go.

29:57So, frankly, it depends on what state you're in, who's charging you, and how lenient or tough a judge is. Awesome. Hans, I appreciate the work you're doing to chronicle all of this. It does feel to me like we're on the road. Hopefully, this next election cycle, we can get there. 35 or 36 states require ID. So, it seems like that's some progress. We only have 14 states left. So we're 70 % of the way there and you've been able to compile these 1600. If we assume there's 10 or 100 times that feels to me like our elections are very safe and we have watchdogs like yourself and I guess the other side have their equivalent really granularly trying to find examples here.

30:47and we found a very small number of them overall. So I feel good about it and I feel great that you're doing this kind of work to obsess over it because somebody needs to. So thank you for obsessing over it. Well, thank you very much. I appreciate the time. I talk about prediction markets all the time on my podcast and prediction markets are brilliant because they allow people to buy and sell contracts on future events with prices set by market opinion. So if you're interested in this, we've got some amazing news. Calci, the world's largest prediction market, just made it legal to trade on the upcoming U.S.

31:21elections. That's right. For the first time in 100 years, U.S. citizens can now legally place trades on election outcomes. Political polls can certainly be inaccurate. We've learned that. But the markets tell you where the sharps are putting their money. Sharps are the smart people. And election outcomes matter. They impact your taxes, the economy, and of course, the world of investing. Traditionally, only large corporations and the wealthy could hedge their bets against election uncertainties. But now, Kalshi allows anyone to protect themselves if their preferred candidate loses. So go check out the real-time election odds.

31:56And if you feel confident about who's going to win, why not play some trade? I'm going to do it. So here's your call to action. I want you to check out Kalshi. That's Kalshi.com slash twist. K-A-L-S-H-I. Kalshi.com slash twist. And get in on the action. You will get an additional$20 with your first deposit of$100. That's calsheet.com slash twist. I think this is an example, Alex, here of, you know, we can have a logical discussion, even with somebody who's partisan. You know, obviously, the Heritage Foundation is as partisan as it gets. On the left, there's like extremely partisan organizations.

32:31And I actually take them in good faith that they're chronicling the database. When you look at data and you put sunlight on it, it's very hard to debate the data. And I think we saw that here in our discussion. Obviously, he's got a perspective, wouldn't say specifically, you know, that Biden lost or whatever. You know, I think that's typically how people will massage this issue like, oh, we don't have to rehash it or it could have been better. But I don't know how you feel about it. But after talking to him, I feel like both sides are so obsessed over this issue. And the cost is great for election fraud.

33:04Yeah. And there's so many people looking out for it. And we have 50 different systems. the idea that we could actually swing it one way or the other, the Russians could swing it for Trump or a bunch of immigrants could be paid a hundred bucks each to vote for Biden. It feels like the numbers don't match to me of there being systematic election fraud that would actually switch an election. Now there's fraud in every system, but I don't think there's enough here to swing any election other than maybe a very minor local one. What's your takeaway from this as you and I learn here? Because that's what we're really trying to do is like learn and get the data and then examine the data from first principles.

33:42I have a lot of thoughts. So first of all, it's always very enjoyable to watch you Jasonize someone who doesn't know you that well. I don't know what that means, but okay. Well, you don't let things go and you have an ability to ask a follow-up question when someone wasn't excited about the first question. And I appreciate that. It's a little uncomfortable, but I've done it to both sides. Oh, no, no, no, no. You're equal opportunity here. I always enjoy that. But the thing that really blew my mind, because we were talking about this last week, talking about this week, I'm sure we're going to talk about it more, more startups, more technology.

34:19It seems that even with all of the intricacies, the different systems, the patchwork quilt of our democracy, it kind of works. And I think, frankly, it's very encouraging to me. It's not like we have a system that's on fire, in the ditch, broken to all pieces. we can improve it but generally speaking seems to go okay and hell yeah that makes me happy yeah and you know we've had close elections many times al gore conceded eventually i don't trump has never conceded but we we still the democracy and and the democ the the democratic institution we have known as america seems to have even though a person did not want to concede allowed the transfer of power you know even with january 6th so you know you you can have these big debates about was january 6th the walk in the park was there an effort to overthrow the election were there people who you know were using immigrants or whatever at the end of the day uh the judicial system seems to have done its job trump filed 60 plus lawsuits he lost them sydney powell was obviously a liar and a conspiracy theorist she got i don't know if she got disbarred but she was found guilty obviously or she pled guilty yeah so you know it does feel like i too feel like we're trending towards reality here yeah i know it's hard for people when they lose whether it was al gore it was hard for him to concede or hillary claiming the russians you know caused her to lose because of the email leak i understand like but at the end of the day it's kind of hard with 150 million voters to do enough votes to swing an election and then even in these um states that are, you know, the swing states, they're under a lot of scrutiny.

35:57So I don't think it's all that easy to admit fraud. And if you do, you're going to go to jail for a long time, I think, in the national election. You're throwing in, well, I know we have Rafa and Ali waiting in the wings, so we're going to get to him in just a second here. But one thing prepping for today's show that I was less happy about is a little bit of like causing the injury and then offering the bandage. So Heritage is doing good work. I think the database is useful. I got to go through it. It was lovely to have. I appreciate learning stuff. Sure. But the Heritage Foundation is also going out and doing in-camera footage to Latino voters.

36:31This was in the Times. And trying to over-extrapolate from what people are saying to create the idea that there is rampant voter fraud by people who are brown for Democrats. And I did see that story. Yeah. Yeah. It makes me very disappointed because we could have very serious conversations about data, but sometimes orgs don't always speak with one voice. And I just think we could all do better here and we should all move away from fear mongering around election cycle, election time, because this matters a lot. Yeah, matters. Democracy matters. All right. There we have it, folks. It's a simple list of things.

37:07ID. Maybe this makes me, maybe I'm MAGA. Maybe I'm dark MAGA that I believe that people should show an ID. I don't know. Does that make me not? No. I think all we have to do is have both sides of this. Voter ID, generally speaking, very good idea. Disenfranchising voters who don't have access to it, very bad. As Han said, if we have a free ID program, if we rolled that out in every state, I would be 100 % in favor of voter ID tomorrow. So let's just solve it with$50 million, Jason. It's not a lot of money, I presume. This is not a big ask. All right. There's been a lot of talk about Uber's acquisition or the trial balloon of Uber possibly acquiring Expedia.

37:45We talked about it here. We talked about it all in. It's been talked about on CNBC every day. And Uber has got a big valuation, and they have a frisky CEO with us to talk about. It is my friend, Rafat Ali. How are you, sir? He is the... I don't know. You're the CEO of Skift.com, and I don't know if you're the editor-in-chief as well. No, I'm not. We have a very good editor-in-chief. Got it. You gave that to... Full disclosure, I'm a minor investor in Rafat's company, and we worked together on Silicon canelo reporter back in the day you got to hear our discussion i think you were in the waiting room about election integrity what did you think yeah do you really want me to go there uh the you're allowed to have an opinion yeah sure i don't think heritage uh will be coming on your show anytime soon again that's what i was getting at i said jason i think he can handle a tough question or two i think people like tough questions you know i think you're obviously very good at what you do so i mean as uh even though you don't consider yourself a journalist these days.

38:42I still consider you, one, and what you do with obviously the All in Podcasts as well as this. I want to add a point there. When Jason wanted to bring the heritage on, my first thought was, oh, shit. But, you know, I work here. I did it. Happy to help facilitate, prepare the notes. I didn't expect Jason to go that hard on that point he wanted to bring up. And I think that does underscore the value of bringing on non-traditional guests, even partisan ones, if you push them. So, five points to Jason. I was too skeptical. Well, here's the thing. Partisan people, this is what I've learned this cycle.

39:11because I didn't know how partisan some of my friends would become or are or always have been. You know, it's a range. I actually don't care about politics all that much, as both of you know. But I'm getting dragged into it because All In has become obsessed with it. Sachs is obviously obsessed with it. It's very important to him. So, you know, I get dragged into these conversations. But even partisan people want to have hard discussions. We've seen that over and over again, you know, when I asked Trump about the border and immigration or about abortion. I asked JD about certifying the election.

39:43They like answering hard questions. This idea that partisans don't want hard questions. They want it most. I think they want it most. I think Hans enjoyed me asking him about 2020 and pushing him a little bit. Did Trump win or not? He's going to hold that position. Heritage is. I think they have to maintain a relationship with Trump, obviously. And so they're not going to... They don't want the headline out there. Heritage Foundation guy says Trump didn't win in 2020. So they're always going to try to filibuster that answer just like a democrat will filibuster answers you know on their side uh and you can argue who might do it more i think one of the things uh that i'll give credit to you jason with the all in podcast or my journey of giving the other side you should listen yeah is a thing that i learned honestly really starting to listen to all podcasts and i was very one yeah the covet i was on my peloton listening my laptop was on the floor listening to you um on the first one and initially my first reaction was do i really want to listen to david tax and just patience and then seeing his general level-headedness beyond just politics we say well i have to listen to this and then so your journey i'm not sharing I feel like it's a lot of journey that a lot of us have gone through over the last four years, having an independent mind questioning a lot of stuff.

41:12Yeah, I think you have to. And you and I come from a media, all three of us come from a media background. And, you know, we see bias in media. We see, you know, insincerity and people manipulating the media from all sides. And you understand the process of finding the truth. Let's talk about the truth on a business level. you cover uh uber incessantly at skift everybody should go to skift.com it's like the new york times of travel or the wall street journal of travel like it's a really great site if you're um you know in the travel industry you read it every day but let's talk a little bit about the expedia rumor and there's some other rumors out today about it and i know you guys did a lot of pros and cons what are the pros and cons what are the chances an expedia deal happens yeah um i listened to what all of you discussed on the podcast i thought that was a very good rundown that that all of you did and the pros and cons on on the deal i think the the biggest pro is that the frequency of usage and in travel in general is low like how many times would people usually travel two three four times a year maximum this is i'm talking leisure travel and so the cost of acquisition of a customer in an online travel booking site which is a layer on top of the brands anyway.

42:28So it's one level removed from the actual brands that you may have loyalty to in the travel industry, Marriott or Delta, whatever, your airline or hotel of choices. And so for them, they spend tons and tons of money, billions and billions of dollars. It's been rising, we've been covering for the last 12 years of Skiftile. That money went from less than a billion from Expedia and booking each to now seven to eight to$10 billion that they're spending on Google. So the biggest thing from an Expedia perspective is just the cost of acquisition, the frequency usage of Uber on Uber and Uber Eats and other services that they have in other parts of the world.

43:02From an Uber perspective, Zara has been quite clear. You spoke at the conference that Zara spoke last year at our conference that I do with you both on stage separately that he likes the idea of a super app, obviously knows travel very, very well, is not just a dealmaker, but also a really good leader, understand how to create synergies and motivate people to do things profitably. So Expedia, which has struggled a bit in the last couple of years, particularly on its tech integration in the back end, which has been able to do under the previous CEO, but still languishing, the stock's languishing as well.

43:42So from an Uber perspective, if you already have 150 million users being able to upsell a much bigger ticket item, infected these mall ticket items, I think, you know, those are simplest way to put the pros and cons for this. Okay, so I want to ask a question about that, because I use Uber, I use Uber Eats more often than I should. Very much a mobile experience for me, booking flights, booking hotels, entirely desktop. And so when I think about a combination of Uber and Expedia, to me, it just doesn't resonate with how I use these services. I'm just curious, am I incredibly out of date and old fashioned?

44:15No, you're not. No, no, no. I mean, a lot of the transaction, particularly in the Western world, still happens on desktop if even if the research starts on either yeah it does jump back and forth in some of the um for instance asian countries or latin america etc mobile happens to be a much bigger part but at least in the western world that we live in it's a very mobile isn't as predominant as you would assume it to be in for instance versus our media consumption habits yeah yeah but like what i think about um a lot of companies that have an app today like united i customer i went from web to mobile but i feel like if uber bought xpedia they'd be going from mobile to web and i just can't recall a company making that kind of like backwards progression in platform so is that an actual sticking point in this deal or am i over niching down on my own personal concerns well i think uh the product part and again you guys discussed this in the other podcast uh product integration will be the what the biggest challenge there is the only thing i'll say in uber's pro bucket there is there's an overemphasis on assuming that expedia and and everything else in online travel is a huge brand that people have loyalties to most of the customers come through online acquisition channels google but the reality is if uber expedia deal happens and dara says we're just going with uber and uber travel and all of these legacy brands go away it won't kill the company it will take a while but the reality is like there's no loyalty in online travel there's just no is people don't go to hotels.com or expedia because they or kayak conversely because they love that experience necessarily they're just looking for the best price they start at google like most people do they do a search for flights from here to here and sometimes google flights intercepts it sometimes you get an ad they look at multiple sites like i think your and i behavior i think this is generally known we've done research on it that people look at tens of sites yeah before they actually book particularly if it's a once a year big trip that you're looking at right it's a considered purchase if you're doing a big trip and you're spending five to fifty thousand dollars taking your family on vacation whether it's a staycation and you're getting a motel aid or you're going to europe for the summer and you're and you're flying business or first class you're going to take time to get the most out of your and the research does pay off, I think.

46:36What do you think about the third tab in Uber being, you know, you have rides, you got food. The third one would be travel, which would be hotels and flights, or even just putting hotels, flights, experiences. I mean, it does seem like they've done some experiments there. How are the experiments done for them? Well, so I asked our this last year, and I mean, it's not big, so it's in UK where there's a company called Hopper that is also there. It's a startup that is in the online booking space. So they're tied up with them and offering hotels and then rail, et cetera. They're in India. They're a bunch of other things offering as well.

47:13And so while they're not giving numbers on conversion for first-time users, Dara did mention that the rebooking rate for people who do convert time is actually quite high. If I'm remembering it right, something like 60 % of people, once their book runs, then come back again. Interesting. so they trust it right they learn to trust it the way they've learned to trust rides and food it's the same thing with uber eats you're like is my food going to actually show up and can i risk my family missing dinner like if you order that first time you have a little bit of anxiety is it going to show up and then once it shows up you quickly become a two or three times a week person probably oh god yeah that's true ice uber eats is so expensive though jason i really gotta stop it's too easy too convenient it's too fast here's the thing if you look at your time and I know how much you get paid because I'm a part of your revenue streams.

48:04You know, if you look at people's time, if it saves an hour of prepping a meal or two hours and you save 10 hours a month or 20 hours a month, it does add up. I do think the Uber One is, I don't know if anybody here is a member of Uber One. I am. So I save a bunch on fees. How much is that becoming a phenomenon in travel? This idea of a membership that saves you money. I know I'm very loyal to both Bonvoy and United and JetBlue. I actually know my miles and I'm paying attention to that. I don't need to, but I fly so often for free with United or I get hotel rooms so often for free when I'm traveling because of Bonvoy that I kind of like it.

48:46And all my cards are United cards. And I would go probably with Uber cards if the incentive was there. so talk to me about loyalty programs and how uber is sparing with regard to that yeah so um so loyalty obviously is a big part of the travel industry airlines and hotels airlines particularly are have very large programs there's a people say that essentially airlines at this point are sort of loyalty programs slash credit cards and then airline is the loss making part uh that part is true because they're able to pre-sell the miles to some bank or credit card company and then they get advance.

49:24In fact, that's how many of the airlines, Delta, et cetera, got saved during COVID because they were able to get such big advances from like billions of dollars from these banks. Hotels is also the same thing. I mean, as you said, you are a Bonvoy user. Bonvoy is the biggest program in the hotel industry as well. There have been experiments with subscriptions the way Uber One does, but it hasn't been, and those are small scale subscription some airlines alaska air and a few others launching these um some some of the european and asian ones as well airline side launching these subscriptions some of the hotels have done experiments on for instance subscription to co-working spaces this is coming out of it um but they're small scale and then soho house which i know you know well is very much built around working as well as uh staying historically they haven't turned a profit in 27 or 28 or 30 years of uh of their existence and not a not a single quarter of um what is this for them what is amadeus i i saw that this uh firm was uh denying rumors or they explained that uber hasn't reached out to them there's some 30 billion dollar a year or 30 billion dollar market cap company that i don't even know about yeah amadeus is it is so these are called gds is the global distribution system saber amadeus travel port three of the companies in that sort of amadeus is the biggest one saber number two travel port number three in terms of global market share are the green screens that you may have seen that travel agents or if you go to the airport if you are if you peek around you will see these are the antiquated systems this was ai before ai existed this was variable pricing before variable pricing existed in the rest of the world type tech systems that are the backbone on which airlines run hotels run online travel companies run so they're very very powerful backends but in an age where tech has come in and you can directly connect to the hotels instead of using a third-party system there's there's been talk forever and enough progress that competitors are emerging or um you know ai will definitely have a big role to play in there um the there were some the rumor that i actually just heard right before this podcast was um uh was that uber approached amedas that seems unlikely because that's pure sort of b2b tech company um if it approached it the only thing that i can think of is um not to acquire but to then build their own uh expedient expedient kayak competitor hotels.com competitor because they would use that feed correct to present the information inside of their app so there's multiple ways if dara wants to include travel and keep building out the super yeah i can still oh yeah i can surely build it and i think they'll keep doing experiments in terms of what's the attach rate which is uh you know if you're booking one of these two things that they already have how many people are opting in.

52:27So I think it may be from a partnership perspective, as opposed to an acquisition perspective. There was another thing about Grab, which is the big Uber equivalent in Southeast Asia, which I think Darius is sitting on the board of, I forget what exactly the history is. Or Uber used to be an investor. I don't fully remember the history. I think Uber might have had some equity in it when maybe they exited that market, because it was kind of like the DDDO, etc. There were some local players that they gave their business to and took a percentage of yeah yeah and it really is a super app in that part of the world so um so there's some rumor that they may look at it they went public with the spec and um our friend brad gusner took it public uh with the with the spec i don't i i trading at four dollars today it hasn't done very well yeah it's still a market cap of 16 billion if that's correct.

53:21Yeah. Yeah, but according to Crunchbase data, it raised about$16 billion. So I don't know if it's done that well, but it's cheap though. I mean, Uber's worth, Jason, back me up here, $170,$180 billion, somewhere in there. I haven't checked it in a while, but yeah. The general sense is that Dara is focused on a few markets to do well, and you, Jason, are probably aware more than I am on sort of the markets that they're looking to continue to double down on versus expanding back into those big markets. Talk to me about VRBO. that to me seems like an incredible asset obviously airbnb is an incredible company they've got challenges right frequency etc but uber and airbnb were always compared as the um sharing economy uh two iconic companies that came out of that era of that era and so vrbo i used before i used airbnb vacation rental by owner um it seems like that's like an underappreciated asset that inside of Expedia, and that that would, as a marketplace where you have inventory on one side, an asset light marketplace, you got inventory on one side, and then you have buyers on the other, Uber One buyers, Uber's 150 customers with their credit cards in there already, trusting an Uber, Airbnb competitor sounds incredibly compelling to me.

54:39The conventional wisdom that Airbnb as the predominant brand is settled signs. I don't think is correct. I think there's an opening here, um, in terms of if somebody is so dominant, there's always an opening somewhere. And so one, there's that. And VRBO has been there for a long time. As you said, it was there even before I'm going to say pre web maybe. Um, but home away, you may have, some of you may remember home away, uh, used, um, so home away, bought VRBO and then home away got bought bikes, by Expedia. Um, pre-covid i forget exactly the year and then so they renamed everything verbal v-r-b-o as in verbal that's kind of how they pronounce it it did really well during the pandemic some would say it even saved expedia in many ways you can imagine the reasons airbnb and uber in summer of 20 21 22 did gangbusters business yeah people didn't want to be in a hotel risking getting covet in the hallway at the front desk at the gym and having your own home wow you don't have to worry about that so it's kind of hard to remember that craziness but that that was what it was it has come down hard because um because verbal um as people came back to the cities verbal which historically was focused on non-city areas uh which is the opposite of what airbnb was which is cities and then they added obviously the other stuff over the years has languished in terms of growth particularly this year verbal has has struggled um so there's an opportunity potentially i don't think well by dillard controls expedia so you can never count out any deal with him meaning he's the probably the best deal maker in the media tech world in history and so if you bring a good and and and dara and barry are very close it was a mentor right it was his mentor yeah he gave him the shot at expedia and so there is you know if barry and barry is up there in age right he's 80 years old so if he was looking to retire that would be a good landing place for expedia and then you have this existential threat there's always an existential threat as you alluded to even when somebody is the number one player so the existential threat as you've described it is ai helping you find the perfect hotel or place yeah i think that's gonna happen i mean i've said this for many years since maybe 2015 the tyranny of the search box as in the travel search box box that was there in 1990 pick pick a year 80 98 99 is is essentially the same travel search box that it is today which just seems mind-blowing based on the changes that have happened and so i think a lot of back-end technology in terms of of removing the friction the seamless trip that people have talked about which is how do you make sure that from airport to hotel to airline the thing that blows my mind every time is the second you step off the airline the airline has no idea where you're going has no idea what hotel you're staying at has no idea what you're going to do right in an in 2024 right how's that possible and so or even vice versa hotel does not know the second you leave the hotel are you going to the airport maybe they do but then are you um which airline you're taking what's your choices so this seamlessness of travel which somebody like uber could potentially do the examples of this in for instance china metuan and a few others uh that are trying to create super up around travel uh are there because we brought up ai rafat i want to ask about the ask a skift service you guys have put together on your company's website um you just relaunched it and it's much expanded much better including um even sec data and so forth i'm just curious how much consumer use has there been of your ai q a service yeah so uh so keep in mind this is a b2b service so not a consumer in the sense that it's not a it's not a we we're purely focused on the business to business so we launched this in may last year so we're early with chat gpt uh with GPT, I'm going to say 3.5, I think it was at that point.

58:54And so now we're in 4.0. We relaunched it a month ago from a chatbot sort of paradigm to an answer engine paradigm. So it's trained on everything that Skift has done over the last 12 years, all text on our daily news, research, our conferences, transcript. And now we add SEC filings, not all filings because then it overwhelms the system. There's a lot of garbage that comes back. So only on 10K and 10Q and even then parts of 10K and 10Q that are relevant to the industry. We cover about 140 public companies across all parts of travel. So it's quite accurate. I would say if you test it yourself, you'll see that it gives.

59:33How do you feel like as a media guy with people indexing your information? I'm sure Skift has been indexed into perplexity or cloud or open AI with or without your permission. Correct. And have you been able to prove that they've done that? And are they coming to you saying, hey, we want to make this right if we're going to use your data and pay you? Yeah. One, I take it as a compliment. The fact that Google, there was a Washington Post story last year. They got hold on the sources of data that Google's AI has been trained on. and we were like number 500 site on the web. I thought that was not bad.

1:00:19Like for a tiny small company like ours to be the 500 most useful site that they found on the web seems like a compliment. Well, you hire a hundred percent, it's a hundred percent human created with high quality editorial with people getting paid real salaries. And so when you think about the information that's available on the web, whether it's Twitter, Reddit, 4chan, message boards, like it's very variable, right? Even the reviews on Amazon are bought and sold. to the extent that people even create Chrome toolbars to try to identify the fake stuff on the web, right? So if you have an unbiased source, that's high quality, man, that's, that's super valuable.

1:00:54Yeah. Yeah. We, we have blocked GPT from crawling our site now over the last year, because they started that as an option. You can opt out of being trained. Certainly using Ask Gift, we don't send the data back. This is just the API call. Also, they actually, that's an option that you don't use your own chat, but they create on top of GPT to not train the overall chat GPT thing. I don't spend tons of time worrying about it because there's nothing I can do about as a small company. My thing is how do I use the tools to, this sounds like a cliche, but how to give value to our user. At all times, it's how do I make this better for the people that are reading us, they're paying us money, how do I get more money out of them if I continue to provide more value out of it?

1:01:43And so AskGift is an example where it's not a free, you get like one query a month, two queries a month for free. And then if you're a paid subscriber to Skift's paid services, you get unlimited access. So not that this converts people in droves to Skift, but it's a good retention to for our team, especially our sales team to say, and a lot of it today is education. Like the fact that our sales team is presenting this to people who are enterprise subscribers of Skift, they're paying anywhere from 25 to$100 ,000 a year to have access to daily news and research and other things that Skift has. And for them to know that, oh, this is available, this is our people can use it.

1:02:25So Alex, to answer your original question, we get about a couple of thousand queries a day. I mean, for a B2B product with a very specific use case, that doesn't sound bad. So put it into growth terms for us. What was that six months ago compared to today? Well, so I think we're sort of fair to say it's not been like hocistic growth there. But there was an initial bump in early days because it was novelty. and it's sort of petered off and stayed around that number between thousand and two thousand a day type range. One more question about AI costs, because I was going back to your corporate posts and when you guys talked about the initial launch of it, you were using GPT 3.5.

1:03:10I think because you said at the time that using four was just prohibitively expensive. How has the cost curve of AI models impacted how you guys build AI drive products? so again small usage in terms you know in the overall scheme of things and and and opina has also dropped the cost quite a bit since then so we're in 4.0 now it doesn't it's not significant for us we're comfortable with uh the cost we're there that's awesome i mean i'm curious to see how fast oh one can come down in price because it seems like people that are building with that set of models are getting consistent discounts it feels like the early days of aws in the best possible way when everything was getting cut and cut and cut and cut i love it yeah well one is a lot better i've i've i mean i'm sure you guys have used it too oh one preview just is so much better reasoning um it's crazy how good it is i mean you think about the average analyst that you would hire at a venture firm or you know at skift or we might have at silicon on reporter back in the day the what would take you a two days of research it will do in two minutes and then you start your research where you would have been in hour 20 maybe as an analyst in the 90s right and you obviously got to check the data and the facts it's going to hallucinate and get things wrong but man it is so great at giving you quick overviews and making you smarter i'm just amazed at the jump oh one has made yeah and you know i use it a lot in my day-to-day duties as a ceo whether it's coming up with ideas for memos that i need to write for the team or or, you know, I have to make a list.

1:04:47It's going to like a lot. And more and more, my usage is I talk into it versus typing. And I'm sure you've seen that as well. I, in my car, when I'm driving, when I drop my daughter off at school, we will put on the chat GPT. I made my action button on my phone, the voice one. And I have it now doing a series of history questions. So I told it, hey, my daughter is learning history. I want to rush up on it. start with the easiest possible questions who's the first president united states when did the revolution happen who wrote the declaration of independence and like i'm like wow i don't remember this stuff it was 40 years ago when i was 10 years old in high school whatever you know trying to learn this stuff as a publisher i hear that the green shoots of hope in the publishing business are uh subscriptions obviously you're doing that corporate and you're a b2b publication so great.

1:05:38But I also hear licensing data for models could become a viable business model. And the third, Apple News and the revenue sharing from Apple News seems to have been significant for certain people making seven, eight figures. So that sounds a little bit dangerous because people subscribe and you have to give them your entire corpus. Is that right? Are you on Apple News and how does it work? So we are, but I think this is the challenge when anybody talks about, you know, Facebook's efforts in media back in the days, Twitter and now Apple News and a bunch of others. We are a niche industry publication, so it's hard for them to come down to beyond the top players to then come to somebody like us.

1:06:24So we don't have tons of visibility into or access into these revenue programs. and beyond a certain point, I just cannot run my company based on that anyway. And so for us, it's always about direct usage with users, whether it's email or even honestly LinkedIn, which is huge for us. Really? Explain. Yeah. I mean, just in terms of everybody in our industry is on LinkedIn. Got it. This is true for a lot of industries. So Skift has, I'm going to say 500 ,000, so half million plus followers on LinkedIn. I have about half million followers on LinkedIn. So between us, about a million, and then you add all of our teams, et cetera.

1:07:11So it's quite large, substantial. And so many times when I meet people at our own conferences, they say, well, we're huge fans of Skate. We read you. I said, where do you read us? On LinkedIn. I said, oh, great. Can you please click through and come to the site? Please hit the blue link. Come on, people. But honestly, it does. and the commentary that we do i do these videos that i do and our team does as well so net net linkedin is where the action is for companies like ours versus apple news or certainly you know facebook etc and people starting to get their news on tiktok i saw you guys doing some shorts some vertical videos yeah you think there yeah i am out of the demo i'm out of the tiktok demo so uh probably the wrong person but um but i actually did not even go on instagram so just twitter was enough for me as a as a famous back in the days but um in b2b i think it's less but um but certainly we have to be where people are and if tiktok is it we'll figure out way some of our some of our employees are doing better videos than we would ourselves do which i think is great um and that's kind of where we are and back to office you have people back in that amazing office or you're still remote no we're fully remote we're almost 100 people uh we're fully remote we're 15 different countries now post-covid we gave up our office that you came in pre-covid and gave a talk to our team but we're a very different company than uh than those days certainly so we're i'm i live in new york i live in astoria if you remember yeah um in queens and And so at the size we're at, we've made it work, the virtual part.

1:08:50We certainly, the whole company comes together twice a year. And so that helps a lot. And then different teams are able to get together at least a couple of times a year. So four to five touch points per team seems to be the way to go for us. Yeah. What about mentoring? Can you mentor young analysts remote? Or is that challenging? I'm curious. It is, but I think the honest answer is surviving as an independent media company profitably from everything else. Yes. And so for us to be able to have a company that has lower cost and is a globally distributed cost base, if you will, has outweighed the cons of not being in an office.

1:09:31In other words, you can hire people in places where it's less expensive to live, dramatically less expensive than the U.S. The U.S. or East Coast, particularly where the cost, particularly in a city like New York, is very high. So I think the balance of, we still are majority, I would say, people in the U.S., they're definitely not New York anymore. um but we have people we have about 15 people in uk we have about nine people in in india philippines panama etc got it yeah and those workforces are dramatically lower cost and they really want the work so you probably say they're loyal i mean you have you have to give them marks for that like they um they are definitely uh more loyal they'll stay at the company for four or five six years where an american might be looking at it as a springboard especially as somebody who is you know let's say comes from some level of privilege in america we all do like they probably care more about their career or work-life balance yeah i don't know if there's anything wrong with that i just think that um having a mix of nationalities for a global industry like ours and having a diversified cost base um i mean we cover the globe skift is truly truly global like if you look at the skift homepage you'll have five stories from five different countries and for all i know five different continents on it so it's a very global um output that we have everybody go subscribe to skift.com and you're a rafit on twitter i think right yes r-a-f-a-t uh anything else alex or should we we let our just one one quick thing i didn't know you were an investor in skift before we started this chat and so now yeah that's fine Well, Rafi, one thing Jason's been talking about a lot is distribution dire straits, a lack of DPI from investments.

1:11:23All VCs want to see some exits. So I'm curious, when is Skift going public? When is Skift going public? Fair to say that Jason is not looking to his retirement based on his investment in Skift. Let me just put it that way. Well, get out of mode and get after it. Come on now. Here's the thing. Sometimes when you make investments, they're slow and steady. compounding 25 % growth, 50 % growth every year, doubling every two to three years in terms of revenue base. And you have that downside protection. Then other ones, people are shooting for the moon and they might want to do 10X year over year revenue, but they burn out.

1:11:58And some of those rockets wind up just blowing up before they even get off the launch pad. So you want a mix of those in your portfolio and slow and steady has won the race for Skiff. I mean, it's an amazing brand and product. Brand and product profitably. so profitably amazing yeah i mean and there's i think with this media consolidation um that has occurred you see penske bought every like back-breaking tumbling brand uh vox bought a bunch of brands shut down a bunch of brands i mean it is a chaotic period of time but it does seem like skiff the information certain sub stacks just really subscription-based b2b business b2b publications have done wonderfully and they always have right i mean yeah i mean b2b media has been there forever and and grown the only way it exists uh is probably there's not because there's not like tons of people clamoring to invest in that sector no yeah you gotta make it work yeah and these companies some of them are large like the hundreds of millions of dollars if you look at some of the largest and billion dollar companies as well in these things awesome uh we'll talk soon and i'll see you when i'm in new york uh everybody go to skift.com check it out and if you're in the industry subscribe obviously and check it out great publication okay nice talking to uh rafit and congratulations on the success that's gift um i guess there's one more story we should just hit real quick yeah yeah real quick but i i have a i have a small detour jason i've been holding this in for 10 minutes so our chat about uber and acquisitions got me thinking because we talk about uber partnering with different companies and so forth i mean uber is worth like seven united airlines you and i both use uber you and i both use united airlines come on dara why not get yourself an airline they're cheap you can buy two i mean the problem with those airlines, I think, is they are just not profitable.

1:13:44I know that Airbnb had looked at doing an airline at one point. And I think everybody has aspirations and they get frisky about airlines or even hotel chains. Virgin, obviously, did airlines and a hotel chain and buses and everything. So yeah, people get a little frisky and think about it. I do wonder if a commuter airline specifically would work. And maybe what you want to do is skate to where the puck is going and obviously where the puck is going is um vtols so i would be more likely to see you know uber have a big vtol uh footprint and be buying a lot of uh joby's vtols because i think those kind of folks just want to give the inventory over to operators yeah that would be interesting if united or delta american airlines were buying vtols but uber said you know we'll do we'll get in the veto game as well or they'll just say put your inventory into our system and get all the uber one subscribers to use your product all that's lovely and smart i'm just saying dara it turns out jet blue's only labor is 2.6 billion so come on now you can buy a dozen of them they're cheap uh i mean bonvoy marriott i wonder what mario it's like 70 billion i think jason um i looked this up for you earlier yeah 75 yeah so that would be really hard and the problem is, you know, when you're good at a marketplace and those dynamics, owning and operating things is very hard, right?

1:15:08That's a different skill set. And it's a much lower margin business. It's not a very resilient business. So once you have assets like hotels and something like COVID happens or a government terrorist attack or whatever, you know, those companies can go out of business and they frequently do get recapped or government, you know, subsidies. It's just not an asset-light, scalable business. You saw Airbnb and Uber and Lyft, DoorDash. They were all able to weather COVID quite gracefully, right? Except for ones that maybe Airbnb, when everything shut down, it wasn't graceful, but they were able to rebound.

1:15:41They just had to get rid of full-time employees at the mothership. They didn't have to have hotel rooms that were empty. No, no, not at all. Now, I did promise one last quick story just before we jumped on. Anthropic has released some new stuff, Jason, that I think is really cool. And I want to show you a very short demo. So just to give everyone a quick level set, there's a new version of Klaude 3.5 Sonnet. Klaude is the model family from Anthropic. I'm sure we all know this, but they have a new thing out called the Computer Use API. It's in beta and it allows the model to take over your computer and do stuff.

1:16:15Your browser, take over the browser window. I think it could take over more than just that. Let's go ahead and play this demo, Jason, and let me know what you think of this guy. I think this is super cool. So this is a query that they've given to the service. Essentially, it is a plain text introduction, and then the service is going to pick it up and do stuff with it. And I think this is great. Wow. Yeah. So they're giving it instructions. Hey, take a bunch of data from this database and fill out this form. And then the large language model takes over the browser and does that behavior. And so this has been something people have been doing for a long time on the web with scripting agents, scrapers do this.

1:16:55And so, you know, here you can see telling your browser, telling Claude, go find me all the flights using Google Flights, Expedia, and report back to me the pricing and do this every day or do it every hour and then tell me the price trend. By the way, that sounds like a crazy search. There are services that do that. They go straight to the web for you, check the prices, and they make a graphic. They've been limited to people who are technical or in other countries where business process outsourcing is popular. Yeah. This is now letting your language model run your browser. What could go wrong?

1:17:32Well, we've had RPA, Robotic Process Automation, for a long time. We currently have agents via companies like Sierra, Microsoft, Salesforce, but offering this up as an API for developers to use and play with more generally, more openly, and also to allow, I'm hoping, homebrew development with very modern AI models. That's so cool. I can't wait to see people dream up to use with this because I can imagine about 20 different use cases as you've already come up with one yourself. So I'm optimistic. I think it's good to show that the AI models are improving and they can do more stuff, Jason. It's not all hype.

1:18:05It's going to release your AI to do anything that can be done in a browser window. This could get a little bit crazy. You could have it, as an example, go and create a thousand accounts on social media services. and trick the catchphrase and then say, go make me a thousand versions of Alex and Jason on these social networks, then DM people that we're going to book you as a guest on the show, have conversations with those people who DM you back and then ask them to send you, you know, a thousand dollars as a booking fee to be on the podcast. And this sounds crazy, but there's somebody out there doing that for this week in startups and other podcasts.

1:18:44uh what's the guy something cynic who did the habit book somebody contacted me like oh this person simon cynic or whatever wants to have you on their podcast and i was like oh i think i know him and then i looked and i was like wait a second so i said oh just have him email me he knows me and they didn't get back to me uh and so somebody did this and the hack was um they thought they were getting booked on all in and they were in fact getting um they gave their instagram credentials over to the hacker who was going to, I guess, get these credentials so that they could stream it to this celebrity's following on Twitter.

1:19:21I'm sorry, on Instagram as well. They went in there and then they redirected their creator funds to their bank account. And I think they changed their password in time or whatever, but my Lord, you can think about the shenanigans this could do. Of course, they will have somebody monitoring this kind of a service, but yeah, be prepared folks. But this dovetails with your least favorite thing, Jason, over-employed employees. Because people use mouse jigglers now to appear active, but now you can just have an API do it for you and look even more realistic. I mean, we could literally have it, yeah, be DMing other people or building spreadsheets and you would think they're doing something at work.

1:19:57I mean, a small company, that's not going to work. But a big company, sure, of course it could work. Yeah, fall between the cracks. But it's good to see Anthropic making news, pushing the boundaries a little bit. They've been quiet, talked a lot about OpenAI on the show. So here's the other major American private market model company doing cool stuff. Awesome. And we will see you tomorrow on This Week in Startups. Bye-bye. Okay. I host every six months or so a workshop called Angel University. And this is where I teach people how to become professional angel investors. And the next time I'm teaching the course is on November 6th.

1:20:31And I teach it with my pal, Mike Savino. He's a partner here at launch, one of my best friends. And it's based on my book, but everything I've learned since then. Obviously, you know, I've invested in over 400 startups. And if you've met me for more than five minutes, you know that Uber, Robinhood, Calm are amongst the ones that I've hit that have gone supernova. In fact, Uber is considered the greatest investment over the last decade or two in Silicon Valley. Robinhood, you know, doing fantastic as well. And Calm, not yet public, but another great company. In this course, I teach you the fundamentals, my personal philosophy.

1:21:04And then I compare and contrast it to what other people say. And the most important thing is how do you source and decide which companies to invest in? And then how to evaluate those companies? I have a criteria. I have 13 reasons to invest in a company and about 30 reasons to not invest. And then we call those pink or red flags. Pink flag, that's something you can clean up. Maybe the cap table's a little messy. You know, red flag could be, you know, a patent lawsuit that you don't think they could ever get out from under. And when we talk about those criteria for when you're picking a company, we'll also go into adding value as an investor in startup and then portfolio construction scenarios.

1:21:40Like how many investments do you need to have a chance at hitting an outlier? If you haven't read the power law, you don't know what the power law is, the Pareto principle, go ahead and look it up. We talk about securing pro rata, very important, getting investor updates, what information rights are, and just so much more. We had 1200 individuals join us for this workshop last year. We did four of them. So 300 people at each. Many of these accredited investors have also joined my angel investing syndicate, which is the syndicate.com and you get to see our deal flow. The workshop is open to all investors, whether you're retail or accredited and all the proceeds of this go to charity.

1:22:16You can see a full list of the donations we've given at angel.university slash charity. You can sign up at angel.university. So whether you're an accredited investor or non-accredited or you're just interested in learning, visit angel.university to learn more and register. Again, the next class is November 6th. I'm going to be moving to twice a year for this because my schedule is very busy. So if you don't get in on November 6th, you're going to have to wait six months, clear your schedule, unless it's something really important for your family, you can take a couple hours and learn about how I make decisions and our team make decisions on which of these early startups to invest in.

1:22:50It's not like investing in public companies where you can see how many subscribers Netflix has, or how many Uber rides were taken, or how many DoorDash deliveries occurred last quarter versus a year ago. No, this is a whole different set of criteria when you're dealing with a company in years one or two, or even in year zero. So I hope you come. It's for a good cause. Again, angel.university slash charity to see where all the proceeds go. I'm very proud of the work we've done, Mike and I and the team over the last, I don't know, six or seven years of doing this. We've inspired people to find this new career.

1:23:23People say it's changed their lives, and they love being an angel investor. A lot of times it's young people who sold their company or it's young people who are professionals making a little bit of money. They're making 200 grand or 300 grand working at Google or something. And they just want to learn how to do this. And then all of a sudden it becomes a path to becoming a venture capitalist. Because think about that. If you have no venture capital experience and then you go apply to be a venture capitalist and then I apply and I've made 15 angel investments and two of them have done well and the founders speak highly of me, who's the venture capitalist going to hire?

1:23:54The person who took the initiative to make 15 bets, or the person who just wants to be given a chance, right? You're going to pick the person with more real world experience. And then, you know, a lot of people who are retired and post money, they're 50, 60, 70 years old, and they're sitting there at home on a mountain of cash, and they want to do something fun. We know it's a lot of fun to hang out with people who want to change the world. They're called entrepreneurs, and they're lunatics in the best sense of the word. They have crazy dreams, crazy ideas. And when you're an investor, an angel investor, you get to spend time with them.

1:24:24But you don't have to drag yourself to an office, you don't have to put in 60 hours a week, you can put in five hours a week, you put in 50 hours a week, or anything in between being an angel investor, you make your own schedule, you meet the most interesting people in the world, sometimes you hit a big winner, sometimes you lose. And that makes it just so exciting. And so I think it's a better pursuit than going to Vegas and playing blackjack or betting on sports. I love the idea of betting on startups, because you get all these non-financial rewards that come with it, which is you get to see where the world's headed.

1:24:54You get to see and you get to hang out with inspiring people and see their plans to change the world. It's just an awesome, fun career and pursuit hobby, however you want to look at it. I hope you come. Angel.University.