When Trump Hit Pause & Amazon’s Andy Jassy 4/10/25

24:33Yesterday it happened. And indeed, 3000 points. Well, first of all, I don't really like the characterization that we're divided amongst teams. In fact, I would say that the person who right at the final minute was with the president when the president decided to do the pause that we had been talking about all along was Secretary Lutnik. Secretary Lutnik and Secretary Besant were in the room for the final call with the president. And I think that's a nice clarification of how we're not on different teams. Wait a second. I know boys will be boys, according to Carolyn Levitt. But are you not on X?

25:09I still don't get anything except talk about that whatever you want to call it, that food fight between Elon Musk and Ron Varro, Peter Navarro. Well, as you know, I practice a martial art called Aikido where basically what you try to do is redirect negative energy so that it works against the guy and doesn't turn it into an all-out scuffle. I think verbal Aikido is a better choice. It's a better choice. So, Kevin, I characterize it. it as the art of the deal on one side and hitting your head against the wall until, you know, because it feels so good when you stop and that the actual what happened yesterday was somewhere in the middle.

25:52I guess you'd say it was all art of the deal. This was the plan all along to go China into being isolated. Well, I think that what's been going on all along is the president recognizes that in order to get the big change that we need for America's workers, like these are really fundamental changes that we're talking about that we need to create enough pressure on our trading partners that things that American presidents have been asking for for the last decades are actually offered at the table. I can say that I was on a call with the Swiss president yesterday with President Trump. We've got a whole system set up through Susie Wiles and the cabinet secretaries that we're starting to have a meeting today after lunch where we prioritize our allies and other countries and decide who or what is going to visit and when.

26:38But I can tell you that the amount of stuff that's coming to the table came up really a lot, you know, long before the market movements this week. And so it was looking like that there had been enough action to get us to the point where we should really genuinely start negotiating. And if you go back and look, you can see the negotiation signals were out there already, even on that Monday where somebody made a fake tweet that suddenly said that I was saying something that ended up being kind of mostly true on Wednesday. So anyway, I think that everything was moving forward in an orderly fashion.

27:09And there's no doubt that the Treasury market yesterday made it so that the decision that, you know, it's about time to move was made, I think, with perhaps a little more urgency. But it was going to happen at the moment anyway. You expect real negotiations over the next 90 days. People say it's happy talk. And in fact, all these trading partners are going to be emboldened now to not do anything because they know that the president's not serious or he's going to blink. So they can just wait out. They've got 90 days now and they don't have to do anything. I talked with Jameson Greer, the U.S. trade representative, yesterday, just to get an update on the number of countries.

27:50And, of course, saying you've got a lot of countries calling, that's one thing, right? But, you know, they could be calling to say, how's the weather out there? But then having people really begin negotiations to make offers, that's a different number. That number is like closing in on 20. There's a little bit of discussion about exactly, well, did they really make an offer? But for sure, there's a bunch of offers that are really sensible offers. And they're coming from our top trading partners. And it's some of the most progress in trade negotiation that I think probably the most progress that we've ever seen.

28:17And so I think that this system that the president set up to say, hey, we're serious about negotiation. We're serious about reciprocity. If you guys come down, we'll come down. To do that, we've got to have some negotiations. The fact that people started to send stuff. We had two deals almost closed as of last week. Almost closed. Getting them closed requires the lawyers to work forever and so on. And so getting them closed by yesterday was not quite going to happen. But, yes, there's a big inventory of deals that are right close to the finish line. All right. That's interesting. Kevin, how much of the decision was led by, in terms of to pause yesterday was led by the bond market and what you were seeing there and anxiety on that front?

28:59My view, and you know, I've been on the show for 20 years, right, is always just shoot straight. My view is that they're very, very close. It could well be that the announcement that came out yesterday was going to happen anyway, because there were so many deals teed up. There was a discussion about whether a general pause or a few deals that are just about finalized are the way to move to the next step and to make it clear to markets that we're serious about the strategy that we're following. And I think in the end, everybody decided as a team, all the way from Peter Navarro to whoever else you think is on the other side, agreed to do it this way.

29:32And I think the fact that the bond market was telling us, hey, it's probably time to move, certainly would have contributed at least a little bit to that thinking. But it wasn't the bond market that made a panic move, because there was a very systematic well-planned move that was just about to happen that just turned out to be at the same time. Right. Now, the reason I ask about the bond market, Kevin, is as this plays out over the next 90 days, and we all watch the bond market as well, and I imagine you folks will too, if we see a spike in the bond market in a similar way to what we saw in the last 48 hours prior to the pause, how do you think that changes the negotiations, or if it does at all, and also how our allies or folks who may not be our allies or negotiating against us on tariffs are going to be thinking about that bond market?

30:20Well, I think that you can think about there being two phases. Phase one is going to be where we demonstrate the truth that with the guidance of Susie Wiles, the chief of staff's office, that we've set up a process so that these negotiations can proceed in a very orderly, well-thought-out fashion with timelines and so on and priorities. and again where there's a big meeting about that with senior cabinet officials today here in the White House and then once we show you well here's how we're doing it then I think that'll be very reassuring for markets because I understand not only are they serious but we can even see the schedule and we can see the priorities so the priorities make sense and then I think that the second phase will be when you start to see the actual deals and see the characteristic of them and they say oh not only are they talking about it but they're delivering and oh my that thing they delivered is really great for American workers.

31:09That second phase is going to be much faster than you might think because basically we've got a few deals that we've been working on ahead of this that are really, really well advanced. Kevin, I think that's a big part of the question, what those deals look like. I mean, there had been questions asked because even renegotiating NAFTA took two years to do. Are these just kind of straight across the board, bring tariffs down here, cut these non-tariff barriers? Is it a formula that's kind of working or is every country different? Well, in the end, the asks are going to be put forward by the president, guided by the same trade team that got us to this point.

31:46And I think that the asks will be sometimes extremely clear to the public, and sometimes there'll be quiet things that involve things that people don't even talk very much about involving military matters or so on. And so it's all of the above approach to make the relationships between our countries much stronger than they've ever been before. And I'm really highly excited about it. It's going to be a heck of a time. And when you talk about timing, when I talked about, well, 60 days, 90 days, the book around the world in 80 days kind of ran into my mind. And I'm thinking, well, if we say 90 days, let's shoot for 80 days.

32:19And I think that when I look at the schedules that we're getting ready to debate and talk about, I think it's very doable, very doable. Kevin, when we were still in negotiation mode, we were hearing that we could go to zero, zero, zero with a lot of these countries, but that doesn't cut it. There's so many other things in terms of currency manipulations and, you know, vats and all these other protectionist things that you can't that went into that crazy measurement that the administration used to try to summarize all of them using, you know, trade deficits. That seems like no way you could get some of those deals done in 90 days.

32:59Are we going to accept much less than what we were promised in terms of getting rid of all the impediments to trade? I don't think there's any chance that if Secretary Besant or Secretary Lutnik, the two of the main negotiators for this, are going to go out and come up with a deal and then they show it to the president that if the president doesn't think that it satisfies his objectives for the American people, then those negotiators on our team are going to hear from it. And boy, I can tell you, I've seen what happens in the past in the previous administration. when deals are negotiated that the president doesn't like.

33:31He sends them back to the drawing board until he's ready to strike a deal that he thinks is right to those American people. What are we doing with China? Are we isolating China to finally decouple, which is something that's been talked about, but I don't know in practice whether it's really possible. Or do you expect them to come to the negotiating table? Because most pundits say that Xi is backed into a corner and there's no way he's going to come around at this point. You know, President Trump has a strong relationship with President Xi. They've had conversations in the past that have been very productive.

34:12Those conversations haven't begun yet. When they begin, I think, is a decision that President Trump will decide. And I don't have any clarity on whether that's going to be soon or later. But I know that the president has had a lot of progress, made a lot of progress in the past by speaking directly with President Xi. And I'd be shocked if at some point that doesn't happen. We had Professor Jeremy Siegel on earlier this morning, and he brought up a few questions about this. He said, look, first of all, it takes the markets time to recover from the psychological damage that they've experienced the last eight days.

34:44And he also said he did not think that the United States holds as many cards as we think we do when it comes to negotiating with China. We're not out of the woods on the tariff. I mean, there's, you know, we'll see what gets negotiated. It definitely, you know, will have to be negotiation with China. I don't know whether Trump holds as many cards as he thinks. What would you respond to that? I think that given the massive trade deficit that we have, that there are really quite a few cards that are held that are very difficult for China to adjust to. So think about it right now that there are whole container ships filled with stuff that might be sitting in port in China right now wondering what they're supposed to do.

35:24And then the factories that reload the incoming container ships wondering whether they should shut down for a while because they don't know where they're going to be shipping the stuff. It's an absolute extreme disruption for the Chinese economy right now that's come because of repeated terrible behavior on their part, including aiding and abetting the fentanyl trade. And so I think China's going to have to figure something out. And my guess is it's going to involve reaching out to us and setting up talks. But their economy is certainly right now going through the kind of shock that they probably never expected.

35:59Hey, Kevin, how does this whole pause change the timeline to the extent it does around the budget talks, around taxes? That's something that, you know, House Speaker Mike Johnson has talked about. He thought completing in the month of May, given that I imagine these talks are now going to play out in a longer way in terms of the tariff piece and potentially the revenue that could come with that. And I know that was something that was important to the administration in terms of thinking about how to therefore construct the budget. I imagine we're now talking about sometime in the summer. Does this get pushed to the fall?

36:33Where are we? I don't think the very ambitious timeline that the White House has signaled has been changed at all. If anything, things can even go faster. What's, recall, happening is that the Big Six group, which includes Scott Bessent and me, and then the leadership of the House and the Senate, we've been meeting every week and negotiating the tough negotiating things so that both the House and the Senate, when they move forward, that they're moving forward in a kind of ratcheted way so they don't slip back. And I think that we're going to hopefully see that this week when the House passes the Senate budget instructions.

37:07And then from there, all of a sudden, the committees are going to start writing the stuff. And a lot of that stuff, the committee's been working on since last summer. And so I think that absolutely the big, beautiful tax bill is right on schedule. And then it is a technical matter that I know that market participants like to think about is that now that we've got, say, $2 billion a day in tariff revenue coming in, That's a lot of revenue that can help reduce budget deficits should be good for bond markets. How does that influence the package? And the bottom line is that what will happen will be that the CBO, when it sees the revenue, will start to adjust the baseline for revenue to include the tariff revenue that comes in.

37:47And that will reduce the deficit. It'd be good for bond markets, but it won't affect the overall numbers that they agreed to. It won't count as a tax increase. right? It's just like a baseline adjustment. Kevin, I appreciate the perspective that it won't be part of the full CBO scoring. That's something that we've talked to Secretary Besson about. But I always thought that the administration and at least some Republicans were thinking of, you know, trying to include some of those numbers, at least in terms of how they were going to sell the plan to the public and to others. When, you know, if we're talking about taking down, taxes on tips and things like that.

38:26And part of it was creating some kind of permanence around those tariffs so that you knew that that revenue was also permanent. If that revenue is not permanent because there's still talks going on, that will therefore change the dynamic with which you would negotiate around the budget, no? Well, what I'm meaning to say is that the tariff revenue, if Congress decides to write into law the president's objectives, then we would, of course, celebrate that. But there's a plan right now that gets this tax bill done, you know, before we all go home for vacation in August and, you know, well before then, and that it doesn't require any tricky maneuvers with the tariffs.

39:04The ball is on the tee and they can hit it. And the tariffs are going to be something that should make the budget hawks really happy because the deficit will be lower than they ever dreamed possible, but it won't affect the reconciliation process. So, Kevin, you add the 10 percent, the tariffs on Canada and Mexico, and you add in China, and these are still the highest tariffs I think we've had people say since Smoot Hallway. So is that permanent? Is this now, we heard the other things are going to be permanent. It's hard to really gauge what's real and what is sort of more posturing. But do you expect this to be a permanent situation or this is fluid as well?

39:51I think everybody expects the 10 percent baseline tariff is going to be the baseline and that it's going to take some kind of extraordinary deal for the president to go below there. In the way that I don't know if history is ever going to really come down with sort of the administration spin on how it all happened. And you've seen some of the articles that, you know, that the president was watching some TV yesterday and, you know, that it, you know, saw different things happening that finally forced his hand and that some of his closest advisers didn't find out, you know, exactly what was happening.

40:31It just seems like a lot of this, if you're going to call it the art of the deal, should we just assume that it is an art, it's not a science, and you can get to this point and it was all deliberate when it's all said and done, or are you just sort of being buffeted by circumstance and finally arriving at a policy that affects the entire globe, frankly? No, this policy was discussed at length with a whole bunch of trade advisors over many, many weeks. In the end, the president had multiple options to decide between, and he knew that he had to create leverage to get people to the table to begin negotiations.

41:09He did that before the final decision, and the negotiations lifted off like really crazy, right, a couple days before markets started gyrating. And so then the question is, how do we start to show that we're genuinely going to have these negotiations? And we have a couple of deals that are close to finished, And we also had the idea that we could have a pause. And the president decided to pick the latter approach, which was something that had been studied all along, all the way to the beginning, about, like, how do we go from first, let them know we're serious, to second, getting negotiations that actually delivered for the American people.

41:45And the president chose this path that it was not something that he thought up the last minute. But was it deliberate deliberate overkill on the original the calculation used for those very high tariffs that allowed the United States, you know, to cut it in half and have very high tariffs? Was that all part of the part of the plan? The original IEPA was meant to address the trade deficit, which is viewed as a national emergency. And having the tariffs be a function of the trade deficit seemed like the logical thing to do for the lawyers at the USTR. and those numbers ended up being pretty high.

42:21And I think, you know, did the higher numbers help get people to the table, help people understand the urgency of what we're trying to accomplish? I'm sure they did. Okay, good. Kevin, thank you. Kevin Hassett. Thanks, guys. Great to be here. Hope to see you again soon. Thanks. Cheese will be next. Coming up next on Squawk Pod, an exclusive interview with Amazon CEO Andy Jassy. He runs the biggest retail name in the world. He'll join us on rising costs, tariffs, and how Amazon is spending billions on AI. If you sat in the meetings with the AWS team right now, they feel like it's their responsibility and their mission to make the cost of AI meaningfully less for customers than it is today.

43:12You're listening to Squawk Pod from CNBC. Today, our Andrew Ross Sorkin sits down with a special guest, Amazon CEO Andy Jassy. Jassy released his annual shareholder letter just ahead of this discussion, predicting a rapid advancement in the mega cap tech's investment in artificial intelligence. Stand, Andrew, bye. In three, two, one, cue Andrew. Here's Andrew. We are in Seattle this morning at Amazon's campus in downtown Seattle. We have an exclusive interview with Amazon CEO Andy Jassy out with his annual letter today. And there's so much to talk to you about. Couldn't be more thrilled to talk to you, especially right now.

43:52So what is going on right now in terms of the way you're thinking about the tariff position? Obviously, the China piece of it is still on. And a huge portion of your sellers and manufacturers that are selling through Amazon manufactured in China. Well, it's hard to know what's really going to happen. You know, there's a lot of flux right now. And, you know, we're going to spend time on where we spend all our cycles trying to figure out how we have the right customer experience. So we know customers care a ton about selection. We're going to try and do everything we can to keep prices as low as possible for customers.

44:31And we're going to ship things as quickly as possible as we can for our customers. And that's what we're going to spend our time focused on. There have been some reports that you've already started to cut back, though, in terms of buying certain product from China. Is that true? Well, what I would tell you is whenever you have any threat of any kind of discontinuity as a team, you have to think about what are the things you can do to help customers. And so we're doing everything we can to try and keep prices the way they've been for customers as low as possible. And, you know, so we'll do things like we've done some strategic forward inventory buys to be able to get as many items as make sense for customers at lower prices.

45:13You know, there are some cases where we have deals we were negotiated that weren't done where we'll renegotiate terms to make it easier for customers to have lower prices. So we're going to do everything we can to keep prices as low as possible. Are you seeing, by the way, any pull forward in terms of folks saying, I need to buy? My son just yesterday, who's been desperate for a new laptop, said, Dad, maybe we've got to buy the laptop now because it could cost a lot more later. You have a smart son. How old is he? 14. Oh, my gosh. You know, I think, again, it's such little data at this point since, you know, the first set of tariffs came out and then yesterday there was a pause.

45:49But we have seen, you know, we see people have not stopped buying. And in certain categories, we do see a little bit of people buying ahead. But it's hard to know if it's just an anomaly in the data because it's just a few days or how long it's going to last. But, you know, we're going to do, as I said, everything we possibly can to keep prices low for customers. Who is going to eat the cost, though? Because I think that's one of the big questions. How much can get passed on to the customer? How strong do you think the American consumer is right now? and how much of it gets eaten by Amazon, how much gets eaten effectively by potentially a third-party seller?

46:25I think we'll have to see how it all plays out. But if you made me guess, I'm guessing that sellers will pass that cost on. I think they'll try. And I understand why. I mean, depending on which country you're in, you don't have 50 % extra margin that you can play with. So I think they'll try and pass the cost on. And, you know, as I said, it's so early right now, but we haven't seen any change in consumer behavior in a meaningful way yet. Is there any change in terms of your own behavior, meaning the company's behavior? One of the things you're doing is building data centers, for example, and having to invest heavily in that space.

47:07Obviously, there's component parts that are coming in from China and all over the world. And how you think about that, for example? Well, every finished good that's got any degree of complexity in it has parts from all over the world. And that's been true for a while. And actually, in our AWS business, really starting about five years ago, we very significantly diversified our supply chain. And so we have components that are from everywhere. It's not just one country. But we have such high demand right now for AWS, and the AI growth is so significant that we don't see any attenuation in demand, and we're going to keep building.

47:46And so depending on what happens with the tariffs, there'll be no change to that view. I don't see us attenuating or building data centers right now. We have very high demand. Can I ask one related question in terms of data centers? One of the other sort of major shocks to the system was now a couple months ago was the creation of DeepSeq, which was this new large language model that people thought was remarkably more efficient. And there was this view among some that said, maybe we don't need all of these data centers. We're not going to need all of this processing, as much processing power and as much energy as some had forecast.

48:20Did it change your forecast when you saw that? No. I mean, in some ways, we would like nothing better than to make the cost of AI meaningfully less expensive for customers. The cost of AI today, because of the chips and because of the cost of inference still, even though the cost of inference has come down several times, but it's still more expensive than it should be and than it will be. and then what you need for companies to be able to deploy AI as expansively as they want. And so when DeepSea came out with this model, which people were excited about, it felt like it cost less, it felt like the inference was less.

49:01You know, I think if you're building frontier models like we are as well, you're all working on a lot of the same types of problems. And you're going to see all this leapfrogging. You have seen all this leapfrogging every month, and you'll continue to see that. But for me, the lower that we can make the cost of AI, the more customers are going to use it. And I think people get confused. I saw this in the beginning of AWS, too, which is customers love when you take the cost per unit of something down. It allows them to save money on what they're building, but they don't actually spend less. It actually unleashes them to do a lot more innovation.

49:35And in absolute, they spend more. It's the same thing when we started AWS. US would pretty significantly change the cost of compute and storage and database. Customers loved it, but they actually spent more on infrastructure because it allowed them to build more for their customers. That's something you talk about a lot in the letter, bringing down the cost of compute when it comes to AI. And obviously, you also make an illusion. I don't think you use the word NVIDIA, but I know there was maybe a frenemy partner and a competitor at the same time. They have controlled this marketplace for a very long time.

50:05What do you think has to happen to bring the price down? because I assume that that's going to come from you and AMD and everybody else who's going to try to do this. But how quickly does that happen? It is going to happen. I mean, first I would tell you that if you sat in the meetings with the AWS team right now, they feel like it's their responsibility and their mission to make the cost of AI meaningfully less for customers than it is today. And I think there are two big drivers of it. One is the chips. And, you know, what's interesting is most of the spend today in AI has been on training, these giant training runs.

50:41But when you get the scale of the production applications for AI, it flips. And the overwhelming majority of the spend is actually on the inference, which are the predictions from the model. And so the two things that have to happen is the price performance of chips has to be meaningfully less. And that's part of why we built our own chips. We have our own custom AI, Silicon, called Tranium 2, where the price performance is 30 % to 40 % better than what's out there in the current GPU instances. So we're working on trying to take the cost of chips down, and you have to take the cost of inference down.

51:13And that has to do with a whole bunch of things, including techniques like distillation and prompt caching and how you handle different model architectures. So it's both the chips and the inference. They both need significant cost decreases, and we're working hard on that. One related AI question, which is last time I think I talked to you maybe a year or two ago, there was a different administration in Washington. We were talking about the possibility that you would consider to buy a company like Anthropic or something like that. You were in the middle of the iRobot situation, and it didn't seem like any deals could get done.

51:48In this new environment, do you say to yourself, maybe there's acquisitions that we could make and maybe buy an Anthropic? You know, we never stopped trying to do acquisitions that we thought make sense for our customers and the business before. But I do think that it's possible this environment will be a little bit easier to start to acquire companies. We'll have to see. In the letter, you do talk a lot about the consumer. You talk a lot about the culture at Amazon, about trying to bring down prices. You just said you thought that, and you talk about how optimistic you are. You see no change in the consumer at all in terms of what they're buying?

52:25I mean, we've had conversations over the years where you've said, you know what, actually, I see people going from bigger screen TVs to smaller TVs and other things like that. Is there no shift right now in terms of what the consumer is doing? I'd say not a lot in the last – I mean, over the last couple of years, off and on, as we've talked about, customers have gotten more careful. You know, and they there whenever they can trade down a price, they do wherever they can find a bargain. They do. We've seen significant growth on, you know, discretionary items, but not at the growth rate. We typically see as people have just been careful.

53:01And so that trend has continued. But as I said, you know, we haven't seen meaningful impact yet from all the noise around. But are you thinking there is? I mean, we've heard, you know, I think Walmart around guidance. You heard Delta talking about how they think that we're I don't know if it's a slowdown, but you know not gonna grow as fast We've heard Warner Brothers. I think is worried a little bit about advertising. I know you're in the advertising business What does this feel? I completely understand why they made, you know, companies are nervous right now and you just It's so early you don't know but the worry of course When you have things like tariffs is that it'll create macro issues where you know it depresses demand or drives inflation But we just haven't seen it yet.

53:44Have you either been in Washington yourself or had your team in Washington talking to this administration about all of this? We talk to the administration all the time. And what are you telling them? You know, we talk about a broad number of issues with the administration all the time. And so, you know, we share with them different things that we think matter to consumers, things that we think matter to enterprises. is we have some conversations about tariffs as well, and we share what some of the concerns are, and they're aware of them. One of the things you talk about a lot in this letter is sort of the management and culture that you've instilled in this business over the years and how long you've actually been at this company.

54:29But one of the things that you've talked about is this idea of why, that every question starts with why. Why are we doing this? It sounds like really the question is not why are we doing this, but why can't we do something? Mm hmm. Yeah. How what what is the right now? Is there something that you're looking at where you're saying, why can't we do something? Oh, I mean, we have so many why questions all the time. I mean, we have this concept inside of Amazon of two one way and two way door decisions. And a two way door decision is when you make a decision. And if you get it wrong, you can walk back through that door with little harm.

55:03A one way door decision is when you walk through that door. If you get it wrong, it's pretty tough to walk back through it. And so both those constructs, though, assume that the doors are unlocked. And when you're trying to invent, you're often trying to, you're actually trying to unlock a door that seems previously shut tight. And so a lot of times, why is the question that starts to open up that door? And we're constantly asking ourselves why questions. So today, we're constantly asking why can't we ship items to customers even faster? Why can't we ship items to people who live in rural parts of the country who don't have, it's much harder to get items same day or next day in rural parts of the country.

55:46And most of the logistics providers are actually stopping investing in rural areas, whereas actually we're increasing our investment. Why can't we give broadband connectivity to people in rural parts of this country and around the world? Why does AI matter so much and why do you have to invest the way that we think we should right now? We're asking lots of why questions. I'll give you a why question that some investors ask. They say, why necessarily go down some of these roads which might eat into margin? So one of the questions is, if you were to be able to get stuff faster in rural areas, I assume the cost goes up, or do you think overall it makes the business better?

56:24I think there's no question it makes the business better. And for us, what we're after, and we've talked a lot about this over the years, is we're trying to make customers' lives easier and better every day. And we're not trying to optimize how we do as a business in any one quarter, any one year. We're trying to build a business that outlasts all of us. And when you're actually, what we learn, I get asked this all the time, and we get asked this a lot, which is, haven't you reached the law of diminishing returns on speed? Like, isn't it enough at this point? And I can tell you definitively from the data that it's not.

57:00We have not reached that point. Every time we're able to show customers on a product detail page a delivery time that's faster, they actually choose to buy more frequently, and they actually spend more of their purchasing power on Amazon and make more of their purchases with us when they know they can get it reliably and quickly. So I think that in any short period of time, as you're quickening the pace with which you deliver, you may have cost increases. But if you build, if you innovate the right way and you build the right infrastructure and the right architecture, you actually can relentlessly take your cost to serve down, which we've done over the last several years.

57:38And we're very confident we'll be able to do that in rural areas as well. They deserve to get products same day. Question on cost to serve, or maybe it's maybe cost of cost of everything. When you think about employment at Amazon, one of the things you talk about in your letter is this idea that you want everybody back in the office. You've been a big proponent of that and how successful that's been. I'm curious, though, with the advent of AI and everything that's going on, and you look at how many more people you want to hire in the future, or maybe how many you don't need to. What does that look like in this new world to you?

58:09I think, again, it's so early right now with respect to AI. You know, you're starting to see companies employ it in a really broad way. And they're going to, you know, like everything else, as technology evolves, certain jobs become more important. other jobs that you used to do with people you do in an automated way. But we have found over many years that human beings find other jobs to do, the ones that get automated by AI. So I think we'll have a lot of people for a long time, but they're just going to be working on different things. I want to pivot back to the news, a different piece of news.

58:44You made it a couple of weeks ago, which is that you're bidding now for TikTok. Give me the why behind that. I don't think we said that. Did you not say that, that you were one of the bidders for TikTok? It's just been reported that you were one of the bidders for TikTok? Okay. Well, can you give us the rationale for bidding for TikTok then? Oh, well, you know, we have a long policy, as you know, Andrew, and not commenting on M &A rumors, so I'm going to stick to that policy. Let me ask you a slightly different question. You don't need to comment on whether you're bidding for TikTok or not. Do you think that TikTok should be part of a chess piece in this tariff conversation or not?

59:24Because I think there's a TikTok has always been a question about national security, but it now is also being sort of brought into this tariff discussion. I don't know. I mean, it's I think the conversations that the administration is having probably was having before and more intensely is having now with each country are complicated. And I imagine that each have their set of variables. And I've read that this may be one of the pieces in that negotiation, but I am not privy to it. If this doesn't work out, being a politician might. So that's a good answer. Let me ask you about a couple other pieces of the Amazon business.

1:00:05One of the sort of next legs of the stool for you, if you will, is Kuiper, which is the satellite business. You're putting up a satellite, you believe, this year. It's finally going to happen this year. We're starting to put satellites in the air this year. What's that going to cost? And how quickly do you think that becomes competitive if it does with Starlink? Well, I mean, I'd just start with the need. You know, there are 400 to 500 million households around the world who don't have broadband connectivity. So they can't do things that we all take for granted, like education online or shopping or watching entertainment or doing business.

1:00:40And then there are all these enterprises who have assets in remote locations that they need to communicate with they can't. And governments need that visibility. And so there is a high need for it. We will have our low-Earth orbit satellite. We'll have the first set of satellites up in 2025 with more coming in 26 and 27. I think that people are going to really like the offering. I think that it will have an advantage price point for customers in those three segments. I think they'll like the flexibility and the price point of antennas. And then if you think about it, we have at Amazon, we have in the three segments that are going to consume this product, We have a lot of relationships with consumers, a lot of relationships with enterprises, a lot of relationships with governments, all of whom are excited about this.

1:01:29And particularly the commercial customers, enterprises and governments, are going to want that data to come down into the cloud. Who's going to be the first real competitor to Starlink? Yeah, I think people are going to really like the offering. And I think they're going to like the connection between Kuiper and AWS to be able to analyze that data and do computing and AI ultimately. And so very optimistic about it. How many of the satellites end up going up on Blue Origin? How many end up going up on SpaceX? And is that a concern for you, given that you have to rely to some degree on Elon Musk to get some of this in the air and then compete against him?

1:02:03Yeah, we have four rocket providers. And, you know, all those rocket providers, you know, they're very professional. You have contracts for a reason. And, you know, I would say that we have more in absolute going up on Blue Origin than we do maybe in SpaceX. But we have four providers, and they're all going to be an important part about getting our constellation into orbit. Two other pieces. I want to ask you about Video Prime and want to talk to you about pharmaceuticals for consumers because you mentioned that in your letter. Is that the next big leg of your stool, meaning getting into health care in a big way?

1:02:40You've been talking about it forever. I just don't know if you think there's been some kind of unlock that's happened in the past year. I think if you look at our Amazon pharmacy business, it's really growing very substantially. And I think that there are a couple of reasons for it. You know, first, if you look, I don't know if you shop in the pharmacy, it's really it's a pretty awesome experience. But the team has done an amazing job the last two years or so in changing that customer experience. It's so much easier. You have price transparency. A lot of times you go to the pharmacy, you don't know what something's going to cost.

1:03:11You get to the counter and you can't believe it. And so the experience is better. The delivery is much faster. And then I think people have gotten used to buying their pharmacy items online. If you don't feel well, if you can get it same day, we can deliver same day now in lots of cities around the U.S. If you can get it same day and you can get it actually in a few hours, it's a huge advantage if you don't feel well. What's going on in terms of Video for Prime? You've been investing heavily in sports. Does that mean you're going to invest away from some of the drama and comedy? What does that look like over time to you?

1:03:44I think it'll be a mix. You know, I think what we've you know, there's a reason why something like 95 of the top 100 rated shows every year are sports. You know, it's kind of the last appointment only viewing. And we've we've learned it with Thursday Night Football. It's been really successful for for customers and for us. And we have the NBA coming in 2025. We have NASCAR coming in just a few months. And so sports has been very successful. I expect we'll continue to invest more there. But we'll continue to invest in shows. You know, I love the new season of Fallout, you know, the new release and Reacher and the boys.

1:04:21I think the content continues to get meaningfully better. The economics of that on a standalone basis, would they work or does it only make sense in the context of the sort of larger Prime package? You know, we started Prime Video as something that we thought customers would love and would be very attractive in Prime and would help, you know, convince people to decide to pick up Prime. And then what we find is when people start Prime via video or for video reasons, they actually end up spending downstream in our commerce business, our e-commerce business. But what we have over the last few years with some of the changes we've made and as the business has matured, we actually think Prime Video will be a very good standalone business as well.

1:05:03Any shows that we need to watch? By the way, oh, First Lady. You have a documentary with the First Lady. How did that come about? Yeah, it was the people making it shared it with a number of folks. We thought the show was really interesting, and we were fortunate enough to earn it. Do you think you'll end up getting into the business of news? The reason I ask is you did that broadcast the night of the election. Yeah, it was an experiment. And actually, it was a really successful experiment. Customers really liked it. I don't know. It's something, you know, it is something that people watch very frequently, as you well know.

1:05:35And we'll have to see. OK, Andy Jassy, it's great to have you. It's great to congratulate you on the letter and great to have you on a day like today when we're all trying to make sense of these tariffs. And we hope to talk to you again as all of this plays out. So thank you very, very much. Appreciate it. That's it. That's the pond for today. Thanks for listening. Squawk Box is hosted by Joe Kernan, Becky Quick and Andrew Ross Sorkin. tune in weekday mornings on CNBC at 6 Eastern. To get the smartest takes and analysis and sometimes a little extra from our TV show right into your ears, follow Squawk Pod wherever you get your podcasts.

1:06:10We'll meet you right back here tomorrow. We are clear. Thanks, guys.