Sharp Tech with Ben Thompson - Netflix as a Sleeping Ad-Supported Giant, 'Max' and the Other Streamers, Whether Nvidia Shall Inherit the Earth

Episode Date: April 24, 2023

Potential implications of Netflix advertising success, checking in with the rest of the streaming landscape, and the long-term questions surrounding Nvidia and the GPU ecosystem. At the end: Two follow-ups on AI and copyright, a Silicon Valley clarification, and Bucks therapy.

Hello and welcome back to another episode of Sharp Tech. I'm Andrew Sharp and on the other line, Ben Thompson. Ben, how you doing? I'm doing okay. Doing okay. Surviving. That was a real Monday morning sigh there. Yeah, well, I mean, as a Bucks fan, there's not a lot to be happy about.

But, you know, nothing we can do on a Monday morning except push forward. That's right. That's right. One foot in front of the other is what we're about today. We're going to come back to the bucks at the very end of the episode. We do have a lot of ground to cover. I want to talk Nvidia in the second half of the show. We've also got some follow-ups from last week's show that we'll try to hit.

Send questions to email at sharp tech.fm. If you've got topics you want us to discuss on Thursday's show. But Ben, we will begin with Netflix almost exactly a year ago, April 4th, 2022. In quarantine, in government quarantine. On top of a in a cement room on top of a mountain. I completely forgot about your, how many days was it? Was it like 18 days in quarantine? Yeah.

I was podcasting with you throughout recording some beta episodes. And I remember thinking to myself, how is Ben staying sane? You were much cooler about the whole thing than I would have been. But we're just happy that you made it out. So in quarantine, you managed to write, why Netflix should. sell ads. And then right on Q, a couple weeks later, Netflix leadership announced on a call that they would, in fact, be offering an ad-supported tier in the near future. Reed Hastings at the time said, quote, those who have followed Netflix know that I've been against the complexity of advertising

and a big fan of the simplicity of subscription. I'm a bigger fan of consumer choice and allowing consumers who would like to have a lower price and are advertising tolerant to get what they want. And 12 months later, here we are. The ad-supported tier is live, and last week, Netflix said that in Q1 this year, the ad tier generated more revenue per user than the subscription tier. So, Ben, big picture, what do these numbers mean in the context of the larger business? How much does this development actually matter? And how encouraging is this early on for Netflix here?

Yeah, I think the numbers are super important because, so Netflix has three tiers. They have like basic tier, which is like fake HD, which is like 720 or whatever, which is an upgrade from being like standard definition like it was for years. You could have one concurrent stream and that's, I should, should have the numbers in front of me, like $8 or $9 a month or something.

And then you have the standard tier, which is $15.50 a month, two streams and one. real HD and then you have the Ultra, whatever, you know, big plan. That's four streams, 4K, and is sort of the most expensive plan. And what they did was they started out by offering Basic with ads, where it was cheaper than Basic,

but had the same sort of limitations that Basic did. There were a couple other limitations. Not everything in the Netflix library was immediately available in the ad-supported tier because they actually had to get permission from some of the shows that they didn't have rights to to adjust the business model. But they said that almost everything is on it now. But the important thing here is it's not that basic with ads, which is cheaper than basic,

made more money than basic. Although that was sort of the case, I think, worldwide. In the United States and Canada, which are their most important markets and are the largest advertising markets in the world, basic with ads made more per user than the standard tier, the one that is like $8 or $9 more. And that is a huge deal. Now, there's a lot of caveats that go into this, which is, you know, the ad product is just getting started. At the beginning, there's always sort of a honeymoon period because you're in like advertisers' experimental budgets where they always have money to throw.

Like, that's where like companies like Snapchat will build their business or like Twitter back in the day, right? Or TikTok today. It's like there's less focus on sort of ROI and more let's figure this platform out and see what's sort of available. And Netflix is still in that one can. assume. And also, they don't have that many subscribers sort of in this tier. So if they have a lot of demand from advertisers and not that much inventory, relatively speaking, the prices are going to be fairly high. And so it's not, the caveats are important and worth calling out because we'll see how this sort of scales over time, once it's sort of a consistent thing. But at the same time, you should also expect the ad product to get a lot better over time. They get much better at targeting, understanding who the users are. They're talking about building sort of, you know, an sort of sell side invitation only, but sort of like auction model where you go in and you buy the ads without having to like, right now you go and you make a deal to buy ads. Again, another reason to be wary of the prices that are being charged right now, but they want to

get to where you buy ads on an auction basis like any other digital platform. And again, it's not going to be anyone can just come to do ads. It's sort of invite only. They want to sort protect the quality of the advertisers for now anyways. And so again, caveats are important, but the product should get better. It should get more attractive. And this is a huge deal. It's a massive deal. I mean, what I wrote about,

they should get ads. It was not just that it's an extra revenue stream, but also they could sort of bring on new subscribers, increase their base. They've been saturated for a while. It also made sense in the context of they're kicking people off the plans that are not at home anymore,

password sharing, all that sort of thing, because they have a much lower entry point for those people to sort of come back on. But the fact that it seems to be, doing so well. Again, with the caveats, well, the caveats are important, but I'm going to sort of not say that granting that again, just sort of keep that in mind going forward, is this just transforms, in my estimation, the long-term outlook for their business, right? Basically, what advertising gives you is basically an infinite ability to raise prices because the prices you're raising are not on the

viewers, right? The problem with raising prices on the viewers is when you raise the price, they might churn because it's too expensive or XYZ, either go down to a lower plan or whatever it might be. Whereas if ad prices go up, that doesn't change the user experience. Again, it can change it if you're increasing the number of ads per hour, which they might do at some point. But by and large, if someone's in the ad tier, they're probably mostly fine with it. And as the ad prices go up, their revenue per user goes up without the user having to feel the pain of that price going up, which gives them so much more latitude for increasing revenue. And oh, by the way, also gives them more revenue to raise prices on the non-ad plans

because they like, look, if you don't, if it's very clear what you're getting for some churn. Yeah. If you don't want to pay that much, go to the go to the ad supported plan, right? And like, and it's a very this, instead of this weird like somewhat HD, real HD, 4K, like the differentiation that they were, you know, in some respects that 4K plan was, really the password sharing plan, right? Because that's how you do like that. True.

Four streams. That's right. Now there's like it gives them more latitude on that side as well. Because look, if you don't want to pay for ads, absolutely. We will continue to give you an ad free products. But we're going to make you live up to the deal, which is we give you content you want. You're going to pay for it, which they've always been, they've always been transparent about that. That's why we're going to raise prices.

And so this is just sort of a win, win, uh, all around. And I think as even more of a success than I expected this early, again, with the caveats in place, but no, it seems like a home run. Yeah, you know, I want to be careful of confirmation bias here. But when we talked about the ad supported model a couple months ago, I think I came on the podcast and said, you know, this could be how Netflix conquers the world over the next five to 10 years. And this could be the model that really comes to dominate their business. and on my basketball podcast, greatest of all talk,

I like to call myself write about the big things sometimes. And I do feel like this is an early write about the big things call on Sharp Tech. And what I find most interesting is that Netflix has been this shape-shifting company for like 25 years now. And with the news last week, they also, they announced they'd be shuddering the DVD business. So that's topical. but they've gone from DVDs to streaming other shows to then making their own original content. And this could be another shift that would be pretty consistent with their history, where four or five years down the line, like the ad-supported product could just be an absolute

monster here and could be accessible to a much larger scale of audience. And, you know, again, it's early, but I'm very curious to see how they build on this. Yeah, I'm going to steal your job. It sort of go to the next question. John says you wrote last week that the more subscribers Netflix has on its ad plan, the better. He asks, why charge so much? Why not charge less and add more subs? It's not exactly as if 2 million quarter over quarter, and it adds is a strong, robust number.

Shouldn't they be more aggressive, if not now than way this year on price? And the answer, John, is yes. This is like, this is the other big sort of opportunity. You can actually see, and I think someone asked us this question on Sharp Tech when we, I think it was when we were actually officially recording. I guess so because we had email feedback. back, but like, wouldn't there be a free Netflix? And I kind of pooped the idea a little bit, but now it's like, why wouldn't there be a free

Netflix, right? Isn't that where this eventually goes? It seems pretty clear, right? And suddenly that dramatically expands the scope of sort of what's available. And it's funny, you kind of end up with where YouTube and Netflix are in some respect. I've always said for all along consistently YouTube is Netflix's biggest competitor, right? But there have always been two big differences. difference number one is Netflix is all sort of produced content and YouTube is primarily UGC content.

And their brief foray into making their own shows was sort of kind of a disaster, right? That's difference number one. Difference number two has been the business model where Netflix has always been subscription and YouTube has been ads, but you can sort of get a subscription to take the ads away, which as I know what I'd do they is like one of the best deals in the world. I don't care how much they charge having no ads on YouTube is an amazing experience. You know, you mentioned that on dithering. I was not aware that YouTube had a premium product,

and now I feel like a complete room for sitting through the ads over and over again the last several years. It's life changing. It's life changing. It makes it, you don't even realize how bad it is or like how convenient it is not of the ads until you like actually go in a browser that you're not logged in or something. Oh my God. What is this? This is ridiculous. Ads in the middle of a video. We're going to finish recording and I'm going to go subscribe to YouTube premium.

No free ads on the podcast. but that's just a heartfelt endorsement from Ben. And all right, so tell me more about YouTube and Netflix. Well, I mean, it seems pretty clear in the long run. This is the optimal outcome for Netflix is where there is a free ad supported version. And then you can pay if you don't want ads, right? Like it's kind of like where basically where YouTube ended up. And, you know, I thought it was super telling one of the most important takeaways,

I think, from the last earnings call is they are now changing. the memory, it was basic with ads. Now they didn't change the name. They're just like, oh, it's now full HD and you get two streams. It's basically now standard with ads. They're increasing the quality without changing the name because I think they recognize this is actually a really compelling product. It's in our interest.

Before everyone at Wall Street, I wrote this whole time saying like, this is a stupid thing to worry about. Everyone was like, oh, they're going to cannibalize their business. Like, cannibizing their business is good. you want to get more people on this tier. You're going to have the more people you have with advertising, the more compelling the product is going to be, the more targeting you can do,

the more sort of focus you can be, the bigger base of advertisers you can draw in. And I think they're on to it, right? Because they basically completely changed uplevel the quality of the ad support of products. I like what they did. My, you know, I called for, I think, one advertising product. But then when they did,

I'm like, well, at least have an average. advertising version of the standard plan and the ultra plan or whatever it might be. Yeah, I was surprised how basic it was. Well, and also just to answer, John, they're not lowering the price of the ad supported product, but they are now delivering more value at the current price. So in a way, it's like lowering the price. But I agree that eventually it's going to make more sense to just get really aggressive in terms of getting people onto the ad supported tier.

Yeah. And, you know, ads opens the door to lots of stuff. I've always been, you know, maintain Netflix is not going to do sports, not going to do sports, not going to do sports, not going to do sports, it doesn't make sense. And there is still as an extent, I think they said on an earnings call a quarter ago, they're like, yeah, like, that's renting, that's renting content and it's very expensive. And all that is true. But, but there is, like, there is room at least, a possibility.

There's no way it makes economic sense to do sports unless you have an ad business. Now, again, Amazon Prime and Apple will dink around. with it, in part because their business and their goals are very different. But for a stream for service, part of the benefit of doing evergreen content is it's there all the time. It's a benefit for not just current subscribers, but future subscribers or reason to sign up, where sports has to be live, right? And that is what makes it compelling for advertising.

And now that's at least a possibility. Now, that's, I almost regret bringing that up. That's not really the most important thing here. The most important thing here is the prospect of dramatically expanding the, the, their addressable markets, right? Like literally, if there's a free version Netflix. It seemed like it had sort of plateaued, at least in the U.S. Oh, yeah, yeah, it plateaued like two or three years ago.

Yeah. Right. And so now we're just expanding the pie in this massive way. And the possibilities once you start looking at the market that way are pretty interesting. I mean, it's funny. I go back to the why Netflix should sell ads and it feels so timid in some respects, right? It was timid. I remember talking to you at the time.

And, you know, it was like, well, they've got this really distinct brand identity and they want to be really careful with it. And there was like a lot of concerted debate about whether it's a good idea. Now you look back and it's like, of course it's a good idea. Like this is the ultimate no brainer. And suddenly they have to be taken seriously as like a real like toe to toe competitor with everyone in the energy. entertainment space. I mean, they've already been there, but like it could get even crazier over the next 10 years. No, absolutely. I mean, there's almost sort of a life lesson or a takes game lesson here where when everyone is sort of united in a view and there's just like conventional wisdom. Of course Netflix should not do ads. You know, how that's a, it's almost like an unidentified cognitive bias. Maybe it is a cognitive bias and someone's got an email and say this is where it is. But it's like the wisdom of the crowd's cognitive bias. where you don't not only feel like you're out of pocket suggesting otherwise,

but your suggestion is actually too unambitious because you're like, I'm just, I'm already, I've already kind of sort of sticking my foot out here. I'm not sure if I want to go all the way, but it's like, yeah, I wish, because it's one of those things to your point, of course, this makes total sense, right? And in retrospect, it's like, oh, of course you should have done it. I should have written before. It's not just you get these benefits is that you can actually become YouTube in reverse. and yeah, yeah.

Well, and also we can't know the internal deliberations at Netflix, but another funny aspect of the story was the way this was announced a year ago. It sort of felt like a plan that they came up with on the fly as the stock was taking a beating. I know. It's like these results are really bad. We're going to get killed. What can we do to placate investors? This is absolutely what it felt like.

Yeah, exactly. So who knows, but they got there eventually. So one question, the rest of the streaming space continues to look like a disaster and there's all these people lighting money on fire. Do you think there are lessons here for some of those players or is Netflix a unique case because of the massive scale they were starting with? Well, everyone else sort of already knew this was what they should do. So everyone had announced or was already doing ads. Right. And I brought this up in the article.

I was like, look, Hulu makes more money from their ads supported customers as they do their regular customers, right? It was sort of a thing that everyone knew, which makes it all the more remarkable that, you know, the conventional wisdom was sort of in the opposite direction. It didn't be clear, it wasn't just me seeing they should do ads. There have been people on Wall Street that have been calling for it for ages. So I don't want to like pat myself on the back too hard here. But yeah, I mean, I do think it's a big problem for everyone else just because the advantage Netflix has had. And the degree to which that advantage manifests has always been sort of tricky to ascertain. There's been a lot of debate going back and forth.

Like what benefit do they have by having just way more customers? I overstated the benefit probably in the midterm by saying because they have more customers, they can pay more for content because their per subscriber costs will be lower. And I still think that's the case in the very long one. That's one of the reasons why I've been optimistic about a lot of these other streaming services going out of business because they'll realize we can just make more money by selling the Netflix. Like we're actually content makers to add on this distribution. We're not only spending all this money to build a capability we don't have.

We're also diluting the money we can make by just selling to someone else that does it better because they have a larger user base. This takes that argument to 11, right? If Netflix's user base actually expands significantly because, say, they have a free tier and those extra users, Netflix is suddenly uncovered, are actually worth, you know, tens of dollars a month. like Netflix's relative buying power, their buying power on a per viewer basis is going to be so much higher than everybody else. That argument is just going to be even stronger. And I think a lot of investors are going to look at these companies

and say, look, you're spending so much money on this streaming service. And again, it's not just the cost of building it, but Netflix has sitting here willing to pay, what are you doing? Like, and, yeah, that's the biggest impact. Everyone already knew you should do ads. It's just going to exacerbate the Netflix scale advantage is what I would anticipate. Okay.

Yeah. Well, one day we'll have to come up with a separate power rankings of all the dumbest streamers out there. You look at like Paramount Plus. Paramount Plus and Peacock, they're both insane. They make no sense. They made no sense from day one. What do you think about the Max name? I do have to ask you.

Oh, boy. So I know you came out and defamation. offended it, a typical Zag from Ben here. I don't like it because it just sort of, it underscores what little expectations we have for the masses these days. The idea that HBO would be alienating to mass audiences is just very depressing to me. It may be accurate, but at the same time, I just associate HBO with such high quality that I think I reflectively. We want to say, we want to be accurate. We They want to put a lot of, they want to put a lot of direct on it. I know.

And so they're like, look, it's very important that we market this direct to people and they understand that they're not going to be inundated with quality. I just feel like that's a really grim commentary on where we are as a culture. But again, it could be the right play, you know. And HBO, I am very curious to see how successful Max can be because they're offering it at a pretty high price point. and I don't know that there's that much more of a market for it. I mean, are you bullish on what's possible

there? I do think the reasoning for the Max name and not calling it show Max does make sense. I think part of the allure of the service, why I've been relatively optimistic about it, is I think a streaming service needs to be all things. And especially in the internet,

it needs to be sort of bifurcated. You need the tent pull stuff that gets people in the door and then you need the, you know, discovery sort of content that people watch. If you're going to endure, I think it has to be basically a stand-in for TV itself or cable and everything that people used to get from TV. That's what Netflix is. That's what Max is. And most of these other services, I don't think, can play that game. So if you're not Apple and Amazon with limitless resources, it does feel like the days are numbered.

It's just a question of how many years it's going to take before people give up. Yeah, but I mean, one thing that does concern me about the name is there, it does, there is a bit of shipping the org chart. Like, this is a phrase that, that comes up sometimes, used to come up all the time to come up all the time, they're kind of some Microsoft, where it's like product decisions are being made because it's a reflection of sort of internal dynamics and like, you know, the, whatever, the, the realities of, of the entity that's sort of doing it. And I think there is a bit of that here where it's like, well, we have HBO and Discovery in one company.

So we're like everything flows from like those sort of like internal dynamics, which I think is is certainly a challenge. I mean, I do disagree with people like they should just ship two apps. I think that's a mistake. I think the the challenges of customer acquisition are very large. You know, like just getting one app is is tough. and having you like, oh, you got an HBO up over here, another app over here. And there is a bit where, to your point, the HBO Halo ideally sort of wends itself to the whole service. But I just, a lot of people were mad.

They called it HBO Max in the first place because they're like, you're diluting the HBO brand. And now they're taking HBO. They're like, why are you not using the HBO brand? There is a little bit of a lack of consistency in the takes game, I feel like, about about this move. But I don't know. I mean, it's one of those things where you go back and it's like, we were so clearly in the best structure for everyone, like five, six years ago where make money on TV the first time around and then sell it to Netflix and they can make a lot of money on the second time around. And they just people could, the media companies could not leave well enough below. They just looked at Netflix's stock price.

They're like, why is there multiple so much higher than ours? We could do that. What they're doing is not special. We could copy that. And we've ended up in this situation where they destroyed the cable bundle, right? The cable bundle was declining. It was always going to decline. But what happened over, and this is a point that Matthew Ball made on an interview

sort of a few months ago with me, the cable companies destroyed or the networks destroyed the value of the cable bundle unilaterally. They went in and took out all their best shows. They said, no, go get streaming. And they gave people a reason. It wasn't just like natural churn as people sort of shifted over time. It's like, no, this is no longer worth it to you unless you like sports. That's kind of it.

And I don't know that you can put that back, you know, put the genie back in the box as it were. Like I think I'm mixing a few metaphors there. But yeah. It's all right. I love it. It's going to be a tough few years for sort of everyone, everyone involved. In the long run, probably the best place to. be will be

would be better for everyone if sort of Netflix was sort of the place where they all sold to. Disney's probably in too far. Again, they have to figure out what to do with Hulu and all this sort of stuff. Again, I think, you know, the Disney content, the reason why it's so compelling is you don't need to have that much stuff for

people to still pay you like $15 a month or whatever, right? Especially if you have kids. But now, like, all the costs of Disney are not for the Disney stuff. It's for all the Hulu stuff. And are they actually maxed, like the same questions I just raised for the other streaming service applied to the Hulu part of Disney for sure.

And that's the other aspect about the landscape we have today versus where we were five or six years ago. It sort of feels like everything is more expensive now and everything is worse now than it was in like 2016 or 2017. And that's true for a lot of different sectors that were subsidized by people willing to light money on fire for many years. but at the same time, it would be nice to see more consolidation, if anything, because the quality has really sort of taken a hit over the last couple of years.

And it'd be nice to just pay for a couple streaming services that were really good, as opposed to all of these streaming services that are slowly bleeding content and relying on, like, kind of thin IP that they're trying to generate themselves and isn't really working out even for Netflix. Like their original content adventures have not been particularly, fruitful. Right. This is the interesting thing, right? Netflix's seeming strength right now, still really, to your point, isn't a function of like them becoming HBO fashion than HBO came Netflix, which is sort of like a common catchphrase. No one, I don't think,

thinks Netflix is HBO. I think everyone's given up on that front. It's like that's a empty fear. In some respects, well, that's an affirmation of sort of the structural argument that that has undergirded my Netflix optimism all along. It's still mad like the structure still matters. The fact that they have the larger user base, it still sort of matters, matters most. I do think your overall point, though, is sort of an internet story where just in media, in sector after sector, the middle gets hollowed out, right? And it is kind of a bummer where all the money goes in these super, it is like, you know, it's almost like an analogy like AI or whatever, right? We've read these massive centralized services that are heavily controlling what comes out and, and trying to, you know,

make it say the right thing and not do something that's risque or you're going to have like these scrappy open source sort of like alternatives and like that's kind of like everything right it's like you're going to have open user drainage stuff at one side or you're going to have like we're going to double down on our bets we're going to least fewer movies increasingly expensive yeah and like that whole middle ground like where's our rom-coms right we need the rom I mean it's funny like with AI like one of the things is the whole like romantic connection erotic talk discussion, right? I was talking to the replica COS week, which is that was a story

with them. It's like, yeah, where's the rom-coms? Like, it sort of applies in industry after industry. Robcom's going to get super weird over the next 25 years as AI proliferates. Vid says, I've been a Stratory reader for a while and an avid

listener of the podcast in the Stratory Plus family since they started rolling out. Ben, I've always accepted your thesis that Netflix is missing out on quote, a massive cultural opportunity by not releasing episodes to shows weekly, but I've started to see lots of people in my group chats adopting a similar binging strategy, even with shows that are released on a weekly schedule. In other words, people are waiting to start new seasons only as the penultimate episode or final episode is released and then watching all the episodes

in one or two sittings. So I wonder, do you think Netflix actually does have data to support its release strategy. And alternatively, do you think your thesis was right until Twitter became rubbish and conversations shifted to group chats, making weekly drops less valuable? So, Ben, I'll let you answer Vids question. And also, this is an opportunity to talk about your takes on Netflix's movie release formula here and their refusal to really lean into feeders as a revenue stream. But what do you think? I mean, I think it feels like there's a little bit of Stockholm syndrome here where it's like, oh, actually, this is great because we all coordinate on ourselves of release schedule and do XYZ. It's like the streaming service could do that for you, right?

As like it's been done sort of for all sort of history as far as this goes. I do think, I still think it's a mistake, like especially like the large tent pole shows. And you don't have to treat every show the same. Like when they release like their top chef ripoff like you sure release it as a binge thing right it's not like a cultural happening right like like the but for like you know stranger things which by the way we only ever talk about stranger things that's really the only truly culturally impactful show I think they've you know they've had arguably speaking but yeah that and house of cards right lead into it right and so like not everything has to be treated the same just like not all Netflix customers have you treated the same you can have some that prioritize and value. you, no ads, and you have other ones that don't really care and are a huge market opportunity, right? You're a big company. You're not like a scrappy startup stuck on having to do everything universally

the same way. And so on one hand, yes, I'm sure Netflix does have data that supports their position. It's always sort of like, it's with those companies, you're always a little hesitant to go against because they do seem to often have it right. At the same time, as you said, they opening the podcast. They also could wildly change directions. That's one of their strengths.

and that also affirms the fact that there's room to criticize and give feedback because they might end up going that direction, right? So I'm going to stick with my they should have for certain shows and they should have like a category of shows. These are tent pole shows. These are important shows. We're pretty marketing behind. We want to leverage the entire ecosystem. Like there's an entire world of content creation that is desperate for Netflix to do this so they can write. Like they want to do free marketing for Netflix. They're begging for it.

It's like our only purpose in life is to do free marketing. for, you know, to create cultural moments out of shows that might not be ones. This is like an entire world of media that exists to recap shows and build these hype cycles and everything else. And Netflix just leaves all the money on the table. And to Vids point, I would just say that I have found myself watching shows like that recently, where I will wait for, you know, eight to 10 weeks to where almost all the episodes have been released and maybe catch up in time for a finale or something like that. But you're not really, sacrificing anything if you're releasing the episodes like three at a time or something,

if you're Netflix or Amazon Prime, because there are people like me who are willing to wait regardless and we're going to be there no matter what you do. So if you're looking to like potentially create some cultural momentum, then yeah, trickle them out. And then the people like me are going to be waiting no matter how they release it. Yeah. Again, maybe I'm missing something, but I am emboldened by the ad thing. It seems like a no-brainer. So I am sticking with it, uh, of it. So I think it's the right thing to do.

But I don't know. As far as the movie thing goes, this one, it was sort of interesting. Maybe as being a little generous to Netflix, they're like, look, like other companies are leading into the windowing strategy, which is, you know,

this idea that you're going to, uh, you know, make, like the original winning strategy is first they're available in theaters, then they're available like those cheap theaters that are like, you know, like super cheap way. on that it's like pay-per-view, that it's like on airplanes, that it was like available on

HBO, and then it shows up on like regular TV, right? Like a year down the road or whatever might be. And the idea was you're spending a lot of money on content once and then you can monetize it multiple times over time. And that's exactly what a content company should do. You want to leverage your fixed costs over as many ways as possible. It worked great for 75 years. Right.

And to the point, this is sort of a point that David Alsav has made it. at Warner Brothers Discovery, there's a marketing benefit that comes from it, right? The marketing you spend to promote a movie does not just apply to that movie's release. It's in people's head so that when it does show up in your streaming service, they feel that streaming service is more valuable because they remember the movie marketing. They remember people talking about it. Like, oh, I have a relatively recent movie on my streaming service. This is great.

I'm going to sit down and watch it, do sort of X, Y, Z. Like, it's not, these aren't distinct buckets that don't sort of tip over sort of in other places. that said from the movie perspective, number one, marketing movies does cost a lot of money, right? Like it's like half the cost, right? Like it costs like $100 million to make a movie and $100 million to market it. Like it is really expensive. That's number one. Netflix is not in that game currently.

And I can understand the reticence to sort of get in that game. They have to share a lot of the revenue, like 50% with the theater or whatever might be. So you're making less revenue on that. And there is a bit where I think. there is a real Netflix brand proposition, which is, like, we're the place that you go, like, you can turn on Netflix on a Friday night and there's going to be a new movie for you to watch. And in a way that, you know, there is a something unique and discoverable and interesting. And we have a marketing engine, which is our home screen, right, which is you turn on Netflix

and they can put there what they want. And that's marketing. They sort of get for free. And particularly to the extent that Netflix is trying to do this middle ground, like, like the rom-com ground and like all those like movies that no longer exist in theaters. I think there is some logic to that. It's brand affirming for Netflix. I think in a fairly meaningful way.

And because the marketing cost of movies is so large, like to shift in the windowing direction would be a pretty significant expense that Netflix is to some extent capturing because they sign up for Netflix knowing there will be original movies they haven't seen before on Netflix. It's sort of a, I mean, I'm very fuzzy out or very soft on this point. I'm willing to kind of give them the benefit of the down on this. I think TV shows are different. I think thinking about movies, TV shows in a different way is, is it useful?

And I think the TV stuff because of the sequential nature and the whole industry around writing about it and stuff like that does should be the weekly thing. But for the movie thing, I think they have a decent point and I have no real objection to them doing this way. It arguably might be a better way to do it. So is it related at all to their unwelcome? willingness or reluctance to negotiate deals with talent that require them to share profits that that come from theaters yeah residuals um yeah i mean buying large netflix very early on was like we're not going to pay residuals we're just going to pay more up front and that was sort of pretty disruptive to hollywood where like whether it's a hit or not you're going to get paid um and sort

of the the upside is if it's a bad show i still got paid if it's a good show that you don't sort of benefit from the you know you're not making money from friends 30 years later right like Like Jennifer Anderson is, right? So sitting on an empire. There's way for from a creative perspective, it cuts both ways. As I understand it, I think Netflix does pay more residuals now. Like figuring out how to account for streaming revenue as far as residuals go was part of the recent, you know, negotiation a couple years ago with with the actors guild, I think. So I think it's gotten, it's gotten more fuzzy.

And I think in general, there's more convergence in a more of a shared model. although I do think Netflix, by and large, is relatively allergic to residuals. And it makes sense, right? Their whole model is pay, like, it's over the content and then you can make money over time. Yeah, it doesn't work that well. But that said, obviously it's important to actors. You could certainly understand the allure. So I think they've had to give ground a bit.

Okay. But the reason I ask is just because culturally, it does make sense to drive everybody to Netflix, and they've always emphasized that. And at the same time, culturally, they've always been reluctant to pay those back-end fees. And so it all adds up to not really wanting to put too much emphasis on theaters unless they're trying to push like Oscar fare. Right.

Just to double down on the movies versus TV thing, right? With movies, even with TVs, there is value in recently it at once doing the binge release. My argument is there's more value from dripping it out. With movies, you're not. going to have a 10-week build-up of, like, people writing about your show. It's a, it's a one-time event no matter what. So I think the benefits of doing it traditional way are much less than in TV, and the, the benefits of doing it their way still remain, right? Like, releasing a movie, just you have a new movie in Netflix on Friday night, that is similar to, oh, release a TV show,

you could binge the whole thing over the weekend. You're getting, like, it's just the tradeoff, I think, is different for movies versus TV. And so I, I'm more amenable to their approach for movies for that reason. Because I think they're foregoing less by doing it there. No respect for the magic of a movie theater. I'll also just note for the record, the windowing strategy industry-wide, as far as movies are concerned, is in a really strange place. Because over the last like six months or so,

there have been so many movies that hit theaters do fairly well. And then like two or three weeks later are on a streaming service. And at that point, like, audiences have no incentive to go and pay $20 a ticket to see a movie in theater. If they can just assume that it's going to show up on streaming services a couple weeks later. Like, you need to go in one direction or the other. This sort of trying to have it both ways is it's bizarre. Like, who thinks that's a good idea? All right.

So now for something completely different, let's talk about Nvidia. Ben says, I'm not sure how to think about Nvidia long term. My understanding is that they mostly make the GPUs and tied-in software to power most everything interesting on the internet along with some seriously high margins. Microsoft, at least, is looking to develop their own chips to get out of having to pay that margin. Is developing the silicon and something like Kuda actually hard enough to make hyperscalers pay up? or is the allure of capturing that extra margin too tempting, and industrial GPUs will become something of a commodity over time? Or am I just hopelessly backward on all of this?

And I have a related question, which is that as AI proliferates, I'm wondering whether Nvidia is going to become the most powerful company on Earth. So you can take either question, Ben, and take it in any direction you want. Yeah, for Ben, I think it's a little bit of column A, a little bit of column B. in that he's right to be uncertain, and also he is a little hopelessly backwards

on a couple of the specific details. So like specifically, the internet runs on CPUs, like traditional processors that are mostly made by Intel, but AMD has been making sort of large progress in that space. And meanwhile, the hyperscalorosur is trying to build their own arm processors to sort of replace the whole thing. Like Amazon in particular has made the most investment there.

So most computing is on CPUs, most servers are on CPUs. So the sort of like powering the internet is not in video. Okay. So GPUs are used for this AI processing. And they're used on both sides, both for training and also for inference, which is where you're actually like someone makes a request and you're sort of running that through the model. And you run it through the model, you want to be massively paralyzable.

So basically, CPUs can handle much more complicated tasks. They're a general purpose processor. And a GPU is tons and tons of very simple processors. So if the task is very simple and you can run multiple things at once, then they're good at that, right? So in the context of AI,

you're simultaneously, like say you're predicting the next word, you're running tons of predictions simultaneously and then picking the best one, right? Like that, so it sort of went stuff to it. And it's downstream from video graphics where you can cut a screen up into an infinite number of tiny squares and render each of those squares individually.

And so it scales perfectly, the bigger the chip, the faster the graphics processing because it's called embarrassingly parallel. So, Nvidia is rising in sort of prominence because of AI, but the internet still runs on CPUs to be clear.

Secondly, Microsoft is not new to be doing this GPU thing. Google has had their own AI processor for years called the tensor processing unit, And so they were first in this game. Facebook's been building their own for a while. Microsoft's been building their own. You know, there's multiple startups in the space

that have been around for several years that are building AI-specific processors. So the Microsoft news was interesting and that they're now in the game, but I think they've been in the game for a while. This might have been rumor for a while. And of course, everyone's in the game. And AMD is obviously the other competitor here

because where Ben is right. And sorry, no problem, Ben. The details wrong is useful for me to spout on them. He is right about this, the overall importance of this question and what's going to be meaningful and does invidia have sustainable long-term differentiation in the long run? Because, yes, they do charge high margins. And yes, everyone wants to get out from sort of underneath that. So you have a very like, so one example is, is Facebook, you know, there's a big open source or project for doing AI called Pi Torch. And they've been trying to optimize it to run on any,

any GPU, right? And Facebook wants to work closely to say AMD to get it working well on their processors, particularly for sort of inference. But then they realize, well, then AMB just raises their prices, right? The challenge right now is there is not enough GPUs. There's not enough data centers for GPUs because these GPUs run super hot. And they have to be sort of like, like the rackings different. And they require different cooling solutions. And so it's not just there's of GPUs is also a question of where to put all the GPUs. It's sort of an open question. And that's being solved. But it's kind of hard to know for sure, you know, yes, Nvidia gets a premium, but how much that premium is a lack of overall GPUs and how much

that premium is a lack of Nvidia GPUs sort of specifically. And there is tremendous, massive, to Ben's point, market pressure to break sort of the Nvidia lock for there to be sort of anyone sort of available. That's why Nvidia's who, software layer is so important because like if you want to hire someone that can program this stuff, it's going to be way easier to find someone that's familiar with CUDA than is familiar with sort of other tools or whatever it might be. Now, by and large, I think I don't really know the answer to this question because there's been a shift sort of abstraction up the stack a fair bit where, you know, particularly if Nvidia, I think had a bet that there would be a lot of

companies built on top of Kuda, when most of the companies are probably going to be built on top of large language models, right? So you're not going to build on top of Kuda. You're going to build on top of the open AI API, right? That's a lever of abstraction sort of higher. That's where the big sort of explosion is going to be. And as far as product companies probably are concerned, and that's a little concerning for Nvidia, because if you're separated from that layer, the,

ability of the intervening layer to sort of abstract you away and then ideally make it so we can run Nvidia or we can build our own chips. It's sort of immaterial. But they're not there yet, right? They're not there yet. And it's not. Everything's running on Nvidia. Right. But I do think one of the interesting takeaways from what Nvidia announced at their last, you know, I talked about this a bit with Jensen Huang is they're really focusing on enterprise customers, right? And bringing in the Nvidia stack and you building your, own sort of AI business. Like you don't want to be on the commodity layer and giving your data open AI and Microsoft. You want to build your own sort of thing.

That is an appropriate strategic response for a company that is concerned about being commoditized. What we know, like why does Intel, for example, the strongest, most durable part of Intel's business is selling to large enterprises and the federal government. And they're selling to them where they are, they have their own data centers. Why? Because once you get, once you're locked on in these bureaucratic organizations, once you're locked on to a certain thing, you're just going to stay there. You're not going to, yeah, AMD, in this case, AMD is almost a drop in substitute, but you have to recertify it and refigure out how to do it. It's just easier to stick with Intel, right?

Whereas the big companies, where it's like the hyperscalers, they will gladly switch back and forth to Intel AMD because they have the motivation and the scale to really push on, we want to. really commoditize our inputs here. Whereas if you're a fortune a owner of a company, yeah, you're big, but it's not worth it in the big picture to make sure you can choose between AMD and Intel. You just buy Intel. Yeah, maybe it costs you a little bit more, but it's not worth the trouble of sort of truly abstracting away everything. It's like your analogy where you're trying to redirect a ship when you're a company that

big. Like, particularly if you're the federal government, like trying to get everybody out a completely different operating system is just a complete non-star. Right. That's the crazy thing is Intel A&B are not really that different at all, particularly when it comes to X86 sort of server processors. And yet, no one's going to bother, right? So AMD's real growth opportunities in the hypers, where they will put it in the worth, it is whatever performance advantage A&B has, they will realize it because they're at such scale. And, you know, the price really matters and things like that. So I see NVIDIA's focus on the

enterprise as a recognition from Nvidia that in the long run, there are very strong forces seeking to commoditize us. And we can't just be selling GPUs to Amazon and Facebook and Microsoft in the long run. Also, we're, you know, our margins are going to get ground down. And so I do think now, that said, everyone's been predicting Nvidia sort of getting commoditized away for ages and it hasn't happened yet.

So I think it's definitely a bit of an open question. But the real takeaway, and this is something I've been observing for a long time, it's TSMC is the real linchpin here. Because they're the ones that are making, if Microsoft comes out with this chip, TSM will make it. TSM makes the Google chip. They make the Nvidia chips. They make the M-MD chips, right? That's what I've been wondering. I was going to say, like, is NVIDIA side by side with TSMC as the most indispensable company for the tech industry?

but it sounds like TSM is in sort of a tier of its own. Yeah, I think if Nvidia went away, there would be a slowdown because I think Kuda, like people can play about Kuta. It's really hard to use. It is hard to use. Programming without Kuta is even harder, right? So like this idea where you have to build a program

that runs across all these chips. And AMD today is not just doing chips. They're shipping whole systems, right? And getting all those different things to work together is really, really challenging. and InVDia does all that work. That said, as far as if one company disappeared tomorrow, where would we be? We'd be much worse shape if TSM went away than if.

Oh, no, I know that for sure. What I'm saying is like as we project forward and AI is central to literally every tech business there is in five years. To replace TSM, you're spending hundreds of billions, if not trillions of dollars. To replace Nvidia, like everyone's trying right now, it's sort of. speaks it sort of makes the point. We'll see how it goes. No one's trying to build a TSM competitor except for Intel because they basically have no other choice if they want to continue to survive. So if your metric is what degree of desperation is needed by a would-be competitor to compete

with us? TSM is a very strong vote. Okay, okay. So looking ahead then, I guess the question is whether there's competition for Nvidia five years from now and more diversity in the market. Yeah, I mean, obviously, AMD with their graphic chips is a sort of direct competitor in a similar way they are to Intel, they don't have a processor advantage because both Nvidia and AMD are on TSM, right?

Like, what of AMD's advantages versus Intel has been, they had the TSM advantage. And Intel is actually trying to negate that by starting to build more stuff with TSMC also. So there's two ways to compete as far as chips go. It's the actual architecture of your chip, like are you faster or not? but what is more important, what dwarfs that is your process faster. And so, Nvidia relative to AMD, I think, and AMD by and large has looked to, they haven't, like, once their chips get better, they raise prices, right? They're not looking to necessarily, like, sort of, in part because they're gated by their production

by what they, by TSMC. They can't just, like, scale up fab, sort of infinitely. The real long-term threat to Nvidia is companies building their own. But what Nvidia will argue is like where is it worth actually investing, right? Are you going to get more gains? Like if you build your own, you're stuck with your own, right? This is a bit I always, an argument I always talked about as far as SaaS companies in Amazon, you know, because Amazon AWS is not cheaper, right?

Like particularly on a sort of lifetime value, yeah, it's cheaper up front. But in the long run, what are you actually paying for? Your cost as a SaaS company isn't just building your own data. center, it's all the effort you spent on building a data center instead of improving your product, right? Like, that's sort of like the content companies. It's not just the cost of building their streaming service. It's the cost.

It's the revenue they're not getting from going with, it's the opportunity cost. Yeah, putting that money into content and everything else. Right. So the question is like at what level scale do you need to be where making, because there's benefits from building your own. You can have, you can build your software stack specifically for your chip and be sort of fully integrated. But is there a cost to just?

letting Nvidia take care of the innovation and you sort of focus on your differentiation. I think this is a very compelling argument again for companies that are not tech companies, but for the hyperscalor specifically for the Amazon's and Microsofts and Facebooks and Google's of the world, it still probably makes sense to build their own. And this is just a question, you know, can Nvidia sort of stay ahead? And that's probably for now, yes, in the long run. This is the limiting factor on Nvidia's valuation. It's like it is this question.

How long can they sort of maintain this advantage, particularly when the market forces driving competition for them are so huge. And their moat is not having spent trillions of dollars on fab equipment. Okay. And are we in the dark, essentially, as far as like the progress at some of the hyper-scalers and Google and how successful- the study saying that, look, our chips are much faster than Nvidia's, but they were comparing it to Nvidia's A100, which is already, you know, is already sort of last generation.

The H100 is sort of the new one. And they're like, hey, we can only compare it to what's available. It's like, you know, you just wrote this for the headlines. You know, can we do a real comparison here? But, you know, like Facebook. Are they succeeding with their internal projects? Or are we basically just totally guessing at this point? Yeah, don't know.

like how is this Microsoft chip going to do? And, you know, can you get yourself in sort of a dead end? Because like there's a degree to which an Nvidia chip is more general purpose than a, like, if you build a chip that's perfectly tuned to your software stack, that will be faster with that specific job it was tuned for. But then when you try to do something else, like, oh, man. Like, it's like building a, you know, we talk about a cruise ship like a bullet train, right? invidia in part thanks to kuda is a bit more general purpose than purely custom built sort of things but um this is going to be a real i don't really know the answer to be honest like i i would say definitively one way or the other but it this is the outline of the of the question for sure okay well um the answer to my question is invidia going to be the most powerful company on earth is no we can be firm on that it sounds like tsmc is going to be the most powerful company on earth well But it's funny how you just go further levels, you go further down the stack,

stuff gets more problematic. The issue with TSMC is in the fab equipment, it's the ground on which the fab equipment exists, which is Taiwan. And so that's the gaining factor. It's a significant one, I think, on TSMC's valuation. Yep. And the early returns on the American TSMFAPs

haven't been that encouraging, but we can talk about that another day. A couple more questions at the end here. Some follow-ups to the AI conversation. Rock says, I agree with Ben that as a general rule, I think the liability would attach to the user who created the output using the AI as opposed to the company that trained the AI. That said, I took Adobe's argument to mean that the AI tools they were referencing wouldn't be commercially viable because their users would get sued for infringement six ways from Sunday for creating infringing works and therefore the tools wouldn't succeed. not that the people who train the AI would be liable for users' actions.

And we'll get to the second point in a minute, but I do appreciate this note from Rock because I was mad at myself for being unclear on the last show. I didn't mean to imply that AI companies would be held liable for individual outputs that users create. But what I was saying is in cases where AI models are allowing users to regularly generate outputs

that are competing with the copy material that all these models are trained on. My guess is that in those cases, that would change the way a court addresses the fair use issue on the training side and the copyrighted material that's being used for training. And my other guess is that copyright holders will end up deciding that it's more efficient to attack the problem that way than try to sue thousands of people every month. But we'll see. We'll have to wait and see how it evolves over the next few months. and years. And as to the second aspect there, the Adobe part of the question from Rock, we also got this note from Alex who says, in the last episode, Ben focused on deliberate

infringement. And I agree with him. In these cases, it's the user and not the tool that's at issue. However, in Ben's Adobe interview, their focus seemed to be more on accidental infringement, which does seem to be something that needs to be addressed at a tool level. If I'm launching a multi-million dollar advertising campaign and my advertising agency uses an AI image generator to create a quote cartoon kid in a yellow jumper, I need to be pretty certain it won't come back with a drawing similar to Charlie Brown. There was a recent case in the UK where a children's author sued over a cartoon dragon used in an advertisement. In that case, the defendants could show the books they'd looked at as references and the steps they'd taken to create the character. If they'd used, if they'd used to,

used an AI tool in the process, how could they demonstrate the similarities were not a result of copying? I think this is the challenge Adobe is trying to address. What do you think, Ben? I think that's a great, great articulation. And I think that's exactly right. Like, if Adobe says all our inputs are safe, then you don't need to double check the outputs. I think that that's an excellent articulation of the value they're pushing on. And probably one I should have given more credence to instead of, you know, be, you know, be grumpy about about the framing.

So I think a very excellent note from Alex. Okay. A couple more notes. David says, not sure if it's an insult to Google or a compliment that Andrew thinks Huli is supposed to be Amazon. I appreciate the correction. This is tough because A, Ben has not seen Silicon Valley, so he wasn't able to correct me. but Gavin Belson, the CEO of Huli, felt more like a Bezos parody than anything related to Sergey or Larry.

But yes, Huli is obviously a Google analog to anyone who's seen the show. I mean, just look at the name, right? Yeah, it's true. It's true. It's okay. You just had a kid. You're tired.

It's fine. You're right, I was not there to hold you accountable. I will share part of the blame. I probably should have watched Silicon Valley at some point. given I write about Silicon Valley. So we can take the all together. If we're ever stranded on an island together, I promise you that I'm going to make Ben sit through

each season of Silicon Valley. I haven't seen the show in several years. Although I did watch a couple episodes to prep for Sharp Tech back over the summer. Jeremy says, Hey, Ben, I've read you since the first sailboat touchscreen sketch and binge lately. I'm writing for the first time to say I'm both sorry

and furious about Janus's lower back contusion. Can we just hit rewind on the playoffs, please? I really hate this. Ben, I was not genuinely concerned about the bucks until I saw this note from Jeremy. Like, even after game three, I was thinking, okay, no big deal, Janus comes back. But big picture, we are officially getting concerned notes

from Sharp Tech listeners. It's a real red flag in my eyes. How are you handling all this? Not well. I would like to apologize and repent to the context of the church of the church of Janus and apologize to Janus. For any comments or statements I made that were in any way sort of overly affirmative in say his teammates, all-star candidates, end of season rewards, any grace I gave Mike Booneholzer. this is all the success, everything that the bucks are, that makes the bucks good. The reason why his teammates look like stars is because of Janus.

It's all Janus. All that praise just serves to deflect from the fact that Janus is the best player in the league. And people should be outraged. There should be more discussion at the fact that these playoffs are fundamentally compromised by the fact the clear cut best player in the league who would be universally acclaimed as such if you were in a larger market is missing. And it's absurd that we have to hear about, oh, injury title, X, Y, Z, and so many other contexts when we're setting up for the biggest asterist of all time that no one will know, except for sharp tech.

An asterisk so large that you can see it from outer space. That's what will be required if the bucks don't get a healthy, honest back. They play the heat Monday night in Miami. We've now closed out several episodes, the last two episodes. this will be our third in a row praying for Janus's tailbone. Everyone out there, keep your fingers crossed. And yes, I appreciate you repenting for your Brooke Lopez boosterism in the defensive

player of the year. For the record, I was, I was not hardcore about it because I knew that, of course the Buck's defense is good because of Janus. But I did engage in a little bit and sort of, I was irritated that Evan Mowgli was getting votes when it's like, Evan Mowgli is like a poor man's Janus.

Like if we're going to. give the vote to this sort of player who's next to a defensive center, then just give it to Yannis. Give it to the right guy. Right. So even my brief Lopez boosterism was actually a pro-Yonis argument, but I should have just been straightforward about it, and I'm sorry for that.

All right. Well, hopefully all of this is rendered moot and Milwaukee comes back and kicks Miami's ass. But for now, Ben, I look forward to coming back later in the week. I mean, everyone's taking strays. at this point. Jimmy Butler is given Drew Holiday 30, 35 points of games. Ful's the best defender in the game. Yadis guard is Jimmy Butler. He barely can crack his team. It's true. He gets... Jimmy Butler, he looked like he was on the way out of the league after two

weeks with Yonis a couple years ago. Brin Forbes outscored Jimmy Butler when Yonis was guarding Jimmy Butler in the series two years ago. Let's get things straight. Yadis is the best defender by far. Right. And the heat have like two good players and are somehow finding a way. And they're losing players and their bucks are still. losing because Mike bull, I mean, look, as long as I'm branting, what was so upsetting about game three is the way Miami won is the bucks, like Drew Holiday, I do love him, but he's not really a traditional point guard. You pressure him, he will get his own shot.

He's not going to like pick apart the defense, right? So even in our playoff run, our real point guard is really Chris Middleton. Ever since Chris Middleton came back from his injury, he hasn't been able to dribble. It's kind of been an issue all year. So the heat get all up in our ball handwors. and they have these terrible defensive players on the floor, which we can't pick on because they're all up in our ball handlers, right? The reason this is so infuriating is, number one, it's a good strategy.

And number two, this was the strategy I was going crazy about last year the Bucks should have done against the Celtics. They didn't have good ball handlers. They were picking on our certain players who didn't have enough good defensive wings. And Bud should have been pressuring ball handlers to get turnovers and to stop the Celtics from being able to just, leisurely walked the ball up the floor, isolate Jason Tatum on Grace and Ellen,

and watch him score every single time. We are watching a demonstration of the strategy that the buck should have pursued last year that was going nuts over that would have resulted in another Yonnas title. Well, we also saw a demonstration of that strategy in last year's NBA finals when the Warriors started pressuring the Celtics.

This terrible Miami team took Boston to seven. Like, like, it, coach butt is terrible. He builds a good system. No, no, he's not terrible. He builds a good system. Like good fundamentals. The team's disciplined.

But the in-game reaction and adjustment just isn't there. And it drives me freaking bonkers. Well, again, hopefully Janus can come back to save the day and make everyone look like 40% better than they actually are. Let's not forget what actually makes these people look good. Yeah, it's right. It's all about Janus. That's the theme of today's podcast. And on that note,

Email at sharptech.fm if you'd like to hear us discuss more bucks takes or anything else related to technology and business. This has been fun, Ben, and let's come back Thursday. Later.