Part of the plan. Thank you for de-queuing the music. I'm Michael Grimes from Morgan Stanley. Elon Musk needs no introduction. We just flashed up the forward-looking statement disclosure. Please refer to Twitter's forward-looking statement disclosures, Morgan Stanley Research Conference website. Please refer to Tesla's public disclosures from their Investor Day, 'cause there'll be a little bit about Tesla here today. With that, why don't we go ahead and bring you Elon Musk direct? He doesn't get a lot of chances to speak directly to investors for this private company, Twitter, we're gonna do that with you directly here today. First is the mission, if you wanna call it that. This is what. Mm-hmm W e've heard you say, promote, protect public conversation, town square of the internet, how would you describe it, the mission of Twitter 2.0 under your ownership, Elon? The goal is to have Twitter be the best source of truth, the most timely and accurate source of truth, even if the truth is something we don't wanna hear or unpleasant or whatever, but have it be timely and accurate and really just where you can really understand what's going on. Not just is something true or not, but what is the narrative? The thing that I think most people don't, well, probably you do, but most people don't quite appreciate is that the media controls the narrative. There are many things that could be talked about in the world, but only a few things can fit on the front page. Twitter doesn't have a limitation like that, the public can control the narrative and the public can inform themselves as to what the narrative should be. That's a really big deal. It's a forum for it's called citizen journalism or for the public to get together and communicate in a way that was never possible before, so you can really know what's going on. I think that's really gonna be essential for a functioning democracy. If you say, like, what is the bedrock of a functioning democracy, it has to be free speech, and a level playing field. That's why it's the First Amendment. It was the first thing they did. Like, we've gotta make sure freedom of speech. Why did they do that? It's because where they came from, there wasn't freedom of speech. Once you lose freedom of speech, you don't get it back. That's why we must protect it at all costs. Thank you. On the core principles, if we wanna call them that, you just went over citizen journalism, power to the people, and a level playing field. That's what you mean by democratic. We've also heard you say core principles of authentic, informative, and entertaining. Yeah A ccurate, you just went over, and brand safe. Where are you on these? Yeah. Yeah, I mean, I think the, some of these are, you know, a little at odds. You know, brand safe I think really means, like, where advertising is displayed, that the advertiser gets to choose what material is near that advertising. If it's something, you know, if it's some sort of like a, like a train accident or a, or a war scene, then probably a family friendly brand is not gonna wanna advertise right next to that, you know? You know, it can't be like, "Here's a bleak war scene. Would you like to buy a hamburger?" would be, like, awkward, you know? That's understandable. You wanna put advertising next to content where it makes sense. The content in general needs to be authentic and informative, even if it is controversial or jarring. Yeah. I think people need to be able to choose, you know, to some degree, what content they want to see, and of course on Twitter you can. Really fundamentally it's a place you go to to learn what's going on and get the real story. That's what, the truth, the whole truth, and it's gonna be more than I would like to say nothing but the truth. That's hard. There's gonna be a lot of BS there. Truth and more. There's gonna be lies, for sure. But you wanna have the truth. You wanna bubble up the truth and be able to sort of sort out. Really want truth with the least amount of error, so. You've said, you know, Twitter is alive. Another way that's described is it's where news happens, as you just said. A couple here, Patrick Mahomes and Steph Curry talking to each other, Rihanna speaking directly to her fans, or in crypto, SBF and CZ. Yeah The conversation directly, on Twitter. Sam Altman Sounds wild. Sam Altman, you know, launched ChatGPT, announces it on Twitter. Paul Graham takes up a tweet on it. You're tweeting on it. I think 50,000 tweets happened there before it was picked up by the mainstream media, just in the first couple of days. Yeah. You know, talk to us about how can the media catch- up? If they're on Twitter, they can be seeing these tweets and posting. Yeah. If they're not on Twitter, is there no way to catch- up? Like, how does the traditional media and citizen journalism and 50,000 tweets on an important subject, how do those intersect, and how does the media catch- up? Well, just as I'm sure many of you use Twitter's happening in real time. If you contrast that to what's happening in a newspaper, they have to learn the information, propose an article to their editor, get it approved, write the article, get it edited, figure out which day it's gonna get published on. The thing that happened is being reported on at, you know, three, four days, sometimes a week late. If it happens on a weekend, then it's like, yeah, at least three days type of thing. You know, ChatGPT was huge news for several days on Twitter before there was any news articles about it in major publications. I think especially if one is thinking about investing in things, you wanna have information that is as timely and accurate as possible. There's no better source than Twitter for that. That's certainly been my experience. Paul Graham, who's a legend in tech, a legend on Twitter, I think he has tweeted that he's 80% left-leaning probably if he had to put him on the scale. Not someone, you know, who's not deemed and revered by all sides of the spectrum. He has said, you know, consistent with your accuracy point, he has said how important accuracy is, and on Twitter he said, you know, you should be default skeptical of any news story about Twitter, and assume it's default wrong because not only do some journalists have the agenda, but the source has the agenda, and it's so easy to go through the chain of inaccuracy or, you know, outright falsehood. Is he right about this? You know, if so, how can accuracy about Twitter and about you be conducted by the traditional media, if at all? Does a PR department help? No You famously don't have. PR departments, I, you know, the right name for PR is propaganda. I also, you know, like maybe we should have like a VP of propaganda. I think just more honest, you know? Also a VP of witchcraft. I've always thought You know, those would be two great ones. I mean, the thing that I mean, if you pick up any given newspaper and go through and read the whole thing and say, "How many of those stories are positive about anything at all?" Almost none. If something is newsworthy, it is going to have a negative slant, whether it is positive or not. There's like something in journalism that they've been trained to basically never write a positive story about anything. Once in a while you see a puff piece, it's rare. Anything that's newsworthy will get written about. Anything that's written about will go through a negativity lens, you therefore have a bizarrely negative view of the world if you draw your information from newspapers. This is simply, I mean, fact. On Twitter, you can get a much more balanced positive, negative situation. It doesn't have that bias quite as much. It's probably still a little bit of negativity bias, but much less so. It, you know, I think it's really... I'm not sure what the legacy media does, I mean, at this point really. Twitter is, by the way, the number one news app in the world. In terms of what people download for news, it's number one. There's 500 million active users, 250 million daily users, of which I'd say there's probably, you know, 180 million what I would call significant daily users, where they're. It's a meaningful amount of time. Like the average amount of time that people spend on Twitter per day of that 250 million is around half an hour or so. what we have is. The thing that's like I think most interesting is there's about 120 million-130 million hours of human attention per day on Twitter, every single day on average. Which comes to a really interesting point, which is to just. It's startling how poorly monetized that is because you have to say like how valuable is that attention? Call it 130 million hours of human attention per day of people that read. These are the generally the smartest people in the world, the most influential people in the world, and you have 130 million hours of their time per day. That's a lot. Currently Twitter makes about $0.05 or $0.06 per hour of that time. I think this is poorly monetized. Paul Graham's half hour of attention is, you know, worth more than that. Yeah. I mean, if I'm spending two hours a day on Twitter, like whatever ads are coming through are getting my attention. Right Getting your attention, getting everyone in the room's attention, your time is incredibly valuable. Now the thing is we need to actually serve ads that are relevant and useful. I think as we do that, we can probably at least get it to like $0.15 an hour, get to $0.20 an hour. Yeah, $0.25. Right. I think the actual potential here for Twitter revenue is gigantic. It's gonna be a win-win situation, which is that if you are served advertising that you find timely and relevant, that is with products and services that are useful to you, that's good for you and good for the advertiser. Advertising in the limit of relevance is content. In the next theme that's out there that's a often an inaccurate narrative is that you're indifferent to misinformation or other narratives get put out in the media that you actually want misinformation on Twitter when, as you've just stated, you want it to be the most accurate in the world. Community Notes, I'm gonna breeze through a couple of these before asking you about them. Community Notes are the fact-checking by power of the people on Twitter. Yeah. Correcting government misstatements from the White House. Correcting candidate misstatement from failed gubernatorial candidate, Kari Lake, that Ron DeSantis had been endorsed by George Soros. Adding context while the spy balloon was over Billings, Montana to a tweet that could have been taken to spiral and mean it had been shot down over Billings with that. Mm-hmm Photo and adding context to that. Preventing a conspiracy theory of train derailments are purposeful or nefarious as opposed to 1,700 per year, quickly corrected before a right-wing conspiracy theory could get going. Company misstatements or statements, misstatements about companies, I should say. The tweet that said, you know, Google set a badge machine to tell you if you were terminated or not, and adding context that no, they had been emailed. On Coinbase, that they'd been told they had to stop staking by regulators. Fact-checked very, very quickly. How can this scale, this neutral fact-checking, which seems elusive in fact-checking, but Twitter's seems to have a handle on it. Yeah. How does it scale, and how does it avoid being hijacked by either side of a, an issue or political spectrum, trying to hijack Community Notes. Sure have it be neutral? Yeah. There's a white paper on Community Notes that I recommend reading. In fact, I'll tweet it out so that people can have easy access to it. It's really quite a clever idea. It takes the viewpoint of someone. Think of it in a way like PageRank for pages as applied to people, which is that as people build credibility in how they review notes, they build up enough credibility to actually write notes. Those notes are then rated by others, and depending upon the credibility of the people rating your note, your credibility score gets affected. It's sort of like a collection of credibility- PageRank. PageRank. It's PageRank for credibility. There were link farms created in faraway places to spam PageRank, so that's gonna be attempted, but you're gonna apply. Yes. Software skills and things to prevent that. In order to be a notes contributor, you have to be a verified person. I see. Yeah. You have to have. It takes a while to get. When you just start out, you will start off with no credibility score. Yeah. Um- Very hard to link farm. Yes. When we actively look at any attempts to game the system and shut them down and remove them from the system if they're determined to be not real people or if they seem to be brigading. There are deliberate attempts to manipulate Community Notes. We also make the Community Notes source code is open and available, so you can basically see everything. That's in the white paper or linked? It's not in the white paper, but it's available on the internet. Okay. It's available. Yeah, you can search for it. You can see exactly how Community Notes is calculating things, what changes are made to Community Notes, we'll keep iterating. The goal is to have truth with the least amount of error. You know, there's always this, like, what is truth? I think like one way to sort of, for example, say like if someone really aspired to the truth, if they really aspired to the truth, they must acknowledge that there is some probability that what they think is untrue. If somebody thinks there's no probability, 0% that what they think. If somebody thinks that they They're trying to claim to you that what they're saying is true with 100% probability, there's a 100% probability they are lying. Truth with acknowledged error, where you aim to minimize that error over time. That's what Community Notes is. I think also, once someone gets Community Noted, they think twice about, being deceptive in the future. You, you start getting noted a few times, people are like, "Uh-oh. I know someone who's been Community Noted. Yeah. this... No one's immune, meaning- Yeah. you've been Community Noted- -on this, CNN, fake chyron. CNN made. Yes no such claim, that, you know, free speech on Twitter by allowing people to speak freely. Yeah. You were quickly Community N oted on a satirical tweet, and community said, "This is not, this is not accurate. Yeah Is this the one you tweeted, "Community Notes for the win"? I think you may have. It might've been- Yeah. I mean, the important thing is that anyone can be noted, including me, and in fact, I wanted to make a note of being noted. That, the point is that if I can be noted, anyone can be noted, including advertisers. We've had a few cases where the advertising wasn't accurate, and it got noted. Oh, the mortgage one. Loan money to yourself, I think I saw that. Loan money to yourself, yeah, sure. Right. It's your own customer, and they can be noted. Yes. presumably change the ad. Yes. I mean, this will, I think, be very helpful in truth in advertising. Okay. I can't emphasize enough the goal is rigorous pursuit of the truth, aspirationally the whole truth, and the least amount of untruth. Yeah. Let's talk back about brand safety. We'll go through two quick ones here. One narrative that's out there that probably affected advertisers and agencies was Twitter was going to become this hotbed of hate speech. Shortly after the acquisition, there was this bot attack you've talked about- Mm-hmm that seemed designed to try to make that a self-fulfilling prophecy actually until it was quickly, defeated by your team. There's 50% less hate speech according to what your team has, put out... Yeah this is a graphic, than pre-acquisition. Not only is it a priority for you, but less, half as much as before. On child sexual exploitation, something that's been so important to you... Yeah 800,000 suspended accounts. Mm-hmm. That's four times more than in any month, of the prior year, of your ownership, and a 99% reduction, in successful searches for CSE patterns. Talk to us about how much of a priority this is and how successful you've been at it, but then also presumably more can be done. I've repeatedly said to the trust and safety team at Twitter that the number one priority, which will always be the number one priority no matter what, is ensuring that children are safe on Twitter, that there's no child exploitation. That is number one priority always and forever. And what I've been told is that we've done more to eliminate this on Twitter in the last four months than has been done in the last 10 years, and it will continue to be the number one priority. You know, a 100-fold reduction in CSE search patterns is pretty gigantic, to say the least. So, yeah, it's absolute number one. On, continued on brand safety, you know, this Global Alliance for Responsible Media's brand safety floor, the GARM, you know, Twitter business had tweeted out earlier in the year that in testing, 99%, at least 99% of measured ad impressions appeared on content that exceeded that floor. you know, what methods. Yeah Do you have for your advertisers when you're talking to a CMO, to the head of an agency, just in total? I think you touched on this a little, if there's anything to add, but brand safety is in fact a top priority with a team all over it. Yeah And technology and humans focused on the subject. As I said, I mean, with respect to brand safety, it really depends a lot on the brand. I mean, if you've got sort of, you know, Disney is a major advertiser on Twitter worldwide, one of our biggest advertisers. Apple's one of our biggest advertisers. Disney of course wants to, does not wanna have their ads next to things that aren't appropriate for a family audience. There are other products that are kind of more, you know, R-rated, if you will. They're more comfortable with being, their advertising being, in the equivalent of like an R-rated movie or something like that. Brand safety varies depending upon what brand you're talking about. Is it a family brand or less family brand? Advertisers can actually adjust, how much what content they're comfortable having their advertising appear next to. Same is true on TV. The, you know, the advertising that you'll see at, say, 7:00 P.M. is different from the advertising that you'll see at midnight. We have the same functionality on Twitter. It's really up to the advertiser what where they wanna put their content. I think by far the most important thing is that the advertising is effective, that it is relevant, and that it moves the needle for a company. The advertising relevance is the most gigantic thing. This, this is gonna sound totally bizarre, but Twitter did not consider relevance in advertising until three months ago. In fact, if you've used Twitter for a long time, which probably many of you have, you should say like, how many products have you bought off Twitter? Probably zero. Judging by the laughter, probably zero. Your time is incredibly valuable. Not-A-Flamethrower? Know anybody at Not-A-Flamethrower? Well, I mean, it's possible they might have bought things from, you know, content-based tweets because the, the content that's recommended is reasonably relevant, but the advertising has not been. As we've moved to shift towards the advertising being relevant and timely, Really, like I said, advertising that is relevant and timely is information, it is content. The, the value of the time of 130 million person-hours of the smartest people on Earth is insanely valuable. Frankly, historically, with the advertising being mostly irrelevant in the past, we've been wasting people's time, and that's not good. Going forward, Twitter will have very relevant, useful advertising. Because it is useful, because it is relevant, there will be a massive increase in revenue because it is now useful. I'm very optimistic about the future. It's been a very difficult four months, but I'm optimistic about the future. Appreciate that. Let's get back to a democratic platform for all. The, you know, you've said there's no permanent suspensions of anyone on the left. You've brought back both left and right. I think what you've tweeted is you had to unban a lot more on the right because that's who was banned. Yeah For the most part. Your goal is equally unbanned, complete equal level playing field for all sides of the political spectrum? Yeah. I mean, the objective reality for anyone looking at Twitter for the longest time was that Twitter had a massive thumb on the scale on the left side. Twitter would ban and suspend accounts on the right at 10 times more than on the left. That's like naturally what you'd expect, frankly, 'cause where are we? We're in San Francisco, which is deep, deep blue. Twitter was controlled by the far left. The natural thing that would happen then is suppression of moderates, not just on the suppression of the right, but even suppression of moderate voices. That's not conducive to a healthy national dialogue. In order to have a healthy national dialogue, you have to represent the whole country, and you have to represent, you know, everyone in other countries too. You know, that's the only way to have a town square. Yeah, there were disproportionately more accounts unsuspended and un-shadow banned on the right because Twitter had a huge thumb on the scale on the, in favor of the left. If you say like, "Have we been suspending accounts on the left? Have we been shadow banning accounts on the left?" No. No, we haven't. Because what exactly what I said is we would do, we are doing, which is to make it an even playing field. You know something is freedom of the speech when it is, when you're hearing speech that from someone you don't like and what you're saying, you don't like what they're saying. This, otherwise it's not free speech. If you don't have that ability, then sooner or later that suppression of speech is gonna be turned on you. It is a good sign if you're seeing people you don't like say things you don't like. That is a good sign, not a bad sign, 'cause it means and provided you can say your piece too. I think this is fundamental. I mean, the reason I did the Twitter acquisition was not because I thought this would be some lucrative goldmine. It and in fact, it has been arduous and difficult with a massive, you know. Being dumped on every day. Well, that's not the most fun thing in the world. If we do not have a strong foundation of free speech, I fear for the future of our civilization. We must have this. That's why I did it. Thank you for doing it. switching gears. a media narrative that, you know, may have some accuracy and not all inaccurate. The code base is difficult to change, tech debt, unwieldy, you know, Rube Goldberg machine, is a term you've used. is this- Fractal Rube Goldberg machine. fractal Rube Goldberg. This. Yes. This is at least partially accurate, if not accurate? Yes. Well, yes. Like the code base is like a Rube Goldberg machine, and when you zoom in on one part of the Rube Goldberg machine, there's another Rube Goldberg machine, and then there's another one. This is your penguin. That's what I mean by the fractal. Yes. A fractal, as you zoom in, it's, there's another fractal and another fractal. A fractal Rube Goldberg machine. It's quite difficult to keep this thing running, and then also difficult to advance the product, because it is really overly complex to say the least. We'll make a change, what appears to be a small change somewhere that then causes a massive disruption. For example, just yesterday, we made what we thought was a small change. We wanna be in the sort of full disclosure of everything, including dirty, you know, gruesome details. Essentially there was what was supposed to be a small change to 1% of the Twitter user base ended up being a catastrophic change to 100% of the Twitter user base. You know, we don't have enough time to go into the details, but there was a Boolean flag in the Twitter front end that should not have been there. Now we fixed that. It's I mean, let me give you like a silly example. At one point we, there was a problem with Twitter Spaces where suspended users were able to join conversations even though they were suspended. We temporarily turned off access to Twitter Spaces, which is, which then re-made anyone who's using the Twitter Android app unable to like tweets. How those things are connected is not clear. If you were In the Rube Goldberg faculty. That's what I'm saying, if you had an iOS app, you could like tweets. If you're on the web app, you could like tweets, but not if you had an Android app because of Spaces. You're like, "What?" you know, one of the things we're doing, there's a lot of work behind the scenes in simplifying the code base, getting rid of extraneous features, and enabling Twitter to evolve more rapidly in the future. It requires a lot of cleanup, essentially. Yeah. You know, you've grown users despite this lean engineering team and the code base being unwieldy and cloud spend 40% down, cut out a data center. Mm-hmm. Is this due to the strength of the engineering team that you've been able to achieve this and introduce subscription and other features while trying to hold it all together? Where does that engineering team go to accomplish both the code base enhancements and all the features you've said you want to introduce? Yeah, I mean, I think on balance we're doing okay because. Just to give you a sense of where things were at the close of acquisition on, say, October 29th, Twitter was tracking to a negative $3 billion a year burn rate and had $1 billion in the bank. That's a pretty dire situation. If 2023 had been a normal year, Twitter would have done something on the order of $4.5 billion in revenue, $4.5 b illion in cost, roughly break even. When you add one and a half billion dollars of debt servicing to that and a massive decline in advertising, some of it's cyclic, some of it's political, call it at least a 50% decline in revenue, roughly a 50% decline in revenue, you've got over $3 billion negative. Twitter has some revenue that's not advertising based. Data services data. Yeah. Yeah. Exactly. Data subscriptions and whatnot. It was In the absence of action, Twitter would have had $6 billion in cost and $3 billion revenue, so -$3 billion, and it was $1 billion in the bank. It would've gone bankrupt in four months. Immediate and drastic action had to be taken, which was. We actually have now cut the burn to the non-interest burn to roughly $1.5 billion. We've got $1.5 billion of debt servicing and $1.5 billion of expenditures. We went from three data centers to two and reduced our cloud expenditures significantly. While at the same time having the fastest product evolution in Twitter's history. Overall, not bad. There've been a few bumps along the road, obviously, but this is to be expected. Now I think we have the opportunity to grow it into something quite spectacular. We have the highest total user minutes in Twitter history. The real number to care about is actually not the MDAU, Monthly... Which is theoretically monetizable Daily Active Users, but it's user time. How many total user hours per day do you have? That's the real figure of merit, 'cause one could, for example, say, go to 300 million daily active users, but if they spent less time on the system, cumulatively, that would actually be a downgrade. It's how much human attention are you worth? That's why I think, like I said, the really profound thing is what Twitter has is roughly 130 million hours of the smartest, most influential people on Earth every single day. There's nothing else that has that. I mean, there are social networks that have more users, but they do not have the smart influential people. They don't have you. The if I do the math on what you've said about the advertiser pause and revenue decline and the cost changes, it's EBITDA profitable today, and then you're looking for cash flow break even after debt service. Well, it's EBITDA profitable. Yeah the D in the EBITDA. The debt service. Is quite big. The D is big. Right. When do you get to cash flow break even? Like- with the D? When do you get to cash flow break even after that D? This is where we need to focus on the E part. Yes I hope we pay taxes. The tea. Yeah. like I said, I mean, we're getting to the point where we're close to having the total expenditures for the company excluding debt roughly equal to the debt. Okay. Cash will make it easier. Yeah. I think, like, we'll be there in Q2. In Q2. like, I don't, I definitely don't wanna count chickens before they hatch or jinx it or anything, but I think we've got a shot at being capital positive next quarter in Q2. Okay. That'll, some of that'll depend on advertisers, so we'll talk quickly about advertising. You've mentioned it a few times. The value is clear. 147 billion impressions on Twitter of World Cup '22, and 50% year-over-year growth in NFL video views. You know, 39% increase in Super Bowl mentions. The NFL putting out engagements never been higher. WPP's CEO, Mark Read, said 10 days or long ago this was, February 24, "Twitter seems to be more stable, and I think clients," advertising clients presumably here, "will look about coming back to Twitter." Are you in touch with agency and CMOs, and is this- Yes a theme? I think we just saw, McDonald's or others- Mm-hmm coming back as they realize you are absolutely focused on the things you were said not to be focused on in terms of, brand safety, et cetera. Yeah. I mean, I think the really what I'd say to advertisers and brands is, you know, use Twitter yourself and believe what you see on Twitter, not what you read in the newspapers. What you see on Twitter is the real thing, and what you read in newspapers is not. I'd like to thank Mark Read and WPP for their support, and Publicis and others, and for the advertisers that have stuck with us, like Disney and Apple. Thank you. That's brand advertising. You've mentioned a few times performance advertising. I think I've heard your story about White Lotus, which you can share here. You know, performance-based advertising has been very lucrative for other companies and. Yeah can be made more relevant on a more narrow cast basis, if you will, than pure brand advertising. Mm-hmm. when do you introduce performance-based advertising and scale it? From the moment the acquisition closes, like, performance-based advertising is really just advertising that is relevant. Relevant. In fact, we should really have aspirationally zero non- Non performance-based advertising. You know, we want advertising that matters. People's attention is precious. We should not serve them ads that are annoying or irrelevant or strident or ugly. It's interesting you mentioned White Lotus. I actually was talking to David Zaslav, who's great, and he was like, "Hey, why can't we put a White Lotus ad, trailer every time someone mentions White Lotus on Twitter?" I'm like, "Absolutely." Like, one of the, like, super obvious but profound things that we're doing is enabling keyword advertising so that you can enter the keywords and, like, White Lotus, and if somebody mentions White Lotus, you put the White Lotus trailer there. I mean, this sounds very obvious. You don't need GPT-3 for this. We don't need advanced AI for this one. You know, it's sort of just Google AdWords but applied to tweets and the home timeline and replies and everywhere else. 'Cause you'll often have very sort of long, deep conversations, people going on, talking about movies, TV, products and whatnot, and that's the perfect opportunity for advertisers to provide their message. You know, if I think about something, for example, like Starlink, which does advertise in various media. For Starlink, would wanna advertise to users in regions that are not already saturated. Starlink tends to be saturated in urban areas, but is not saturated in rural areas. What Starlink would like to do is like to say, please, show the ad in to rural users with a slow connection. The simple message is, do you want faster internet for less money? Click. Probably you do. Twitter needs to be able to do a simple thing like that, and it will be. In fact, it is already able to do that. We just haven't fully rolled it out. We're I think around 20%-ish, but by the end of this year, almost all advertising will be, should be reasonably relevant. That, your Starlink example comes back to your disposable income, if you want, or the education level and the ability to buy and afford a Starlink subscription of the Twitter user. Like, it also dovetails back into that. Yeah. Yeah, absolutely. So, you know, I think there's also an opportunity here to have advertising be much. Like really improve the relevance of advertising using AI, like, in the sense that if you. Based on what tweets somebody views, likes and whatever, you can actually populate a parameter space, an M, you know, ML parameter space, and then you can take an ad, and even if you say nothing about that ad, after it's dropped in the Twitter system and it has 10,000 views, you populate a parameter space of that, of the ad, and then you correlate the user parameter space and the ad parameter space, and then you don't need to do any demographic targeting 'cause you could be like, say, it's a gardening ad. Well, you could be 20, 30, 40, you could be 70 years old, any sex, whatever, it doesn't matter. What matters is you like gardening. That's the ad that should be shown. I think we can get away from a lot of the sort of targeting by age range and sex and whatever, in favor of targeting by interest. And a lot of these sort of demographic targeting was done, coming from a TV or newspaper era where you don't have interaction with the user. You just have to kinda guess 'cause it's a one-way street in TV. On Twitter, it's not a one-way street. There's continuous interaction. I think we can just have a profoundly more useful advertising experience. Let's close on Twitter before getting to another couple companies I've heard that you run. The vision you're building towards, I think you've called it Twitter or the Everything App. Yeah. Tell us about it. You know, beyond just improving the advertising and, you know, what will we be able to do on Twitter in your grandest vision? Yeah. twitter.com is, it, you know, you know. I think it's possible to create a very powerful finance experience basically. Like PayPal is kind of like a halfway version of what I think could be done in payments and finance. Like, let's say, you wanna be able to send money easily from one account on X/Twitter to another account effortlessly with one click. You wanna be able to, I think, earn interest on the money. You wanna be able to have debt, so you can your interest can go negative. I mean, basically, I think it's possible to become the biggest financial institution in the world. Just by providing people with convenient payment options. We don't have the time to go into it in detail here, except if we just make the app more and more useful, people will use it more, and it'll be great. I mean, yeah, so you'll see. Okay. We're still on Twitter for a moment. We had, you know, in Austin, thank you for hosting at the Investor Day. You showcased 16 executives on stage at various times, with you, an incredibly deep and built-out management team. I think the executive team, with you with Twitter, you know, is perhaps a bit leaner. Maybe there's a media narrative here that's accurate. Look, he does have a black turtleneck as well. So, so- Do you need anything more? I don't think so. When does the Twitter management team have that bench like you showcased at the Gigafactory? Well, I think it takes a long time to build a strong management team. We built the Tesla management team over 20 years. I think Twitter is an easier problem than Tesla, by a long shot. It'll take some time to build the team and, I don't know, probably a few years. Okay. Yeah. Switching gears, you shared Master Plan Part 3 at the Gigafactory and, you know, the edit that came to my mind was, you know, Master Planet after your first piece there on sustainable energy for all of Earth. Can you take us through that positive, optimistic, mathematically, underpinned vision for our planet? Okay. There's not a lot of time to do that. I guess the overall message is that we can absolutely make Earth or turn Earth into a sustainable energy economy, fully sustainable, using lithium-ion batteries, solar, wind, as well as geothermal, nuclear, and other things, but primarily it'll be solar and wind and lithium-ion batteries. Our calculation is you need roughly 240 TWh of lithium-ion batteries. Most of those will be iron phosphate lithium of the iron phosphate variety. It's, you know, those with a primarily iron cathode, which is a plentiful material. In fact, Earth is. The number one element on Earth is actually iron, a little factoid. I think Earth by mass is about 32% iron and about 30% oxygen, and then everything else is miscellaneous. We're like a muddy rust ball. Plenty of iron. Basically the materials needed for to make 240 TWh of batteries are actually plentiful on Earth. We don't need to mow down the Amazon or anything like that. We don't need to basically do anything terrible to the environment to create 240 TWh of batteries. In fact, there will be less mining required. In a sustainable energy economy than is currently required. Really this is a message of hope and optimism grounded in physical reality. It is not wishful thinking. We should be excited and inspired about the future. I'm not suggesting complacency, or anything like that, or that we should. Getting there faster is better than getting there slower, but what we don't need to live some terrible austere life, and give up the things that we like. You can have the things that you like, in fact, even more of them. The environment can be good. All the good things are possible is what I'm saying. We should be excited and optimistic about the future, which we need to go build it. It's a lot of work, but you should not feel sad about the future regarding sustainable energy. It will happen. We just wanna make it happen faster rather than slower. That was the first big takeaway. You know, the next one that I had was this, your next phase of vertical integration, the relentless first principles thinking on vehicle design, battery design, factory optimization, you know, at the same time, as the vehicle that could lead to a target, I guess, of this, you know, 50% step change in cost when the new gen eventually comes around. Can you just take us through the quick summary of that and it unlocks the next wave of the TAM because there's price elasticity is the what you were sharing on this subject is the second big takeaway for Tesla? Yeah. I mean, there's a clear path to making a vehicle, a smaller vehicle that is roughly half the production cost and difficulty of a Model 3. That vehicle will be, or, you know, really used almost entirely in autonomous mode. The thing that is really gigantic for Tesla is autonomy. If people have used the Tesla Full Self-Driving and have seen how rapidly the Full Self-Driving capability has been evolving, it should be obvious that that is by far the most profound thing. The sort of total addressable market stuff is like, guys, you... This is, like, actually not the right way to think about it. It's like, passenger vehicles right now only see about 10 to 12 hours of use per week. There's 168 hours in a week. If those vehicles are autonomous, they're probably gonna get used for 50 or 60 hours a week. That's a 5x increase in the value of a car, and it costs the same to make the car. At that point, you basically have software margins in a hardware product. It's insane. Total addressable market is everyone, all humans. Powerful. let's switch gears to SpaceX, hitting, first Starlink. What can you tell us about Starlink and the scale and deployment and how that's going? Yeah. Starlink's, the Starlink team is doing an amazing job. More than half of all satellites in orbit right now are Starlink satellites. If you add up all satellites launched cumulatively, they are less than Starlink. Starlink is currently providing global connectivity. You can get connectivity anywhere on Earth from the most remote part of Antarctica to San Francisco, anywhere. It's full global connectivity, high bandwidth, and low latency. The latency is important because unless you're in low-Earth orbit, you cannot get low latency. The geostationary satellites, you know, are up very, very high. You've sort of got sometimes up to a second of latency on a, from a geostationary satellite, all things inclusive. With Starlink satellites, we believe we can get the latency under 20 msec. In fact, for international communications, an interesting thing is that in fiber, light travels much slower than in air or vacuum. In a rough approximation, light travels about 300 km per millisecond in air and vacuum, but only just roughly over 200 km per millisecond in fiber. You've got like a sort of roughly 40% increase in speed of light going through the Starlink system than through fiber. It can also follow a more direct route. Instead of following the sort of coastline of the continents, it can actually have a more direct route. It's a shorter route and inherently faster from a physics standpoint. It connects the world way better than fiber. And it will provide, and is providing connectivity to people that either never had it before or where their options were extremely expensive. Or very low bandwidth. It's helping out a lot of communities that never had access, and especially when you consider that education is digital these days. That's really how you can learn anything. You can basically learn anything for free on the internet if you have the internet. In terms of providing education abilities to remote communities, Starlink is doing a lot of good in that regard. Lastly, you know, the launch, whether, you know, Falcon Nine Heavy or Starship, the static fire test, that went well and, what can you tell us about the next phase on launch for Starship? Yeah. We're getting ready for the first launch of Starship. This is a very difficult program. The rocket is roughly two and a half times the thrust of a Saturn V. If or once it reaches orbit, it'll be by far the biggest rocket to reach orbit. More importantly, it is designed to be the first fully reusable rocket, orbital rocket, ever. The key to extending life beyond Earth is a fully and rapidly reusable orbital rocket. This is a very hard problem given the constraints of Earth with Earth has a thick atmosphere and strong gravity. It is only barely possible to do this. That's why it's not been done before. We're getting close for our first orbital attempt of Starship. Hopefully in the next month or so, we'll have our first attempt. I'm not saying it'll get to orbit, but I am guaranteeing excitement. It won't be boring. I think it's got, I don't know, hopefully above 50% chance of reaching orbit. We're building a whole series of Starships in South Texas. I think we've got, I don't know, hopefully about an 80% chance of reaching orbit this year. It'll probably take us a couple more years to achieve full and rapid reusability, which I can't emphasize enough is the profound breakthrough that is needed to extend life beyond Earth, because it lowers the cost of access to space by orders of magnitude. In the same way that if let's say there were no airplanes that were reusable, how expensive would air flight be? It would be insane. You'd have to buy a new airplane every time you flew somewhere, and you'd have to tow a small airplane behind you for the return flight. You know, that's just, you're not gonna scale. If many things go well there, this vehicle is, it could make life multi-planetary. That's a really big deal. It could make life on Mars real, and I think that's, I mean, that's one of the great filters that any civilization has to pass through, which is does the civilization become multi-planetary or not? This is one of the elements of the Fermi Paradox. I mean, I sort of wonder that if we are able to get to multi-planetary, that'll be a forcing function for ultimately improving space flight to become multi-stellar, to go to other star systems. I think we may discover that there are many long-dead one-planet civilizations. We don't wanna be one of those. We don't wanna be one of those lame one-planet civilizations. Please. I think we're gonna wrap on that. Thank you, Elon Musk. All right. Thanks.