He goes, "If you get things done, I love you. And if you don't, I hate you." I thought that was so great. What a weirdo.
When I heard it, I was like, "That's how Sam is."
Can I tell you about this Elon Musk podcast? Did you listen to this Cheeky Pint episode with Elon?
I saw it. Didn't listen to it.
You got to listen to this thing. This is an amazing podcast. I think...
Who's the Indian guy? I started seeing him pop up like out of nowhere.
Dwarkesh. Dwarkesh is a podcaster. He's got his own YouTube channel and whatnot. He is this unicorn because he's a really good podcaster. Like he's got good energy, good vibe. You like listening. He's like an authentic nerd in a way, but he's also technical. So one of the reasons this podcast is really interesting is because normally Elon says something and he'll be like...
I predict that in four years we will have, you know, whatever, 10 more data centers in space than we have on Earth cumulatively. And then what me or you or Joe Rogan or any of us would do when he says that, oh my, who?
How are you going to do that, right? That's like, come on, bro. Like, you know, that's all we're capable of. Like our little peanut mind can only do so much.
This guy, like, but Elon, that doesn't make sense because electricity doesn't work the same way in space or whatever.
Yeah, he was pushing back the whole time to where I was like, is Elon going to throw his drink in his face? Like, what's going on here? Is he going to get mad that this guy keeps pushing back on every idea that he says? And because he pushed back, so he would be like, um, but why do we need that? We can, uh, you're talking about this much energy that could be done by this many solar panels that could fit in just Nevada. So why do we need space to power this? And then Elon would be like, well, here's the reason why permitting and blah blah blah blah blah blah. And he's like, but won't there be other problems in space, you know, that are just as difficult? How are you going to fix a broken GPU when it's in space? Like, what are you going to do for maintenance? That's going to be incredibly difficult to do when it's floating in space. And Elon was like, you know, then had to come back with a counterpoint each time. So I thought that was one remarkable thing about the podcast was somebody with enough technical acumen to on the fly calculate. So he, so Elon would say something about space and then he'd be like, so if you're going to do 10,000 launches, that means you're launching a Starship every hour. Every hour you're going to do a Starship launch. How many have you done the last, you've done like, you know, three or whatever in the last year? Like, is that even possible to do an hour every hour? And then Elon would be like, well, uh, you know, there's more planes than that taking off from airports. So yes. And then they would like go into why that's like either correct or incorrect. Uh, and by the way, it's really funny because the founder of Stripe is sitting there, right? It was like a collab. It was on Cheeky Pint, which is made by the Stripe guy, and he probably asked five questions total in three hours, and Dwarkesh asked 500 maybe. But it was exactly what was needed. And shout out to the guy from Stripe for letting that play out the way it did. I thought it made for a great podcast.
Did he best Elon in anything?
No. No, no, it wasn't like a one-upsmanship, but it was definitely a, I'm not just going to let you say that. Like, wait, so what do you know that I don't know? Because I would just think this and he would like kind of earnestly ask the question and he'd be like, okay, I buy that that does this, but I don't see the connection you made to this. And so he forced him to keep going. Uh, which I thought was great.
Did you feel the same sense that I feel when I see a white guy win the 100 meter dash or like...
A sense of national pride?
Like, uh, like when Chris...
Anderson, white cornerback in the NFL.
A little bit. If I'm being honest, a little bit.
Actually, I felt scared for him. I was like, "Dude, tone it down. This is a little crazy. You just get a little feisty here." Uh, but he was also like autistically unaware of, like, so I can't not ask this. I have to ask this because I want to know the answer to it. And I thought that was so great. He's so earnest in the way he does everything. It's great. Um, all right.
Like, what? What? What? Like high-fiving your wife watching it?
Like a World Star hip hop?
Where we at? Where we at?
Next. I may or may not have bought a Dwarkesh, uh, you know, hoodie after that. Um, so can I tell you some of the crazy things that Elon said?
Uh, did he mention Epstein at all?
He did not. They didn't go into, they basically were like, we're going to use this time to talk about like technology, not like the woke blah blah blah politics. Elon, they like did a very minimal amount of it, did a little bit on Doge at the end, but like not much.
Um, I'll tell you first just like a useful thing and then I'll tell you some of the crazy things because the crazy things you're going to be like, we're both again going to be like, whoa, man. Uh, but let's just do the useful things. So they asked about hiring. So they were basically getting into, they're like, "So, um, how do you do, like, what are you doing differently than the rest of us? Like, what is it that you're doing that's leading to all of this incredible, like, individually and any one of your companies is doing incredible things, right? SpaceX, trillion dollar company. xAI came into the AI game like, I don't know, seven years late or something like that and became a $250 billion company."
I saw the xAI thing. I don't even know what it is.
Have you ever used Grok? It's Elon's competitor to ChatGPT, which the product is called Grok. It's his revenge company because when OpenAI kind of screwed him and did exactly the opposite of what he wanted, right? OpenAI, open-source AI that's, you know, a nonprofit and now it's a for-profit closed-source company. And he funded $50 million and now owns none of it. Like his revenge company was xAI. But again, seven, eight years late to the game and still caught up. Tesla obviously worth more than the next 20 car companies combined. So it's like, you know, any one of these companies dominate all of these companies supernova. What is this? What's even happening? And so they were asking about hiring and they were like, you must have incredible people because just by logic there's no way you could be in five of these companies at the same time doing the things yourself. So you must have incredible people. How are you getting these people? And so they were like, "What do you look for when you hire?" Because he hired, I guess, the first thousand people at SpaceX or thousand plus people at SpaceX. He interviewed them himself. And he just goes, "I'm looking for evidence of exceptional ability." And he goes, "Uh, sometimes I would see a resume and I'd be like, 'This person sounds great.'" He goes, "But I've learned you trust the conversation, not the resume." So he's like, "If in that first 20 minutes I'm not saying, wow, I believe the conversation and I don't believe the resume." And he goes, "All I ask them for in the interview is tell me about something exceptional you've done. I'm looking for evidence of exceptional ability." And if they can't tell me a story, he goes, "If they say one thing or, you know, one to three things that I'm just saying, wow, wow, wow." That's all I'm looking for when I hire.
Does he define that of exceptional ability?
Yeah. Like could it be like I was an NCAA athlete and I won? He's looking for technical achievements or like things you've built or done technically. And he was saying, he's like, you know, it's not that I'm batting a thousand, right? I make a lot of mistakes. He goes, but at this point I have probably, you know, to use like the AI terms, I have the most training data probably, you know, out there of hiring like technical talent. You know, I've been doing this for a long time and I've interviewed with so many of these people myself and I've gotten to see what worked and what didn't and then I revised my training. You know, I have a big training set and I revised my own. I RL'd myself and I redid reinforcement learning on myself in order to figure out what should I actually be looking for? What do I actually want? And then they were like, they asked him this great question. They go, "Oh, what do you look for in like a sparring partner?" So, like, what does it take to work well with Elon? And, you know, what do you want? And he goes, "I don't want a sparring partner. I want you to execute well." He goes, if someone executes well, I'm a huge fan. If they don't, I hate them.
He goes, he goes, I don't care about my own idiosyncratic preferences. He goes, if you get things done, I love you, and if you don't, I hate you.
I thought that was so great. What a weirdo.
You say that, I feel like you're the same way, dude. I literally feel like when I heard it, I was like, that's how Sam is.
Well, I just took Andrew Wilkinson's like personality test. You see him tweet that thing out the other day? Yeah, your boy's autistic apparently. So, that's literally what the test.
Wait, why didn't we lead with that?
Yeah. And yeah, breaking. Tell TBPN. I need to get one of those little like announcement things. Yeah.
This is better than your birthday. Feb 9. That was your awakening day. You found out the good news.
Maybe now our podcast can get into the top charts here. We can break into the top 10. We've been lacking.
Yeah. It turns out, yeah, capital A autistic. And so maybe Elon and I, I don't know. Maybe he's got the good type of autistic though.
I got the math one and you have...
Yeah, we both got the one where we are just awkward in social situations and we say things like...
I got the social one and he got the math one. By the way, for anyone who's, we say this with extreme love and a sense of humor about this. There's a long-running joke which is like in San Francisco, like it's one of the highest status things you could have, and that's where the origin of the joke come from. All right, so a lot of people watch and listen to this show because they want to hear us just tell them exactly what to do when it comes to starting or growing a business. Now a lot of people message Shaan and I and they say, all right, I want to start something on the side. Is this a good idea? Is that a good idea? And again, what they're really just saying is just give me the ideas. Well, my friends, you're in luck. So, my old company, The Hustle, they put together a hundred different side hustle ideas, and they have appropriately called it the Side Hustle Idea Database. It's a list of 100 pretty good ideas, frankly. I went through them. They're awesome. And it gives you how to start them, how to grow them, things like that. Gives you a little bit of inspiration. So, check it out. It's called the Side Hustle Idea Database. It's in the description below. You'll see the link. Click it. Check it out. Let me know in the comments what you think.
Can I tell you another thing that he said in this? So he was like, they kept asking. They were like, okay, so like what are you doing differently? He goes, I just have a maniacal sense of urgency. And he goes, I shoot for a deadline that I have a 50% probability of success. And people make fun of me because that means half the time I'm late and I'm wrong about my deadline and I miss my deadlines. But it's worth it because he goes, "Work is like a gas. It expands to fill the time you give it. And so I just don't give it much time." Um, and so, okay, that was good. Now let me tell you about some of the crazy stuff. So...
Wait, that's actually really interesting. I get criticized within my own company of doing that, of creating like these super urgent and people get worn out. And then like Hormozi, who I think is a really smart business guy, he like preaches all the time of like have patience and everything like that. That's a hard dichotomy to handle.
Uh, yeah, I think it's sort of like long-term patience as in you're not going to give up and you're going to stay at it. Um, I think short-term maniacal sense of urgency is, you know, on the task. So, it's like patient with the mission and maniacal about the task is probably the way to do it. So, here's another thing that I picked up. He might have said the phrase limiting factor 400 times in three hours.
Have you ever said that phrase?
So if one thing came out of this, it was how Elon thinks and works is you have a huge mission, right? We're going to go to Mars. We're going to build the first electric car company. We're going to build artificial general intelligence, right? Like whatever it is, huge mission. You have maniacal sense of urgency. Great. That's the second component. And what's the third one? It's this idea of the limiting factor. And so at any given time all he's doing is he's sort of scanning for what is the limiting factor, as in what is the bottleneck that's preventing us from getting to the outcome sooner right now. So for example, with AI right now, they're like, you're doing these space data centers, that sounds crazy. Why are you doing this? Why can't we just do it on Earth? And he's basically like, well, to make AI smarter, we need bigger data centers, more compute. And so you're either going to be limited by chips. Initially we were limited by chips. So I started building chips. So like Tesla's building like their own GPUs to like, you know, like the equivalent of the Nvidia chips. And he bought huge orders of Nvidia chips and, you know, told Jensen like, "Give me everything you got." The second thing is power. And so he's like, the limiting factor is now power, not chips. He's like, I suspect that actually in the next, you know, 24, 36 months, there will be more GPU chips that can't be turned on. So the limiting factor will not be chips anymore. We're not constrained by chips. We'll be constrained by power availability. And so then he's like, um, and so they walked through, they're like, so what do you mean? So like, why can't you just connect to the grid? And he's like, you can't connect to the grid for these reasons. They're like, okay, well, why can't you do it off grid? Why can't you just buy like a power plant and power your own chips, not connect to the grid? He's like, well, there's three companies that make the power plants and they're all booked out till 2032. We can't build the power plants fast enough. There's not enough people or projects and time to build enough power plants as we're going to need. And they're like, well, specifically like what's the... So then he's like, specifically why? He's like, actually it's the turbines inside the power plants that we cannot procure in time and there's this many companies that make the turbines and actually within the turbine it's the blades and vanes that you can't do. And so we're going to try to make them at Tesla, but even that will take time. And so we're limited by this. And so I'm throwing all my weight into whatever is the limiting factor at any given time in this business. And so it ended up with like, we're going to launch these data centers in space. That's the way to get around this limiting factor. But over and over and over again in the whole podcast, it's basically like the formula is like identify the limiting factor and then go ape shit to get over it. And most people don't do either of the two. They don't actually correctly address or identify the limiting factor and then even if they did, they don't go ape shit, right? Um, and so ever since this podcast, it like any meeting I've had, so like, you know, I check in with kind of my portfolio companies and I've just gone full Elon on them where I'm just like, so we want this outcome, what's the limiting factor? And then the CEO will talk for five minutes. I'm like, cool, you didn't answer my question. What's the limiting factor? Well, the limiting factor is X. Okay, great. Why are you talking about all these other things besides... It's X. X is the limiting factor. So all we're going to focus on is X. Okay. Well, what does that... Okay. So what does that mean? Now once you identify that, you need to drop everything else. Just focus on X and then go ape shit on it. And he would say, he's like, in my companies, if something's going well, you're not going to see me. Right now the Boring Company is going well. I don't spend any time on it because it's going well. If it hits a roadblock and there's a limiting factor and they can't solve it, then I will come in and I'll throw my full force, my full weight at it. And I just thought this simple operating philosophy I think can serve a lot of people is to identify the limiting factor and then figure out how to throw your entire weight against it and, you know, sort of if you got to go over the wall, over, through the wall, around the wall, under the wall, whatever you got to do, you got to get past that wall. Like I don't know, like I don't know if you want to play the game, but like right now let's take Hampton for example. What's the limiting factor in Hampton? Thought exercise.
Yeah. So we basically only do two things. We get the best people to join a core group and then we allow them to have good conversations. And so for us, the limiting factor in order to expand to chapters globally is to get more of the best people. So more people.
What's the limiting factor on more good people?
Well, I could probably make a list of five things that we could do. For example, we could ask for referrals more than we do. We could probably spend more on advertising.
So, okay. So, those would be solutions, but just one level deeper into the problem. So, like we have a lot of people, they're not good. Or we have good people, but we just don't have enough of them. We don't have enough people applying, period. We're limiting our prospects to only 13 cities and so we need to open that up in order... We have thousands and thousands of people applying every single month but we limit it only to 13 cities. So what's the limiting factor of opening up 13 more cities?
Just bandwidth probably of like having someone really focus on a particular city because once one person joins or a prospect shows interest in let's say Birmingham, how do we find seven other people who are interested? And then the way that we will eventually scale that is we will instead pivot our business to where moderators are basically franchise owners and they have to go and help recruit seven other people once we give them one person. And so they're running their own little coaching business.
Right? And so a useful exercise in this case would be basically what's stopping us from doing it tomorrow. So let's say tomorrow you wanted to open 10 cities. What would actually break if you opened 10 cities tomorrow?
Well, a customer would potentially be angry because they'd have to wait a long time until we found seven other people in let's say Birmingham.
And so then you say, is that true? Is there not... do we have... so to have 10 cities you need how many members like to activate that city?
Um, well, you just need one core group of eight people. So you need 80 people. So you need eight per city. Um, do you have eight that are ready to open one city tomorrow or no? Like, could you open one tomorrow? Yes. Could you open two tomorrow? Probably maybe. But like it would get hard. I would need more people. And then the question would be like, well, why do I need more people? You know, why can't I just do it with less people?
Uh, yeah. Yeah. Okay. So you start to get into like the nuts and bolts of it, right? But like if we just even just look at what just happened, right? It was like started with a high level problem. We need more power, right? In your case, it was like we need more good people. And then you got to turbines and blades and vanes is the limiting factor, which was like actually Mississippi has four people and we need eight, right? That's the turbines, that's the blades and vanes level of detail. Um, and so, you know, could you do... if we want to do, we need to do 10? Could I do one? Yes. Could I do two? No. What happened at two? Why did one to two break? Right. So you sort of... Okay, there's a limiting factor somewhere between one and two. What's different? What happened there? And also, a second ago, you were like, we could ask for referrals. We could do X. We could do Y. It's almost like we jump to solutions really quickly. And what I think actually happens in these companies is we sort of already have in our mind a bunch of pet projects of good ideas, generically good ideas, but not specifically problem-solving ideas. And I think one big source of waste in companies is when everybody wants to do their pet projects or everybody wants to do their generically good ideas rather than the specifically effective ideas that are going to solve the current bottleneck that is right in front of us that we could unblock for tomorrow.
I think this makes sense. I think it's hard to pull off. I think that like it's hard to pull off just because of interpersonal relationships. A lot of people would like approach this and be like, well, I'm doing all these other things. Like how dare you like tell me just do this one thing. Like you don't understand. And like how are we going to keep the business running? You know what I mean? That's the pushback mostly.
Totally. Uh, and I think that's by the way, that's a very real thing like, hey, we're going to focus on this and it's like, well, actually you don't really realize I'm holding up this entire ship by keeping these six trains running on time and I have to keep doing that. So a lot of people, a lot of times I've seen this where it's like I could tell the good people in my company, they're like, yeah, I hear what you're saying and I'm going to do that. I'm not even going to... I'm going to have to keep some of these trains running and we don't need to discuss it. I just know it needs to get done and blah blah blah. What I found is pretty effective is to address it in the call. So what I'll do is I'll address the trade-off. So I'll say, so here's what that means. By focusing on this, what that means is that this area where we actually know how to improve it and we have a bunch of good ideas and maybe even two things that are in flight, we're actually just not going to do any of those. And we're going to accept mediocre progress here for exceptional progress here. And nobody wants to say that out loud, but I'm going to say it out loud because it's true. We have a trade-off of energy and focus and intention. And if we're going to move it all over here, that's what that means over there. We're all saying out loud, yes, we understand and accept this. It's like accepting the terms and conditions when you sign up for a website. It's like, "Yes, you're going to sell my data." It's like, "Yep, over here we're going to make far less progress than we want." And I think this happens throughout life, right? You have a kid and you're like, "Wow, this is really important." Well, guess what's going to be a little less important is like your work for the next three months and your gym routine is going to go out the window. And you just have to like, you know, if you don't accept those trade-offs, you're just going to feel a constant underlying state of anxiety or stretch yourself too thin and do a poor job of everything. Rather, it's actually just better to say, well, great, for this season, this is what matters, and I'm okay with a less than stellar rate of progress. I will only do this amount to keep this running, but beyond that, I will not be doing for this next period of time.
Are you good at not bitching at them about why this thing sucks? For example, for The Hustle, it was grow email list, therefore social media does not matter. And then I... Why does our social media suck? Yeah.
Yeah. Yeah. Yeah. Uh, well, I give them permission to... I mean, we put the thing right in front of us, like front and center. I'm an obnoxious repeater. So like every day when I start my day, I will say, "Hey, here's our goals for the year. Here's the things we said matter. Here's what matters this week." And I'm like a bot. But because I realize like if I don't keep repeating it, people don't even know what we're trying to do. Um, and we don't like... so I try to get a really crisp set of things, really crisp clear set of things we're doing. And then I will repeat it so much to the point that I myself remember we said social media doesn't matter. All that matters is growing email list. And they have permission to be like, wait, but social media doesn't matter. All that matters is the list. And it takes some time to work together where you kind of screw that up. It's like you reflect, you like, why was this last month not as productive as we wanted? And then you sort of come to the pow-wow and realize like, well, it's because we said we were going to focus on these things but then we kept getting distracted and, you know, wanted to also do these five other things. And guess what, when we tried to do it all, like we can have anything, we just can't have everything all at once. And you learn that the hard way once, twice, and it's like, you know, George Bush, you know, fool me once, shame on me, fool me twice, don't fool me again. You know, it becomes that at a certain point. So, the good thing, you know, I've worked with Ben for like six years, so we kind of know like, yo, we've stepped in that gum before. Like, do we really want that on our shoe again? I can kind of see it coming if we're going to do this or you're saying this, but remember this is like that other time, so let's just stay focused.
You threw that George Bush quote in there like you're rapping.
That was great. What's the quote? Fool me once, shame on me. Fool me twice, don't fool me again.
Yeah, he like loses the quote. So he's like, fool me twice, uh, don't, you can't keep fooling me. It's like an old timer.
Can I show you one other clip that I thought was pretty just, uh, I don't know, wild from Elon?
Yeah.
Is it the clip about how he lied about not going to Jeffrey Epstein's island?
Um, I'm not sure if he went or not, but he said that he refused to go, but the emails say that he was actually begging to go.
Yeah. Yeah. Yeah. Yeah. Hey, we're going to be around the town. You guys got anything going on? Yeah, definitely. Um, all right. Watch this.
More intelligent than humans. So, in some sense, you're like a doomer and this is like the best we've got. It's just like it keeps it around because we're interesting. I'm...
I'm just trying to be realistic here. Um, if AI intelligence is vastly more, if AI is like, you know, let's say that there's a million times more silicon intelligence than there is biological, um, I think it would be foolish to assume that there's any way to maintain control over that. Now you can make sure it has the right values or you can try to have the right values.
Okay. So, I don't know if I... I think I might have picked the wrong time step a little bit there, but uh, here...
50 minutes in and they barely drink their drink.
No, there's refills. There's refills. Well, one of the funny jokes is that Dwarkesh doesn't touch his pint. Uh, he just keeps bringing it up to his mouth and it just wets his beard and then it goes back down with the same amount in it while the other guys are just like crushing them. Um, all right, so here's what he says about AI. So the topic is, um, how do you make sure that AI, you know, ends up being good for humanity? And he says something where he goes, they were like, how do you stay in control? And he says, um, if you have intelligence that's 100 times smarter than any human, it's hard to imagine that the humans stay in control. And then the second thing, then they go, so wait, are you like a doomer here? Like, you know, you just think we're doomed? And he goes, I'm just trying to be realistic here. If AI is vastly more intelligent, like there's a million more times silicon intelligence than biological, I think it would be foolish to believe that we can maintain control over that. All you could do is try to make sure it has the right values. Like kind of like basically like it goes on like that it would keep us around.
And I just thought that's a pretty stunning admission for a guy who's building AI to basically say out loud, um, this thing is going to be so much smarter than humans. And when it is, the idea that the chimps are going to stay in control of the humans, like that the humans are going to be in control of this thing that's a hundred times or a thousand times or a million times doesn't make sense to me because it's foolish. Like, uh, you know, the ego is the reason why like Napoleon's Napoleon, you know, like the reason to do like there like does a computer have the need to dominate? It's not even that it dominates, it's that we don't control it. So for example, like if we say, cool, I want you to run all decisions by me, but it makes decisions a thousand times better than any human alive in the history of mankind. Realistically, if you gave it the mission of being successful, is it actually going to run the decisions by you?
But that's what I'm asking. Does it actually have the mission to be successful like a human has?
Well, you're going to give it to... even if the human is the one prompting it, say, hey, I want you to make this thing really... I want you to help me become president. I want you to build this successful company. I want you to build this technology.
Do you think after that it's going to care what you have?
But you can turn it off if you turn it on.
That's my point. Like humans have a compulsion to be right. A lot of humans have a compulsion to be nice, to reciprocate. Does it have compulsions? And so like, can you turn it the way that you just told it to dominate or to win? Can you also say win less?
You might. Then the next guy says, "No, I want to win. Stay on." All it takes is one guy to not turn it off. Right. Like that same thing with one country to not have safeguards on it. Um...
Have you seen the AI social network where they're like all AI guys talking about like how they're going to dominate humans?
Yeah. Although that might be like kind of faked. It's unclear at this point if that's fake or real. Like it might be that humans are saying say something like this because it freaks everybody out. Goes viral. Have you seen this? Elon tweeted this meme. I don't want to live in a world where someone else makes humans irrelevant before we do.
That is kind of how it feels. Um, but like I hope he's wrong.
Yeah. Even with SpaceX, uh, you know, he's talking about like, you know, the goal is, you know, the original mission with SpaceX is to like preserve the, he calls it like the candle of consciousness, like basically like what if something happened to Earth? If we're a single planet species, like all consciousness could get wiped out. Uh, it should be really important that we're a multiplanetary species so that like human consciousness or intelligence like, you know, lives on. And in this interview, he kind of backs off that. He's like, you know, consciousness is we don't really know what that is, so let's just say intelligence. Um, because xAI just merged with SpaceX and he's like, so you know, at least intelligence will propagate through the internet. It's basically like the robots and the chips and the AI will definitely be multiplanetary whether we are or not. Like he basically his argument. And so Dwarkesh is like, why do you think that they would care that humans survive? And he's like, well, um, you know, there's no reason to kill us. Like that's not very reassuring was essentially the way that he put it. He's like, you know, humans are interesting and we're part of consciousness. So like if you want to maximally expand consciousness, you would keep humans around, right? There's not a big advantage to doing it. It's like, whoa, if that's the moral victory we're clinging to, this is a pretty startling thing. So I just thought that's a kind of crazy admission that he was making. Can I tell you two other just insane things from this? So he's talking about what he's doing with his AI project, which I don't think he's ever really talked about before. Have you ever heard of this Macro Hard project?
No. What's it called? Macro Hard.
Yeah. Like the 70 for 30 or what's it called? 70 for 70 where you got to like run a mile every day for 70 days. 70 Hard.
That's what it sounds like.
No, he made it the opposite of Microsoft. So he's like, what's the opposite of Microsoft? It's Macro Hard.
And basically what he's doing is he's building what he calls human emulators. So here's the Elon philosophy or the strategy that kind of got revealed during this podcast. So the strategy is how do you win? Um, well he goes, if you think in the limit, what is the most that AI can do before you have like robots that are, you know, um, like artificially intelligent robots, which they're trying to build at Tesla, right? So Tesla's building Optimus, which is the robot that can do any of the work that a human could do. Um, and you know, that's robots. He goes, so what is he doing with his software play? He goes, think in the limit. Well, the limit is if you're not in a robot, it's anything a human can do on a computer. And so he's like, um, basically he's, I think he says within like 12 or 24 months he believes that AI will be able to do anything that a human can do on a computer. So he calls it a... so they're building what they call the human emulator. So it's an AI that can work like do anything a human could do with a computer. So like I have an executive assistant. So the idea would be obviously the AI would be able to do everything she does for me. So researching things, booking things, uh, you know, making anything happen, that's obviously there. But also you and I doing this podcast, if we can do this online, that means that AI should be able to do this online. It should be able to produce an interesting podcast twice a week about, you know, business trends, opportunities, uh, ideas that are related to entrepreneurship for business junkies just like us. And it should be able to have a sense of humor, but bring to the table three or four really interesting topics per episode. Um, so AI should be able to do that because this is done online. So he goes, "Anything that's moving electrons, the AI should be able to do." And they go, "Well, how are you trying to solve that?" And he goes like, "So do you want me to just give away all my secrets on a podcast?" Like that would be sort of stupid, right? Like I'm in the most competitive game in the world right now is to create this. He calls it the highest ELO battle in the world. Uh, being like, you know, the chess rankings like ELO, your ELO ratings. So he's like the highest...
ELO needs a wedgie, man. This guy's hilarious. Macro Hard human emulator. This is crazy.
So he's like, um, you know, he's like, it'll take at least three more beers for me to get, you know, to reveal that. But he's like, it'll be something like the way that Tesla solved self-driving. And then they were like, okay, unrelated question. How did Tesla solve self-driving? And he's like, well, what it, you know, what it did was it used very basic sensors, the same sort of sensors that humans have, cam in this case, you know, mostly cameras, eyes, vision, and it watched humans drive a lot, like millions and millions and millions and millions of miles. And then it tried to emulate what they would do. So it tries to, while you're driving, it tries to guess what you...
would do. And anytime there's a difference, it notes it and it tries to learn from other people in similar situations until it can basically match what a human would do while driving. And that's what Tesla self-driving does. And so similarly, what they're doing with the human emulators is they're basically getting tons and tons of data of humans using computers and they're using that to teach computers how to use computers. And that's where this is going for Macrohard. And then he talked about on the robot side,
They were like, is that a, is Macrohard a company?
I don't know if it's a project or a company. Yeah, I think it's a project within XAI. I don't think it's a separate company.
How many people work at XAI? Is XAI considered Twitter now?
They did merge. XAI and X merged and now that merged with SpaceX.
Crazy.
And okay, so now let's talk about the robots thing real quick. So he goes, um, the Optimus robot. He goes, it's the infinite money glitch because labor is the biggest market in the world and if you can have robots that can do human labor and specifically once the robot can build more robots, that's the infinite money glitch. So basically once Optimus can build more Optimus, it's over. Um, and he's like, and they were like, well, how will you do the emulator thing? Because Tesla made sense, right? Humans bought the cars and drove them around. So that gave you all the training data of how to drive. And with Macrohard, theoretically, you could have a bunch of, you know, humans using computers and emulate it. But how do you do the robot one? There's not enough robots to learn from. And so he said he's building this warehouse where 10,000 robots will be able to self-play. So, it's basically this giant warehouse where 10,000 robots are a bunch of toddlers
walking around trying to figure out how to do tasks like pick this up, move this here, try this, work together on this, you know, build this box without it breaking, you know, whatever. And they're not going to tell them how to do it. They're just going to self-play over and over and over again until they figure it out. The same way
They're going to like come in after three weeks and they're going to be like trying to like saw the bars to get out. You know what I mean? Like they're like escaping prison.
So, like this is how, um, they built the best chess engines and Go engines was instead of telling it strategies and rules. They just said play 10 million games and figure it out. All you know is the rules of where the pieces can move, like this pawn can move forward, the bishop can move diagonally. That's all we tell you. And we'll tell you that winning is good. Go. And it played like 10 million games and it became the best chess player in the world. They did it with Go. It became the best Go player ever. Um, now he's doing the same thing with robots, but I thought, how crazy is that going to be? There's just this self-play warehouse where the robots are trying to figure out how to do stuff. Like,
Where is it going to be? Where's this going to be?
Uh, I don't know. He didn't give me the address, but it sounds pretty awesome, right?
They got to put like, you know, they got like puppy cams at like doggy daycares.
