YouSaid · the spoken record

John Carmack

lines on the record: 363
first: 2022-08-04
most recent: 2022-08-04
sittings or episodes: 1
sources: podcast

Every line below is reproduced as it was said and linked to the record it came from. Nothing here is summarised or generated. Directory · Search · Corrections

“I think in many corners he does not get the respect that he should about being a wealthy person that could just retire. And he went all in where he was really going to, you know, he could have gone bust and there's plenty of people. You look at the sad athletes or entertainers that had all the money in the world and blew it. He could have been the business case example of that. But the things that he was doing, space exploration, electrification of transportation, solar city type things, these are big world level things. And I have a great deal of admiration that he was willing to throw himself so completely into that because in contrast with myself, I was doing Armadillo Aerospace with this tightly bounded. It was John's crazy money at the time that had a finite limit on it.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“Options. What can happen with different ways of putting things together? And then in some ways, he was certainly the biggest player in the sort of alt-space community that was going on in the early 2000s. The most well-funded, although, you know, his funding in the larger scheme of things compared to like a NASA or something like that was really tiny. It was a lot more than I had at the time. But it was interesting. I had a point years later when I realized, okay, my financial resources at this point are basically what Elon's was when he went all in on SpaceX and Tesla. And there's.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“Think in some ways we do have a kind of similar background where we're almost exactly the same age and we had kind of similar programming backgrounds on the personal comput Sensibility similarities where we kind of call bullshit on the same things and kind of see the same opportunities in different technology. And there's that sense of, you know, I always talk about the speed of light solutions for things and he's thinking about kind of minimum manufacturing and engineering and operational standpoints for things. So, I mean, I first met Elon right at the start of the aerospace era where I wasn't familiar with I was still in my game dev bubble. I really wasn't familiar with all the startups that were going and being successful and what went on with PayPal and all of his different companies. But I met him as I was starting to do Armadillo Aerospace. And he came down with kind of his right hand propulsion guy. And we talked about rockets. What can we do with this? And it was kind of specific things about how are our flight computers set up? What are different proportions?”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“Because the big company, there are so many resources there, and they do amazing things, do get accomplished. There's so much more that could come out of that. And I'm always hopeful. I'm an optimist in almost everything. I think things can get better. I think that they can improve things, that you go through a path and you're learning kind of what does and doesn't work. And I'm not ready to be fatalistic about the kind of the outcome of any of that.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“Extend eye, and you should probably not say that, but for the most part, yes, I get away with every week I'm posting something pretty loud in opinionated internally. And I think that's useful for the company. But yeah, it's rare to have a position like that. And I can't necessarily offer advice for how someone can do that.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“And I try to point that out and push people. It's like try to go ahead. And when you see something, because there is, you get the silo mentality where you're like, okay, I know something's not right over there, but that's I'm staying in my lane here. And there's a couple people that I can think about that are willing to just like hop all over the place and man, I treasure them, the people that are just willing to, they're fearless. You know, they will go over and they will go rebuild the kernel and change this distribution and go in and hack the firmware over here to get something done right. And that is relatively rare. You know, there's thousands of developers and you've got a small handful that are willing to operate at that level. And it's potentially risky for them. The politics are real in a lot of that. And I'm in the very much the privileged position of I am more or less untouchable there where I've been dinged like twice for it's like you said something insensitive in that post.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“Small company can just go do it while in a big company you do have to worry about I am is there some SDK internally that you should be using because another team's making it you have to have your cross-functional group meetups for different things you do have more concerns about you know about privacy or diversity and equity and safety of different things parental issues and things that a small startup company can just kind of you know cowboy off and do something interesting and there's a lot more that is a problem that you have to pay attention to in the big companies, but I'm not willing to believe that we are within even a factor of two or four of what the efficiency could be. I am constantly kind of crying out for it's like we can do better than this.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“More so than like, yeah, Google goes and cancels all of these projects, different things like that, while Meta is really sticking with the funding of VR and AR is still further out with it. So there's something to be said for that. It's not just going to vanish. The work's going in. I just wish it could be all those resources could be applied more effectively because I see all these cases. I point out these examples of how a third party that we're kind of competing with in various ways. There's a number of these examples and they do work with a tenth of the people that we do internally. And a lot of it comes from, yes, there's.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“On the one hand, we did kind of achieve it and win. And we've got, we're a success right now, but the amount of resources that have gone into it, it winds up getting cluttered up in accounting where Mark did announce that they spent $10 billion a year, like on reality labs. Now, Reality Labs covers a lot. It was, VR was not the large part of it. It also had portal and Spark and the big AR research efforts. And it's been expanding out to include AI and other things there where there's a lot going on there. But $10 billion was just a number that I had trouble processing. I feel sick to my stomach thinking about that much money being spent. But that's how they demonstrate commitment to this, where it's not.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“The plus side, VR has turned out, we've built pretty much exactly what we just passed the 10-year mark then from my first demo of the Rift. And if I could have said what I wanted to have, it would have been a standalone inside-out tracked 4K resolution headset that could still plug into a PC for high-end rendering. And that's exactly what we've got on Quest 2 right now.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“So it's been really eye opening to me working inside a tech titan where I had my small companies and then were acquired by a, you know, a mid-sized game publisher. And then Oculus getting acquired by Meta. And Meta has grown by a factor of many just in the eight years since the acquisition. So I did not have experience with this. And it's interesting because I remember like previously my benchmark for kind of use of resources was some of the government programs I interacted with on the aerospace side. And I remember thinking there was, okay, there's an Air Force program and they spent $50 million and they didn't launch anything. They didn't even build anything. It was just kind of like I.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“Needs to get better. We need to have 100 million of them out there just so that everybody knows somebody that uses this all the time.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“Better than Zoom, better than you, yeah, better than you'd expect for that. It is definitely, yeah, I'd say it's quite a bit better than Zoom when everything's working right, but there's still all the rough edges of the reason Zoom became so successful is because they just nailed the usability of everything. It's high quality with an absolutely first rate experience. And we are not there yet with any of the VR stuff. I'm trying to push hard to get I keep talking about it's like it needs to just be one click to make everything happen. And we're getting there in our home environment, not the whole workroom's application, but the main home where you can now kind of go over and click an invite and it still winds up taking five times longer than it should. But we're getting close to that where you click there, they click on their button, and then they're sitting there in this good presence with you. But latencies need to get a lot better. User interface needs to get a lot better. Ubiquity of the headsets.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“The voice latency is better on that than a lot better than a Zoom meeting. So you've got a better sense of immediacy there. The expressions that you get from the current hardware with just kind of your controllers and your head is pretty realistic feeling. And you got a pretty good sense of being there with someone with you.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“Yeah, so right now you can get workrooms on Quest and you can set up these things and it's a pretty good experience. It's surprisingly good.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“I like cheaper because also lighter and cheaper wind up being a virtuous cycle there where expensive and more features tends to also lead towards heavier and it just kind of goes it's like let's add more features the features are not you know they have physical presence and weight and draw from batteries and all of those things so i've always favored a lower end uh cheaper faster approach that's why i was always behind the the mobile side of vr rather than the higher end pc headsets and i think that's you know that's proven out well but there's i you always ideally we have a whole range of things but if you've only got one or two things it's important that those two things cover the uh you know the scope that you think is most important when we're in a world when it's like cell phones and there's 50 of them on the market covering every conceivable ecological niche you want that's going to be great but we're not going to be there for a while”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“So I have big arguments internally about strategic things like that, where I, like the next headset that's coming out that we've made various announcements about is going to be a higher end headset, more expensive, more features, lots of people want to make those trade-offs. We'll see what the market has to say about the exact trade-offs we've made here. But if you want to replace Zoom, you need to have something that everybody has. And you like cheaper.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“There was a lot of people that especially rich people that would look at VR and say it's like, oh, this just isn't that good. And I'd say it's like, well, you've already been courtside backstage and on Pit Row and you've done all of these experiences because you get to do them in real life. But most people don't get to. And even if the experience is only half as good if it's something that they never would have gotten to do before, it's still a very good thing. And as we can just push that number up over time, it has a minimum viable value level when it does something that is valuable enough to people as long as it's better inside the headset on any metric than it is outside and people choose to go there, we're on the right path. And we have a value gradient that I'm just always hammering on. We can just follow this value gradient. Just keep making things better rather than going for that one, close your eye, swing for the fences.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“So the idea of like I'm doing a VR interview with someone, I'm not saying it's here right now, but you can see glimmers of what it should be. And we largely know what would need to be fixed and improved to, like you say, there's a difference between a remote interview doing a podcast over Zoom or something and face-to-face. There's that sense of presence, that immediacy, the super low latency responsiveness, being able to see all the subtle things there just occupying the same field of view. And all of those are things that we absolutely can do in VR.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“And visit them without everybody getting on a plane and meeting in someplace and dealing with all the other logistics hassles. There is real value in the presence that you can get for remote meetings. It's all the little things that we need to sort out, but those are things that we have line of sight on. People, you know, that have been in like a good VR meeting using workrooms where you can say, oh, that was better than a Zoom meeting. But of course, it's more of a hassle to get into it. Not everyone has the headset. Interoperability is worse. You can't have, you cap out at a certain number. There's all these things that need to be fixed, but that's one of those things you can look at and say, we know there's value there. We just need to really grind hard to file off all the rough edges and make that possible.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“And everybody thought that was dystopian. And that's like, oh, you're just going to forget about the world outside. And I don't get that mindset where the idea that if you can make the world better inside the headset than outside, you've just improved the person's life that has a headset that can wear it. And there are plenty of things that we just can't do for everyone in the real world. Everybody can't have Richard Branson's private island, but everyone can have a private VR island and it can have the things that they want on it. And there's a lot of these kind of rivalrous goods in the real world that VR can just be better at. We can do a lot of things like that that can be very, very rich. So yeah, I want the, I think it's going to be a positive thing, this world where people want to go back into their headset, where it can be better than somebody that's living in a tiny apartment can have a palatial estate in virtual reality. They can have all their friends from all over the world come over.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“The whole point for this place is that there was a time when we were kind of asked to come up with like, what's your view about VR? And I am, you know, my pitch was that it should be better inside the headset than outside. It's the world as you want it.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“Standing still, all the stuff flies at you. It has placed audio strengths. It adds the whole fitness in VR. Nobody was thinking about that back in the beginning. And it turns out that that is an excellent daily fitness thing to be doing. If you go play an hour of Beat Saver or Supernatural or something, that is legit solid exercise. It's more fun than doing it just about any other way there.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“Audio and music turned out to be a really powerful aspect of virtual reality where you're blocking the world off and constructing the world around you and being something that can run efficiently on even this relatively low powered hardware and can have a valuable loop in a small amount of time where a lot of modern games, you're supposed to sit down and play it for an hour just to get anywhere. Sometimes a new game takes an hour to get through the tutorial level. And that's not good for VR for a couple reasons. You do still have the comfort issues if you're moving around at all, but you've also got just discomfort from the headset, battery lifespan on the mobile versions. So having things that do break down into three and four minute windows of play, that turns out to be very valuable from a gameplay standpoint. So it winds up being kind of a perfect storm of all these things that are really good. It doesn't have any of the comfort problems. You're not navigating around.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“Winds up playing only to the strength. It completely hides all the weaknesses of it because you are holding something in your hand. You keep a solid grip on it the whole time. It slices through things without ever bumping into things. You never get into the point where, you know, I'm knocking on this table, but in VR, my hand just goes right through it. So you've got something that slices through. So it's never your brain telling you, oh, I should have hit something. You've got a lightsaber here. It's just, you expect it to slice through everything.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“It actually makes very clear sense when you analyze it from hindsight and look at the engineering reasons where it's not just that it was a magical quirky idea, it was something that played almost perfectly to what turned out to be the real strengths of VR. The one thing that I really underestimated importance in VR was the importance of the controllers. I was still thinking we could do a lot more with a gamepad and just the amazingness of taking any existing game, being able to move your head around and look around that that was really amazing. But the controllers were super important. But the problem is so many things that you do with the controllers just suck. It feels like it breaks the illusion, like trying to pick up glasses with the controllers where you're like, oh, use the grip button when you're kind of close and it'll snap into your hand. All of those things are unnatural actions that you do them and it's still part of the VR experience. But Beat Saber.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“So people will go back and they will name firsts and critical moments, but it's probably going to be a poor approximation of what actually happens. And we've already seen in the VR space where it didn't play out the way we thought it would in terms of what was going to be. Like the modern era of VR basically started with my E3 demo of Doom 3 on the Rift prototype. So we're like first person shooters in VR, matchmade in heaven, right? And that didn't work out that way at all. They have the most comfort problems with it. And then the most popular virtual reality app is BeatSaber, which nobody predicted back then.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“I think that's kind of the fallacy of historians, though, looking for some of those kind of primary dominant causes where so many of these things are, like we see an exponential curve, but it's not because like one thing is going exponential. It's because we have hundreds of little sigmoid curves overlapped on top of each other and they just happen to keep adding up so that you've got something kind of going exponential at any given point, but no single one of them was the critical thing. There were dozens and dozens of things. I mean, seeing the transitions of stuff like as obviously MySpace giving way to other things, but even like blogging giving way to social media and getting resurrected in other guises.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“So, I do think it's going to be kind of like the way the web slowly took over where you're the frog in the pot of water that's slowly heating up, where having lived through all of that, I remember when it was shocking to start seeing the first website address on a billboard when you're like, hey, my computer world is infecting the real world. You know, this is spreading out in some way. When you look back and say, well, what made the web take off? And it wasn't a big bang sort of moment there. It was a bunch of little things that turned out not to even be the things that are relevant now that brought them into it.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“Of social sense of life or whatever you want to add to it. I still think that's quite a good position to take. And I think that while meta is doing a bottoms-up capability approach with Horizon Worlds, where it's a fairly general purpose, creators can build whatever they want in their sort of thing. It's hard to compare and compete with something like Fortnite, which also has enormous amounts of creativity, even though it was not designed originally as a general purpose sort of thing. So we have examples on both sides. Me personally, I would have bet on trying to do entertainment, valuable destination first, and expanding from there.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“And there are other cases where like right now Roblox is basically a game construction kit aimed at kids. And that was a capability first play and it's achieving scale that's on the same order of those things. So it's not impossible, but my preferred bet would be you make something amazing that people love and you make it better and better. And that's where I could say we could have gone back and followed a path kind of like that in the early days if you just kind of take the same game, whether it's when Activision demonstrated that you could make Call of Duty every year. And not only is it not bad, people kind of love it and it's very profitable. The idea that you could have taken something like that and take a great game release a new version every year that lets the capabilities grow and expand to start saying it's like, okay, it's a game about running around and shooting things, but now you can have bring your media into it. You can add persistent.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“All these people Minecraft is amazing case study in so many things. What's been able to be done with that is really enlightening.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“Out very well. On the other hand, we have successful games that started with things like Doom and Quake and communities that formed around those. And whether it was server lists in the early days or literal portaling between different games, and then modern things that are on completely different order of magnitude, like Minecraft and Fortnite that have 100 million plus users, I still think that that's the right way to go to build the metaverse is you build something that's amazing that people love and people wind up spending all their time in because it's awesome and you expand the capabilities of that.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“So the term comes from Neil Stevenson's book Snow Crash, which many of us had read back in the 90s. It was one of those kind of formative books. And there was this sense that the Possibilities and kind of the freedom and unlimited capabilities to build a virtual world that does whatever you want, whatever you ask of it, has been a powerful draw for generations of developers, game developers specifically and people that are thinking about more general purpose applications. So we were talking about that back in the Doom and Quake days about how do you wind up with an interconnected set of worlds that you kind of visit from one to another and as web pages were becoming a thing you start thinking about what is the interactive kind of 3D based equivalent of this and there were a lot of really bad takes you had like Vermal then virtual reality markup languages and there's aspects like that that that came from people saying well what would what kind of capabilities should we develop to enable this and that kind of capability first work has usually not panned out”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“Yeah, I mean, I just talked with him by last year, or I guess it was even this year, about mentioning that I'm going off doing this AI stuff. I'm going big into artificial intelligence.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“Yeah, I mean, he ate it up and he did personify. There was the whole game developers with Ferraris that we had there. And I thought that led to some challenges there. But so much of the... You know, the stuff that was great in the games did come from him. And I would certainly not take that away from him. And even after we parted ways and he took his swing with IDOS, in some ways he was like, he was ahead of the curve with mobile gaming as well, where one of his companies after IDOS was working on feature phone game development. And I wound up doing some of that just before the iPhone crossing over into the iPhone phase there. And that was something that clearly did turn out to be a huge thing, although he was too early for what he was working on at that time. We've had pretty cordial relationships where I was happy to talk with him anytime I'd run into him at a conference. I haven't actually had some other people just say it's like, oh, you shouldn't, you know, you shouldn't go over there and give him the time of day or felt that masters of Doom was played things up in a way that”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“Carmac stands a little bit apart on the programming side of things, but Romero then very gracefully moved into, well, he'll work on the tools, he'll work on the systems, do some of the game design stuff, as well as contributing on starting to lead the design aspects of a lot of things. He was enormously valuable in the early stuff and so much of Doom and even Quake have his stamp on it in a lot of ways. But he wasn't at the same level of focus that I brought to the work that we were doing there. And he really did, we hit such a degree of success. It was all in the press about that. The Rockstar game programmers had.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“I met him, he was the coolest programmer I had ever met. He had done all of this stuff. He had made all of these games. He had worked at one of the companies that I thought was the coolest at Origin Systems. And he knew all this stuff. He made things happen fast. And he was also kind of a polymath about this where he could do, he made, he drew his own art. He made his own levels as well as he worked on sound design systems on top of actually being a really good programmer. And we had, you know, we went through, it was kind of fun where one of the early things that we did where there was kind of a young buck bit going in where I was the new guy and he was the kind of the he was the top man programmer at the soft disk area and eventually we had sort of a challenge over the weekend that we were going to like race to implement this game to port one of our pc games back down to the apple too and that was where we finally kind of became clear it's like okay”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“company more and I didn't because I knew people that were saying that oh companies turn to shit when you got 50 employees it's just a different world there and I loved our little dozen people working on the projects but you can look at and say well business realities matter it's like you're super successful here and we could we could take a swing and a miss on something but you do it a couple times and you're out of luck there's there's a reason companies try to have multiple teams running at one time And so that was, again, something I didn't really appreciate back then.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“It almost certainly would have. I mean, I look back at that and it's like even trying to summon up in my mind, it's like I know I was really, really angry about I am like Romero not working as hard as I wanted him to work or not carrying his load on the design for Quake and coming up with things there. But he was definitely doing things. He made some of the best levels there. He was working with some of our external teams like Raven on the licensing side of things. But there were differences of opinion about it. But he landed right on his feet. He went, he got $20 million from IDOS to go do Ion Storm and he got to do things his way and spun up three teams simultaneously because that was always one of the challenging things in it id where we were doing these single string one project after another and I think some of them you know wanted to grow”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“Mature about that. I was 20 something years old and I am and it did bother me when I'm like, everybody, okay, we need all pull together and we've done it before. Everybody, we know we can do this if we get together and we grind it all out. But not everybody wanted to do that for all time. And I was the youngest one of the crowd there. I had different sets of kind of backgrounds and motivations and left at that point where it was all right either everybody has to be contributing like up to this level or they need to get pushed out was not that was not a great situation. And I look back on it and no we pushed people out of the company that could have contributed if there was a different framework for them. And the modern kind of silicon valley let your stock vest over a time period and maybe it's non-voting stock and all those different things.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“So there really hasn't been for a long time. And even at the beginning, it's like, yes, I did push Romero out of the company. And this is one of the things that I look back, if I could go back telling my younger self some advice about things, the original founding kind of corporate structure of id software really led to a bunch of problems. We started off with us as equal partners and we had a buy-sell agreement because we didn't want outsiders to be telling us what to do inside the company. And that did lead to a bunch of the problems where I was sitting here going, it's like, all right, I'm working harder than anyone. I'm doing these technologies. You know, nobody's done before, but we're all equal partners. And then I see, you know, somebody that's not working as hard and I, and it's, I mean, I can't say I was the most.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“Arguments when as soon as people start talking about qualia and consciousness and Chinese rooms and things, it's like, I just check out. I just don't think there's any value in those conversations. It's just like, go ahead, tell me it's not going to work. I'm going to do my best to try to make it work anyways.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“a matter of coming at it from an angle. Like, I mean, for AGI, we have some of some of the hutter principles about how you can, there are theoretical ways that you can say this is the optimal learning algorithm that can solve everything, but it's completely impractical. You just can't do that. So clearly, you have to make some concessions for general intelligence. And nobody knows what the right ones are yet. So people are taking different angles of attack. I hope I've got something clever to come up with in that space.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“Decision making at that point. That's where that kind of global, holistic, top-down view can work. And I'm really a strong believer that technology should be sitting at that table having those discussions because you do have cases where you say, well, you want to be the Jonathan Ivey or whatever, where it's a pure design solution. In some cases now where you truly have almost infinite resources, like if you're trying to do a scrolling game on the PC now, you don't even have to talk to a technology person. You can just have any intern can make that go run as fast as it needs to there. And it can be completely design-based. But if you're trying to do something that's hard, either that can't be done for resources like VR on a mobile chipset, or that we don't even know how to do yet, like artificial general intelligence, it's probably going to be”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“There's a lot of opportunities for this, but it's not everyone's work a day type stuff. So everyone doesn't have to know how all these things work. They don't have to know how their compiler works, how, you know, the processor chip manages cache eviction and all these low level things. But sometimes there are powerful opportunities that you can look at and say we can bring the future five years faster. We can do something that wouldn't it be great if we could do this. Well, we can do it today if we make a certain set of decisions. And it is in some ways smoke and mirrors where you say it's like Doom was a lot of smoke and mirrors where people thought it was more capable than it actually was. But we picked the right smoke and mirrors to deploy in the game where by doing this, people will think that it's more general. We work and amaze them with what they've got here and they won't notice that it doesn't do these other things.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“Yeah, and that's where a lot of these things become possible. When you're really bringing the future forward, I mean, there's a pace that everything just kind of glides towards where we have a lot of progress that's happening at such a different, so many different ways. You kind of slide towards progress just left to your own. Programs just get faster. For a while, it wasn't clear if they were going to get fatter more than they quicker than they get faster and it cancels out. But it is clear now in retrospect no programs just get faster and have gotten faster for a long time. But if you want to do something like back at that original talking about scrolling games, say what, this needs to be five times faster. Well, we can wait six years and just it'll naturally get that much faster at that time. Or you come up with some really clever way of doing it. So there are those opportunities like that in a whole bunch of different areas. Now, most programmers don't need to be thinking about that. There's not that many.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“Approach, and we wound up with a super optimized assembly language core loop, and then a good system around it which minimized how much that had to be called.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“And make a different decision. And we kind of went through this on the graphics side on Quake where in some ways it was kind of bad where Michael would spend his time writing. Like I'd rough out the basic routines like, okay, here's our span rasterizer. And he would spend a month writing this beautiful cycle optimized piece of assembly language that does what I asked it to do and he did it faster than like my original code would do or probably what I would be able to do even if I had spent that month on it. But then we'd have some cases when I'd be like, okay, well, I figured out at this higher level instead of drawing these in a painter's order here, I do a span buffer and it cuts out 30% or 40% of all of these pixels, but it means you need to rewrite kind of this interface of all of that. And I can tell that wore on him a little bit. But in the end, it was the right thing to do where we wound up changing that raster.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source
“It's not always clear, and that's why the thing that I try to make one of my core values and I proselytize to a lot of people is trying to know the entire stack, trying to see through everything that happens. And it's almost impossible on the web browser level of things where there's so many levels to it, but you should at least understand what they all are, even if you can't understand all the performance characteristics at each level. But it goes all the way down to literally the hardware. So what is this chip capable of and what is this software that you're writing capable of? And then when this architecture you put on top of that, then the ecosystem around it, all the people that are working on it. So there are all these decisions and they're never made in a global optimal way, but sometimes you can drive a thread of global optimality through it. You can't look at everything. It's too complicated, but sometimes you can step back.”
2022-08-04 · Lex Fridman Podcast · #309 – John Carmack: Doom, Quake, VR, AGI, Programming, Video Games, and Rockets · IDENTIFIED FROM THE TRANSCRIPT · source