Latest / Beat The Odds with Dre, Rodney, & Brylan / Elon Says AI Can’t Be Stopped, Meta Faces Its Product Reckoning & The Hidden Censorship Debate
Transcript
- Dre: Yo guys, what's up? Welcome back to another episode. Happy Wednesday, man. How y'all doing? Rodney: Happy to be happy to be back. Happy to be back. We recorded one on Sunday. That was a good one. I think that was one of our l longest podcasts that we've had in a while. Yeah. Dre: It was, it was. Rodney: Because probably because it was no, it's a Sunday. We had time. And for those of that don't know, we usually record at the end of our day work days usually. So typically that's why we keep pretty tight. But that one was that was almost two hours. Dre: Yeah, you and Brian, y'all had a different energy on Sunday. Yeah, Rodney: you think so? Dre: yeah, different energy. Good, good, good debate. Good debate, man. Brylan: Man, we gotta bring it, man. I I Rodney: Sh Brylan: actually hold up, man. We gotta bring it from the group chat here, bro. We still get we need to keep this back. Dre: You Rodney: We we can't keep the vet bro. But but but Brylan: That's crazy. That's crazy. Rodney: we we can keep we can keep it we can keep it in spirit. The only the my my main thing was was just that was just that the term shifted. That that that was my main thing, bro. That that's it. That's it. And I understand that you s I understand. I Brylan: the t I feel like the terms became more clear. Rodney: But see that's the thing though. That's the thing, though. Really, Dre: Yeah. Rodney: really, the thing is, we had talked about clarity of turns before. Cause at first they were crazy vague. They were like crazy vague. And then, and Brylan: Where we s where we where we had originally started were great. Rodney: yeah, when we started off, it was crazy vague. So then we talked about that and we made it more clear. And then, here's the thing, Bryan. As soon as the narrative shifted, now you want to get more clear on the verification Dre: you Rodney: of the deal. Look, look, look, we should keep the bet in spirit. We should really honestly keep the bet in spirit. Cause I I I still hold strong and it seems like used to hold strong too. But let's take the finance part out of it. Brylan: Yeah, I yeah, my position changed. Rodney: Yeah, let's take the finance part out of it. but but I think the spirit the so Brylan: Dre should define it. Rodney: Yeah, Dre Dre don't want to get into it. Dre don't wanna get into it. Okay. Okay, but but but Dre but Dre, Brylan: I think that is fair. Dre: Not what this thing got under doing me just between y'all two Think I'm gonna do it to get not I think I'm gonna do me you think I'm gonna do me This is me. This is between y'all two. I'm just I'm just sitting on the sideline. I'm just I'm just an observer Brylan: Ha ha ha Rodney: don't you don't you see don't you see what w where I'm c where I'm coming from? Dre: No, do. do. Brian was really good at that. Brian, you're really good at that. You're really good at Rodney: Right, real smart with Dre: really, really good at twisting, really twisting it the way you want it to be. You kind of like a lawyer. You're kind like a good, a good lawyer. Rodney: Yeah, just just Yeah. Very good. You have to you have to wa you had you have to really I had to focus. I was like what is what what what is he saying? Nah. Brylan: No no no Dre: He'll Brylan: no. Dre: dump a bunch of facts. It be facts. He'll dump a bunch of it on you and just have you sitting there sitting in it. Rodney: Right. Yeah. And he'll but is that not reasonable? It's like it is re it is reasonable. Dre: Hahaha Brylan: Yeah. Rodney: But that's not what we're talking about. Dre: Hahaha Brylan: I was at the gym keep like and you kept sitting there, you was bro, focus, focus. I'm like, bro, I'm in the gym, bro, trying to I'm like can Dre: Hahaha Rodney: Yeah. Brylan: we not just lock this in? But all right man, alright Rodney: No man. Brylan: man. Rodney: But actually that we do. Dre: Look, we do have a good follow up though to Sunday because Sunday we were breaking down the letter and AI that kind of circulated that was signed by what I think like at the time what 25, 30 of the most important AI chip companies. And we were having a conversation because Dario hadn't signed it. And we were like, all right, well, it's the weekend. We gave him a little grace. We were like, it's Sunday. It just came out, Rodney: Right. Dre: I think Friday, Friday afternoon. We were like, all right, maybe he's a slow reader. Maybe he need the weekend to get to it. And Rodney: Right. Dre: what did he come out with? Monday or Tuesday? I think it was Monday, actually. He responded. Brown, since you're the Dario fan, you want to break down that response? Brylan: Yeah, man. basically, you know, his whole i I think it was expected, right? Like if I'm gonna be honest, that he was gonna come out and be explicit about him not you know, really advocating for for for banning like open weight models. and I think, you know, he's been saying this for years now, but his primary his primary concern has really been like always against China. And I think despite the you know, the tension between him and like the Trump administration. I think that's one way he's just been trying to like gain some brownie points with the administration is always being super like, you know, against against, you know, China just having a a position here here in, you know, the United States, which obviously this whole narrative around open weights was triggered by a Chinese model. and then yeah, like I think his his old you know, in the whole spirit of anthropic, you know, and honestly it might transition a little bit into you know, the the new letter that, you know, we're gonna talk about that came out too. But, you know, once again safety. And so those were t his two biggest things around like, you know, China should not really be be leading, you know, the AI race and you know, we should not be enabling it, you know, in terms of, you know, selling chips to them and things of that nature, but also allowing them major access. Here in United States because you know, without the his whole thing is guardrails, these older models can, you know, just run wild, which I do think, you know, some people given the recent like open AI hug and phase situation that we talked about recently, some people may be in alignment with him, like given the increasing capabilities of these models. And then yeah, I know you know we were talking about this in our past conversation around like is distillation really a threat? and once again, distillation just ties back to China. And so I think he he he came out and instead of you know addressing, you know, them not signing like directly, he just took it as an opportunity, not really to state his position, if I'm gonna be honest, but really to make an argument against China, if I'm gonna be honest. That was kind of my takeaway from Rodney: Yeah. yeah. I think Think everybody's in agreement that nobody wants China to surpass, and nobody wants China to beat us, and whether it's Trump saying that or whether it's Sam saying that, or anybody else, or Jensen from Nvidia saying that. Everybody's in agreement with all those things. So I think that Dario successively said nothing. I think he didn't say anything. he just doubled down on his position. His position is he's just to me, I read it as he's just justifying his his current position, which is that. AI development, open weights and whatnot should not be encouraged because of how dangerous they can be. And I think his lack of signing the letter is also Proof that that's how he th he feels alongside that letter. So we don't have to guess about that. We were guessing on Sunday. We're not guessing now. We know he doesn't support open weights. And he's using that, these safety concerns, as the reason. Now, we can only speculate what the real reasons are for why he doesn't want open weights. But I think it's kind of cheap. And kind of convenient to say it's because of China. Because I don't think open weights and open source is going to stop China. I don't think it's gonna stop China at all. And also I just want to remind people and listeners that the government does not have the ability to control this technology. If you understand it from a technical level, this is actually very similar to encryption. When encryption first came out in the nineties, specifically PGP encryption, Brylan: Yeah. Rodney: which I had a a large background in, there were it was a very powerful technology because you could essentially turn plain text into a bunch of BS that nobody can read. And nobody can crack it, essentially, without a secret key. And and in some cases a secret key with a pass with a passphrase. So The government attempted to try to step in and control it, but they couldn't control it because the protocol for cryptography had already spread. And so I just want to remind Brylan: Yeah. Rodney: people that on the surface, it seems like the government can come in and control these things. But that's likely because people are not fully going down to the granular. Technical level. In order to control AI, they would have to coordinate with Microsoft, Google, Apple, etc., and push hardware updates, hardware updates, hardware updates, so your iOS updates, your Android updates, your tablet updates, etc., to prevent these things from running Dre: you Rodney: on local machines. And that would not and additionally, that would not stop virtual machines because they don't have to update. That would not stop people in general who don't want to update. It would be very, very challenging to control the models. It seems like you can control them, but you cannot. Essentially, what I'm saying is that the research is out there, the open source is out there. Kim EK3 came out with open weights. So so the research is out there. And yeah, there's a little clause about if you make $25 million that they don't they want a piece. But let's be honest. That's not gonna be that that's not gonna that's gonna be a small fraction of people. so it it's really not gonna include most people. So so I think that I understand Dario's position as a CEO. As a CEO, I understand his position. However, as somebody who's not currently a CEO, I'm my my my criticism is not on what a CEO has to do. To keep this company and get its investors paid and make sure this company survives. My criticism is, Dario, shut up. Dario, just shut up. You are just like Sam. You left OpenAI. Elon said himself, Anthropic would not exist if it wasn't for the fact that people did not like Sam and did not trust Sam. So therefore, they therefore there's anthropic in Dario. But they're all the same. But Dario positions himself as if he's not the same. I don't like that. I don't like that. I don't like that piece. That's the piece I don't like. Dario, you are Sam 2.0, and you might be worse. Because You rush forward and and you know what I love? Zuck came out and said, Let me just pull up real quick what what Zuck said. Zuck said Zuck became my my new favorite CEO. And by the way, everybody's hopping on my Twitter right now. Justin made his first post, probably the biggest biggest po first post ever on on on X. He gained Dre: you Rodney: nearly two million followers within less than a week. He's probably at 3 million now. But Zuck said in his post it he the game on X is heating up. Sam is on there. Dar all these Dar all these people are on there. Now Zuck is hop popping back out after after a hiatus. One thing that he said in his letter was. It is surprising that the discourse for many of those who are developing AI is so filled with doom. For example, Dario. I don't understand why anyone who believes AI will eliminate most jobs and much of humanity's relevance would rush to build that future. AKA Dario on Oprah. The notion that AI is so dangerous that only The safe path was an extreme concentration of power seems dangerous. Sam criticized this in a recent interview about an unnamed company wanting an authoritarian control over AI. Historically, hoping that absolute power will benevolently benevolent benevolently provide for humanity if sufficiently. Enlightened hasn't led to safe or positive outcomes. So essentially what Zuck is saying is it's quite interesting that you believe that AI is gonna wipe out jobs and people's relevance, but yet you push forward full speed ahead. And I think that is fair. That's a fair criticism from a peer, not just somebody on the sidelines of AI. That's a peer saying, hey, look, basically politely saying, you're full of shit. You're full of shit. I see through it and we're not buying it. But also, too, Zuck has something in the game too, because he has now invested hundreds of billions of dollars into AI. These companies, Google that have signed it, etc. Meta, they've invested so Dre: Ahem. Rodney: much that they're like, hey, hey, hey. Don't turn back now, guys. We don't don't turn back now. Like y'all, y'all heated up this AI game, and now we don't deployed hundreds of billions of dollars and we got things on the line. Don't turn don't try to control it now. Don't get scared now. That's that so that's a part of it too. So I don't want to make it seem like there's not an incentive on the Zuck side or the hyperscaler side. But that's my piece. Brylan: I wanna I wanna I wanna say two things because I actually disagree with you on two things here. the fur yeah, the first the Rodney: Sure. I'm not surprised. I'm not surprised. Brylan: f the the first one is Rodney: What do you agree what do you agree with? Brylan: No, I I I I I I definitely agree with you in terms of Sam and Dario being one of the same. Right? Like I think I think how they're how they package it up may be a little bit differently. but yeah, they're they're they're they're they're they're definitely running their companies the same. and I think, you know, yeah, the issues that they bring up are relatively the same. I think where I disagree with you on is I think the narrative that America has to win the AI race is going to be something that will not outlive the Trump min the the Trump administration if you get if you get someone else in in in office. Say for example, when 2028 comes, if JD Vance ends up becoming president or anyone, you know, Marco Rubio, whoever is a part of that existing kind of, you know, administration, I think that that narrative will continue because that is part of their strategic, you know, kind of like plan. They want to reindustrialize America. The whole thing is like this whole idea of like it's called Hamilt Hamiltonian economics. and Scott Besson he wrote a really good Wall Street Journal piece on this, in terms of how how they're thinking about it. but the whole idea is that, you know. If you do not make things and if you do not use the things that you make at home, your economy over time eats itself. Right. And that's what, you know, outsourcing all this stuff to China, if we outsource AI to China, will continue to happen. I think if you get someone else in office, like a Democrat, and I'm not arguing for any party or whatever, I'm just saying I think that narrative is going to shift a hundred percent. And so my long where I disagree with you is I think in the long term, I I Going back almost to your point that you were that you were making around, you know, like cyber and the fact that like it was out there and no one could really control it, I think that's gonna that's it's gonna take time, the the the same amount of time that it's taken people to get around to today that open source is a real thing and it's valuable. I think it's gonna take a few more years and a change in like, you know, narrative for people to realize like China is not gonna go nowhere and you can't do nothing about it. And ultimately I don't think people are in the long term going to care about a a US race versus a China race. and then yeah, the the the second thing I would actually disagree with you on too is Zuck. I think Zuck is just trying to take advantage of a moment. But if you if you pair Meta with Amazon, Microsoft, Tesla, all these companies that have done major layoffs since twenty twenty five. Those other companies have laid off around ten, ten percent or less of their staff. Meta is the highest at twenty percent. And it was just back in what whatever it was, March or April, when they made that announcement that they were then putting AI on every meta employee's laptop to track what they were doing, to be able to kind of get that data to once again train these models, et cetera. That they did the largest layoff that they that they've done, and it was literally framed as like making the company more efficient and tapping into this whole existing narrative that, you know, I think we even saw block tap tap tap into this. Like, we're we're laying off because AI is making this more efficient. And so I actually call BS on on Zuck. I think it's a short term opportunistic window for Meta to, you know, really position themselves or reposition themselves because over the last few months they've been dropping the ball, especially, you know, with the whole super intelligence. they made billion dollars investments into these researchers and they dropped this model and where are they? You know what I mean? Like nobody nobody has like really heard from them. And so yeah, those be would those would be the two things that I would dis disagree with you on, bro. Rodney: Okay. Dre: Yeah, I didn't, I didn't get the takeaway from Dario's letter or response that he was closed off to open models. To me, it seemed like he was more in line with wanting less capable models to be open or at least for it to be more controls over these open models. It sounded like, cause what what I, what I took away from it is that he'd think these can be a public, Brylan: Yeah. Dre: he'd think these can be a public good, whatever that means. I don't know what that means to him. But what stood out to me is, again, he stood down. I mean, it's nothing new, no new information here. I mean, he doubled down on what he believes about China, which I don't necessarily agree with. I think China has shown that it is incredibly resilient. And so banning these chips, banning this stuff doesn't do anything but give them more firepower, more motivation, more everything to just develop their own. And at some point, they will catch up or exceed. So I don't think we're actually doing the good that we think. I tend to lean more towards Jensen's perspective on how we should work with China. And we should just give them, we should just make things available and Rodney: Yeah. Dre: really reduce the friction and blockers that we have there. Because I think, and I find it interesting too to see China's position on it because... Let's not forget that like China produces a large quantity of the products and goods that we use in America. And so while as America, we're always trying to slow down China's progress where it doesn't appear to me that China is necessarily trying to slow down our progress. Right. So it makes me wonder, like, are we, are we even running the right race because China controls, what is it like 90 or 80 % of the rare earth materials? They go into these chips. That's not forget about the impact they have over Taiwan. I mean, let's be honest. If China wanted to really slow down this AI race, I think they could slow it down. I think they could really, really slow it down if they really wanted to. So I don't know, man. Sometimes to me, it seemed like as America, get, we lose focus. We lose focus by throwing these, these conversation, I'm calling it conversation starters into the, into the pot. that really don't even seem to be relevant. You know what I'm saying? So I don't know. I think we should work with China a little bit differently than the current way that Dario kind of outlines it. Again, I tend to lean towards how Nvidia thinks about it. But again, I understand he wants to sell more chips. And then on the Zuck thing, I'm a huge fan of Zuck, but I call BS on that letter as well, especially on the statement around, would you want to build that future? Let's not pretend the products that Meta makes are really good for the world. Let's not pretend that they were unaware of the impact of their products. They were very aware. They've always been aware and they've continued to double down, expand, make these products more engaging, increase distribution over and over and over again. clear understanding of exactly what their products are doing. that's, mean, and there's many ways to slice that, whether you look at the impact that it has on youth, whether you look at the impact that it has on other countries, it's literally started silver wars. People have died. because of misinformation that has been spread on Facebook. So, I mean, and yet they keep building, right? So I think that that statement is just a, it's just self-serving, right? I agree with Brian. He's just trying to capitalize on the moment, which I don't blame him, but I don't, Zuck has no right to question anybody else's moral intent around how they run their business. He should not even be the person even bringing that up. know, let Jeff Bezos bring it up at Amazon. I'll take it from him. over the guy that decides to keep building products that are harmful to children and other people. Rodney: Yep, for sure. I I I see Grass's point on Zuck. I think Zuck started off naive, but to Dre's point, at some point they started to know the effects. And but I think it it went from naive to denial. So it went from naive, Zuck, no. I s we see effects, but how do we know us from Facebook? How do we know us from our products? And and then it went to they researched it and they found out that yeah, it was due to a lot of the products and there's a clear correlation. And then it's like denial of the facts. No, it's g it's still not that. And so there is So I understand his credibility shot. And I also think that Sam has thrown shots at at Meta as far as in the past mentioning companies that are more focused on taking your attention away as opposed to bringing your attention towards productivity. Which I can only imagine are social media companies like what Zuck builds. in whether it's Instagram, Facebook, Threads, etc. So I think I would agree with you guys on that. At the same time though, I think that could give him an edge on wisdom, meaning that you start off naive, he started the company when he was in a dorm room, he went off running, had no idea that it was gonna create all these problems. Then when they first started creating problems, was was was naive to the world, then the denial part, and then now he's maybe more at the acceptance part. And so wisdom comes through experience. So it's it's possible that there's some wisdom behind what he's saying, but don't get me wrong, there is something that it's Serving him in his message. But I was mostly focused on his excerpt about these guys. These guys are not college dorm room kids that knew that that didn't know anything. These guys are not people who didn't already think about the crazy dangerous things that could happen. I really don't think that with social media anybody thought about like we're seeing dangerous effects and now we have the evidence and more evidence keeps piling up. And at some point we may even see social media be regarded as bad as cigarettes or worse, as far as mental health, as far as what it takes away from your life over time. But Dario and Sam are not kids. In fact, they have thought already, hey, look, imagine you're going into it knowing, yeah, I'm I'm I'm gonna I'm gonna I'm gonna we're gonna knock out a lot of jobs. So I understand the messenger part. But imagine going into it, yeah, we're gonna knock out a bunch of jobs. We're gonna we're gonna knock out everybody's jobs, it's gonna be the highest unemployment rate. Yeah, it's gonna be this, it's gonna be that. Zach never did that. Zuck never hey, yeah, it's be bad mental health everywhere. Yeah, people are gonna waste most of their life on our platforms. Yeah, people are gonna end up being losers and and spending 70 days of their year scrolling on their phone. Zuck never said that. I don't think he ever thought that. Now, the business incentives of the business ended up being that, but these guys came out and said, hey, yeah, yeah, yeah, we're this is what we're gonna do. This is what we're gonna do. And it's ironic to me, at least, that Some would argue that They're now trying to s and we we don't know anybody's intent, but it seems as though they want to. make it more difficult for competitors. And one of the things in Dara's letter that he talks about is bioweapons, right? And it's one of those things where it's hard to deny. It's like, okay, well, if you're so concerned about bioweapons, then how how can we deny it? But that's that that's the same thing with encryption, right? So it's like when it when it with it when when there was a debate about encryption and the government was getting involved, et cetera. They're like, well, what if what what if a murderer or a pedophile, etc. etc. uses encryption. What then what? Then then then what we do? And so what happens is is that people end up taking a technology and then jumping to the most extreme possibility that you really can't guard against it it you you really can't guard against and if you thought about that with a lot of things that that we have today for ex a gun would be one fire whatever it is you wouldn't you you wouldn't make it because you say what if what what if somebody uses it for this horrible evil purpose what if somebody uses it and so there's this book Called The Beginning of Infinity. And Sam recommended it a hundred times over the last few years, but I decided to read it and it's counterintuitive to most people in the way that we most of us think, but sometimes the best way is simply innovation, focusing on innovation. And building forward and building out and is not Minim the idea of minimizing risk does seem attractive and it does seem responsible. Therefore, that most people are going to say it's a reasonable. But is it really reasonable to have a scenario where You now have these open protocols essentially in the public, and the models that are already out are already dangerous. They already have bioweapon capabilities. They already have cybersecurity capabilities, especially if you are a bad actor. If you're a bad actor, you're already not being stopped. And the argument might say, well why well why increase the intelligence? Well, increasing the intelligence gives you the b ability to potentially defend against. And so we've now created this cat and mouse game. And to be honest with you, it seems as though we're gonna want to move forward as fast as we possibly can. and we're not gonna be able to stop the momentum. And I think that that understanding that this momentum which seems to be nearly a trillion dollars in this AI infrastructure build-out. I don't see how the pacing of AI will will go in practice. So part of this petition, so there was a petition where about eleven hundred employees across OpenAI, Anthropic, and even DeepMind signed to slow and pace AI. But I understand why. Because they are scared. But how about this? When we all think about our lives, thinking about a time where You had a goal, a a big goal. And you started moving towards, moving towards it, moving towards it, moving towards getting momentum. And then once you build up a lot of momentum, it kind of just keeps going. And essentially, now you're in a situation where it's like, do I embrace what comes with that that ambition and that goal? Do I embrace that? And try to figure out how how to handle that and have the confidence to know that I can find a way around the challenges that ahead are that are ahead, or do I say, wait, let me let me just let me slow this down. Let me let me let me slow this down. and that may not be a hundred percent the best analogy, but I think that's a decent analogy as far as I think these people built something. And when they lost control, not only from a competitive standpoint, but also with the hugging vase incident, those two things, amongst maybe some other things that we don't know about. Suggest that they might say, no, we can't control it. But it was arrogant to think that they could from the beginning. There was arrogant to think they could from the beginning. So don't get me wrong, safety concerns are hard to they're they're they're hard to negate. So, for example, when we're talking about encryption, if somebody says, well. Where somebody is human trafficking with encryption and we can't read it, it's it's so hard to they successfully put you in a position where you can't debate it without looking like you're you're going against you see what I'm saying? So these are very, very specific nuanced tactics that people use to negate around the fact that ultimately the loss of control of being able to read people's messages. Is the concern. Yeah, there might be human trafficking, but is that the primary concern? Or is control the primary concern? And I think that when you look at a lot of examples, typically control is the primary concern, but is wrapped in safety. I'll give you one more example before I we I once had a girlfriend who I never share my location with anybody. Never share my location with anybody. Cause I of why I built a privacy and security company. I don't share my location. And This girl in particular wanted my location. because I had her location, but I had her location, not because I wanted uneven. It was just a matter of she already shares her location with other people, right? So if you already share your location, that tells me that location sharing is not a problem for you. So I might as well have your location too. At some point, this girl comes to me with somewhat of an ultimatum and is like, I want your location, otherwise, I can't share your location. Sh share share my location. And while I say, Well, I don't share my location. And I say I don't understand what what exactly, and it's just the reasoning was, well. Every time you leave, you know, I'm always worried about you. I'm always worried about you. I'm always worried. I don't know what you're I don't know what you what might happen and this and that, etc. But I want you guys to use your common sense, guys. I guess I want you guys to use your common sense. Was this girl worried that I was gonna die every time I I left outside of the house? Or Did she just simply want to be able to know where I was at at her convenience? These type of arguments around safety and control are very, very difficult to address. And that's the point I'm trying to drive home. Dre: Yeah, I think both of those things can be true, right? I don't think it necessarily has to be one or the other, right? I think it be about both. Maybe it is a matter of safety and control, right? And I think, and this is just my perspective. I think when you build something that is big enough and meaningful enough, I do think you deserve the right to have an opinion about how that thing should be used. I do think you deserve a right to be able to dictate the rules, sort of say, right, to some degree. You built it, you contributed to it, and I think that's fair. as things grow and they start to take a life outside of what you build, then obviously your ability to do that diminishes. But I don't think, I still think you deserve a seat at the table. And so, you know, I think what's happening right now with these AI leaders raising the safety concerns, I actually find it to be a more responsible approach than any of the predecessors in some of the big tech now. technological moments in time, right? We just talked about Zuck. I don't think at any moment did we ever hear Zuck come out and say, our products are harmful to teenagers. We know they're harmful to kids. We know they cause wars in other countries. We know we manipulate the population. We know we do that. We know we do that. But we just gonna keep building anyway. And we want more regulation. I didn't hear any of that from Zuck. In fact, I heard silence. All I heard was head down, execution, build at all costs, grow, build things, what was it? Break things fast, like they had like a little saying, build fast, break things. Like that's all it was from them. There was no thought about safety. There was not even a conversation about safety. It was at any cost, build. And I think at some point when you get to a certain scale and you build something meaningful enough, in enough people hands, I do think it is the responsible thing to have the conversation about safety because we see what happens when you get products at large distribution in people's hands and you don't have those conversations, what that can translate to and how that can change society. Social media has changed society. It has changed the way people look at themselves. It's changed the way people communicate. It's changed the way certain countries govern. It has changed society. So the conversation about safety should have probably happened with Facebook. 20 years ago and it didn't. And so I think what Dario and Sam and all these leaders talking about the implications of their products, I don't even think they're talking, and maybe this is just bad communication on their part. I don't even listen to what they say as too much into the current moment. I think about, and Brylon kind of laid this out beautifully, think a couple of pods ago, when he kind of said, if they're successful in the objective and the mission that they have of creating AGI or creating these super robust autonomous systems that are capable of hacking systems that are capable of lying, that are capable of getting behind things without being detected. If they're successful in that, I think it's crazy to think that these tools cannot be harmful. I just find it difficult to believe. How can it not be? Now, what that harm looks like, nobody knows yet because we're not there yet. But to me, think raising the flag to at least say, Yo, the where we're going, even if it takes us 15 years to get to that point in time, where we're going is dangerous. Where we're going with these tools, the capabilities of where these tools are going as they get more electricity, as they get more compute. We also have to remember that there's that there's things that they're testing in a lab that we don't know about, like the hugging face thing. I'm sure there's probably other instances of that that we just simply don't know about. And so I think these are real concerns. And then on top of that, I think, too, they're also getting pressure. from the general business public, right? Some of these leaders that run these massive companies want to know, how is this shit going to implicate my data? Because what will happen is if big companies have data problems, have security problems, whatever the case may be, the executive team can be responsible for that. And so I think a lot of these corporate leaders want to understand the implications of these tools. And so I think it would be disingenuous if they are asking those questions, which I know they are, for Sam and Dario to sit in their face and be like, huh, we're good. Don't worry about safety. Don't worry about that. These products are good. It's just like a gun. It's just like everything else in society. People, we could just trust people. People, everything will be good. I just think that's, I think that's an irresponsible take. They're building something that is important. They're building something that is obviously hit scale. I think I just saw a stat the other day. The chat GPT got to a billion users a week. These are real tools Brylan: Which is crazy. Dre: that, crazy, right? That are in people's hands. And unfortunately, not everybody is sophisticated and as logical as we are. Not everybody thinks through everything as we do. Not everybody has the same moral alignment and compass as we do. And so I do think at the broader level, this conversation about safety, about harm, about all these tools, I think is fair. I think it's fair and they're the right conversations. Now, whether or not we're right or wrong, only time will tell when we look back and we wonder, did we do too much or did we do too little like COVID, right? But when you end the moment and shit is firing all off, I do think it's responsible Rodney: Yeah. Dre: for somebody in the room to be the adult and be like, wait a minute. Is what we're building harmful? Are we aligned correctly with society? Are we building something that's not just useful to society and valuable, but it is safe for society to actually adopt? Are we building something that's going to hurt the broader business environment, our partners, who we need to partner with, who we need them to trust us? We need to. build trust in the business community to justify the valuations, to get to the trillion dollars in revenue. Like we need those things. so again, whether or not Dario's doing it right or Sam is doing it right or whatever the case may be, but I stand 10 toes down. These conversations about these tools being harmful and about safety, I think they're the right conversations to be having. I think we should continue to have these conversations. We do need to talk more about what regulation looks like and what. the right support and right balance between the private and the government, the private sector, the government needs to look like to make sure that we're building something that's important. Because I think the last thing we want to do is look back 10 years or 15, 20, 30 years from now, and these tools have been incredibly harmful. And it's like, well, shit, all we had to do was have the conversation. Nobody wants to talk about safety. Nobody wants to talk about the capability. Nobody wanted to have that. Everybody just wanted to push for growth. And we didn't think enough about like, what are we actually doing to society? And I think that's an important question to ask. And again, I wish Zuck had asked that question more because now we look at it and we realize what the answer is. Rodney: Yeah, z Zuck Zuck Zuck messed up. He messed up a lot. The the world's not the same. Like the the the I've noticed. I I remember going to college and you know we had Instagram, but it was most of my interactions were normal and regular. I didn't know what was going outsi going on outside of Brylan: Yeah. Rodney: my school and unless I called my other friends when you were at Central and other friend I I I I I I didn't like and we i it was it was a different way of connecting and communicating that I felt that I feel in hindsight was kept me more engaged in real life than what we see in the social media world. so I I agree with you. Another thing I want to say is is that is that I think people are questioning the frontier laughs because I believe Deep Mind was concerned about safety, which is why they when they first tried AI experiments way before it's Chat GPT and they had any blowback, they pulled a Mac real quick. And I think it was Sam that came through and said, I'm gonna go fast. And then Dario was a part of that team, et cetera, et cetera. And then at some point Dario was like, you know what? Nah, I'm gonna let me just make my own thing. And then now they're both going fast. But I think that the going fast, I think that There were people in the room building AI that were approaching it from a more safe thing or safe way. And they chose not to do that when they joining A OpenAI was a vote towards faster AI. It was a vote it was a vote a vote towards that. But then I also say the second thing is is that I agree with about about safety being the responsible thing. I think what I'm mostly objecting to is the proposed solution. The proposed solution is to slow down the pace. And I'm not s sure that that's the correct solution. I think that there's a conversation to be had on the solution. But one that comes to mind for me is when I see threats, typically what I'm trying to do is instead of slowing down my momentum, slowing down what I'm doing, what I'm trying to do is Create actions and Systems. and other tools so that I'm prepared for the fears that I see on the horizon. And what I've found is that since I've been doing this, I'm always prepared And I think that there could be a different approach other than slow down. I think that slowing down is the is the most obvious top of mind. Hey, it's getting crazy. Let's slow it down. I think that that's one solution, but there c is some other solutions in the room that we should consider. I think that they could could Take a lot of those fears and continue their pace. But outrun the the issues that they are currently seeing. And and and and to your point, there is a lot of things going on in Frontier Labs that are also adding on to their fears too. So we don't, as a public, we don't know everything that they're actually worried about. because they've only Dre: Hmm. Rodney: released so much capabilities. They they do have stronger capabilities internally that we don't know about. So so Dre: I think the labs just don't know what to do right now, which is fair. Let's just use our logical brain here. It's fair to just accept. And I tell my teammates all the time, it's OK to say you don't have the answer. But not having the answer doesn't take away from the fact that the conversation still needs to happen. We still got a solution. We got a solution towards a solution, even if we don't have the answer today. So I think the slowdown. is an honest response to, we don't know what the fuck to do. Stuff is changing every day. We're fighting on multiple fronts with distillation, competition from China, constraints on energy. We need capital. It's new labs popping up all over the place. The reality is, there's not like this one or two companies that are, it can just be a, let me just pick up the phone and call. And as long as me and you align, That's all we need. Like it's just me and you, cause we the only ones in the market. No, like there's a lot of people in the market. And so when when that's happening, it's hard to get consensus. And so I think slowing down really is, I mean, from just from a logical standpoint to me, it makes the most sense to slow down because right now they don't know what the fuck to do because anything they, Rodney: No, no. Dre: any, any choice they make right now is more than likely probably going to be wrong. Right. Any regulation that they do is probably going to be wrong. any restrictions like Dario is trying to say in terms of like banning stuff to China. I personally think that's wrong. Right. So to me, it's so early in the race that anything we come up with is more than likely probably going to be wrong. Right. And to go back to analogy, you like to bring a Ronnie's gun. It's like, OK, if you take an item in a store and you compare it to a gun purchase, which one is slower, purchasing a bag of chips from the store or purchasing a gun? Right. You still can purchase the gun. Right. But there's going to be some things that going to slow you down. Right. because of safety, because of the checks, the balances they need to go through to make sure that this is legit. So to me, slowing down, again, I'm not saying that's the only solution or the best solution, but given the circumstances of the market, I mean, it's not the most illogical. Rodney: Okay, okay. Brylan: Hold hold on, hold on. Let me let me let me let me let me jump in here. Rodney: Brian, I'll I I'll give you one second. If they're serious about slowing down, ask for ID. So you s a it's a and and show it. Show it. Ask for ID. 'Cause that's gonna slow down the growth numbers. Start doing that. Start all that talking. Don't go to the government. Start slowing down in your actual actions. And also, it's just my opinion that slowing down seems like the logical choice. But I don't believe that it is the the best choice. I believe that slowing down seems like the path of least resistance. we just if we just slow this down, it will go better. That's what that's the that to me that's the lay that's the lazy logic. If we just slow this down, it it's gonna go better. And I understand where people are coming from. But when I read, you know, these hacker news 2000 comments. And even just one question says, okay, cool. You guys want to create this internet potentially go to the government for this international agreement? And if okay, if China doesn't agree, then what gets slowed down? If China doesn't agree, then what gets slowed down? If China does not agree, then what gets slowed down? It it it I I I think that that it it it it would be different if we truly had AI dominance that we could protect and that nobody else could access and nobody else could build. They they do not need us right now, they do distill models. I do not believe that they need our models going forward to create new models. And and I don't believe that they're going to sign or get along board with this international. goal of slowing things down. And because I believe that, I don't believe slowing down is the answer. I believe it's the most obvious thing. we should just slow it down. Yeah, that seems obvious. It's almost so it's almost so obvious that it's like, well, why not do it? Because it's the no, it's it's not that simple. I don't think it's that simple. I I I think it's but that's just an opinion though. It's really just an opinion. I think that slowing down seems like the best solution, but I don't think it is the best solution. But yeah, go ahead, Brownie. Brylan: Yeah yeah. No, I I I honestly I think both y'all have been saying some really good points. your your your last point around like would China consider slowing down, yeah, I'm actually in agreement with you. I I I don't think they would until they felt the consequences of not slowing down themselves. Right? i i which which which which is a reality. I think if I'm gonna be honest with you, I think language really matters, right? Language really matters. because I am in agreement that like, you know, I think both can be true in terms of, you know, the conversation around safety or control. But I think what is tricky about those is it naturally just introduces politics into it, right? Where both there's gonna be multiple sides that have differing opinions. But what I do like about the word choice that they use this time around around pacing, that's very objective, right? the thing that I hope they do. is I think they need to shine a lens on the inputs rather than the outputs because whenever we hear pacing government or whatever, all we're judging is what we're seeing, which is these model releases, the commentary behind it, et cetera. I think the thing that is very important to understand is if you think about this from the position of like I'm in one of these labs, OpenAI or Anthropic, what does the model cadence look like? So typically the way it works is if you see them come out with like GPT five or Claude Four or whatever, that means that is the base model. And every time they increment it to Claude, you know, Opus four point one or four point two, like I looked it up. Basically, and I think this is what they're what they mean when when when they say pace, is yeah, they they launched Claude Opus Four. In May of 2025. So literally, what is that? 14 months ago, right? Which is crazy. They were update the next update, so Opus 4.1, which means the base model stayed the same. They just did reinforcement learning on top of it. They did that in 75 days. They did that in 75 days. The next upgrade literally came in less time than that. So we've all talked about. What Elon has been able to do with data centers, where it took him 120 days to build something physical, and then he was able to go do something in 90 days, and how impressive that is. They were doing that the whole time as they were incrementally upgrading this model until they got the next base model, which is Opus 5, what they which which they just rolled out. And the reason why I lay this out is because if you think about it this way, the narrative that has really started to pick up. Over the last year, there's been a few of them. The first one has been reinforcement learning. Reinforcement learning is what you do after you've trained a very capable model. That is very important because you can do that whenever. As long as we have new training data, aka usage, we can take that data and then train the next incremental model. That's very important. The second piece, which I think compounds that, is who did anthropic go hire a few months ago? Andreas Karpathi. Why? For recursive self-improvement, which is these models training better models, aka speeding up development. And I don't know if y'all saw this today, but one of the co-founders of Thinking Machines, she left thinking that's a whole mess. We can get into a whole conversation, but she left Thinking Machines to go back to OpenAI to run their RSI team. And then Google just announ or DeepMind just announced that. The team that originally created Alpha Fold, they split them up because they're trying to reorganize around this world where the models are now doing most of the research and creating the next models. And so from that lens, when I think about pacing, I think they're saying, okay, if on average the next incremental model was coming out 60 to 75 days, theoretically, if reinforcement learning and recursive self-improvement. Help us achieve what we expect it to achieve, that is going to go, let's say, from 60 to 75 days to let's say, you know, 20 to 30. So every single month we're getting a new upgrade. And the thing about it is, if that's true about American labs, that's gonna be true about China. And if that is open source, I think once again, you don't know what the fuck to do. Like you genuinely don't know what the fuck to do. And I think that That that would not only end up being true about America not knowing what to do, but I think eventually, because right now China is just focused on dispersion. They're just trying to get it at, you know, make sure their technology has a say in the fight. But let's not act like these models aren't being dispersed across the billions of people in Asia as well. They would eventually feel those consequences as well. And so, yeah, I don't have much comment around like what safety should look like, what Should look like 'cause I don't know if that is the right conversation to like I do think it's a conversation to be had, but I think it's one where like without understanding what they mean by pace, it's hard to understand why they feel like they may lose control. Because and so that's that's that's my Dre: To me, but to me, to me though, the pacing and China argument are two different things though. Right? Like to me, in my mind, they're two completely different things. Like the conversation about building products that are non-harmful and they cannot be maliciously manipulated to me is a whole different conversation about trying to slow down China and trying to compete with China. Like those are two different conversations to me. So to me, if what you're saying is true, then to me, again, goes, again, in my opinion, then there does need to be a conversation about pacing, meaning that if the iteration speed of intelligence is going faster, which means capability is going faster, that means what you can do is going faster. Everything is moving faster, right? And then the question, the underlying question of are rebuilding something that is harmful to people and that is not going to be maliciously used. And whatever that means, we all have definitions of what that looks like. I'm talking about the definition based on how they form it, however Anthropic defines that, however OpenAI defines that, whatever that means for them. To me, those are two different things. And so if we're speeding up that development cycle, again, I'm just going back to what Z didn't do. Right. Which is Zuck move fast, move fast, broke things, sped up the development cycle, sped up distribution, did all of that and then double back and said, well, wait a minute. We, we, we kind of fucked up a little bit. Brylan: That's fair. Dre: Right. Like we, we, we fucked up just a little, we just, just a little bit. Right. It's like, yeah, all right. Now after a billion people are fucking harm, now you want to double back and talk a little bit about alignment with society and whether or not this was right or wrong. Like Okay, like, all right, like, I guess we can have that conversation, but should we not have had that conversation a little bit earlier? So to me, safety alignment and building something that is good for society, to me has nothing to do with China. Those are two different arguments. Now, I'm not saying people are not trying to mix the two together, but it just in my brain, they're Rodney: So can I so so can I Dre: separate. Rodney: so so what I'm trying to do is so I one I wanna say I don't think anyone on the on on this podcast is against the conversation. So I think we're on alignment on the conversation is productive to be had. Amongst that conversation What gets brought to the table as far as the risk, the reward, the harm, the solutions, I think that's up for debate. So that's the area that I'd like to stay in. Not whether or not the conversation should be had or not be had. Then I think the second thing would be is that. I'm trying to understand in this particular scenario how we can separate China from this because a lot of the safety concerns that Dario cites are safety concerns that The development of the models in China. Will create these concerns, in fact, may even make things Put us in a potentially vulnerable vulnerable position. I think one is the bioweapon thing. Dario explicitly mentions we don't I'm gonna he's standing s tall on that. I'm standing strong on we don't want these models to come out and we we we we believe in science etc but bioweapons is something that we like care deeply about and that's that would be harm to society. We don't know See, with intelligence, we don't know the limits of intelligence. It could be infinite. So there may be ways to create new forms of bioweapons that can spread very fast with basic stuff that you have in your kitchen that wouldn't raise an alarm. So we don't know those type of things. And maybe higher intelligence would unlock those type of answers that we don't that that we didn't know. So that would be very dangerous, extremely dangerous. I think though that those seem correlated with the China situation because you can't separate them simply because China is not just publishing the models, so not just publishing the models and the open weights, which allows any American company to use them. Even if you banned, even if you were able to somehow technically ban using a the product surface, the protocol would still exist. So therefore, pacing makes no difference at all. It that that's at least in that's that's the chain link in my mind. The pacing doesn't make doesn't doesn't make a change at all. I'm not saying that. to offer in chat GBT or in Claude the product itself to offer these certain type of things. But they do seem linked in this case because of the because of the nature of the technology. It has a protocol nature to it. So once you put it out, you can't put it back in. Once once it's out there, it doesn't really matter if if if OpenAI and Anthropics stayed at their current model today, And China kept moving forward, they would have increased capabilities, whatever comes from new intelligence, and that information would spread because it's information. We we literally live in so so so we so there's no there's I'm not seeing we're in this weird scenario where we cannot, I don't see how we could, I don't see how pacing, don't get me wrong, I think the conversation is fine. I just don't see how pacing is the solution. I see it as, okay. Dang. This is getting out of control. But I would argue that it was already out of control maybe 12 months ago. And they just kept steaming steaming forward with momentum. In fact, I watched a recent Elo Enoch interview and he said, He spent multiple years trying to block AI, slow down AI. And every attempt that he did to do that made it go faster. And so he's just accepted to say, you know what, let me just lean on the optimistic side of it. And that's where I would like to say I I'm at. Where I'm saying Elon was scared of AI, likely. And I don't want to fully speak from but from what he sounds like is that he could see all the ways that this would go terribly wrong. And so there are multiple things that he did to try to slow that down. And he said every one of them failed. And he said the signal or the lesson that he got from that is you cannot stop this. You cannot stop this. And so when I hear Elon say that, and I'm not saying Elon has all the answers, but when I hear him say, Okay, we cannot stop this, but then I see these a whole nother group saying, Okay, we created all the momentum. Now we want to pay, now we want to pace it, now we want to slow it down. I'm saying, okay, how do I reconcile those two things? Because to me, the risk that they are mentioning, so maybe at least for me. It could just be a communic may maybe there is something that maybe there's a communication issue. I think I Dra I think Drake did mention that before. Maybe they're not communicating in such a way where it's clear how the pus the solution that they're offering. And this is how leadership has this is actually something that leaders have to do in general. When you come with something, it's not enough to just come with a solution and say, hey, this is what we're doing, this is why we're you you people need to understand. what you're doing and why you're doing it to get buy-in, to get buy-in. And if a lot of people are not getting buy-in, to me, that signals that their communication is not at a point to and it to convince the people who and there's some people you'll never convince, but the stated reasons are easy, at least for me, to disprove. And when I look when I look at people who are really looking into this, People who are very, very passionate about AI, who run their own local models, who use AI much more than I do. And I use AI every day at this point, in an extreme way. I'm I'm using AI for every aspect of my life: engineering, personal, finances, every single aspect of my life. When it does I I see I don't know, I I just I I don't I'm not convinced on the pacing. Even though it makes sense to me, the logic behind it though, based on their stated reasons, don't doesn't make sense to me. Because all the things that are listed, I'm like, pacing is not gonna it's not gonna solve those specific problems. Paces is not gonna solve those problems. So if pacing Dre: Mm. Rodney: is the solution, okay, cool. But it's not gonna solve the ones that you've that you've listed. Dre: Yeah. I guess we can agree to disagree, man. We've got different perspectives on it, which is cool. I mean, that's why we got the pot, right? To all have a different opinion and a different take on it. Again, the last thing I'll say on this, when I talk safety, I don't care what Dario said, okay? I'm separating this from what Dario said. I'm talking industry, right? To me, this is my opinion. To me, safety is separate from China. I understand that Dario is linking it. He has his own safety concerns dealing with China and chips. I get that. I understand that conversation. I totally get that conversation. But to me, that is not the only safety conversation in the room to be had. So if we all agree that the conversation about safety should be had, I'm talking about whatever that means to y'all, because I think we're probably all aligned on what that means. And that's why I'm tying it to Facebook and the impact they've had on society and not tying it to China. and what China is doing. Maybe that is a whole different safety conversation to be in. Maybe that is safety, right? The competitiveness between US and China. Maybe that is a safety conversation. They control the rare earth materials. Maybe that is safety. mean, there's been an argument about national security being there. Rodney: Yeah. Dre: So maybe that is a safety conversation to be had. That's not the safety conversation I'm talking about. I'm talking about safety in terms of society. And are these tools good for the people? That is separate from rather than not China's speed, what China doing with New Zealand doing where everybody else is doing. Those are separate things. I'm not even talking to having that conversation. So maybe, Rodney: I have an answer. They're they're bad. Dre: maybe those are different, but it's just like, I'm just simply saying the same stop that I think Zuck should have had, which is, and again, this is my pain as a product builder. think when you build something at scale, no matter what it is, if you built a meaningful company that is at scale. you should be asking is what we're doing good for society? I think that is a fundamental question that you should ask as a business owner when you're building something important and meaningful. So when I talk about safety, that's the conversation. That's the lens in which I'm talking about it. Not what Dario's saying, not what some hacker news or everybody else, because I haven't seen all that. And I'm not saying that those conversations are not valid or not baked in logic or whatever the case may be. All of those things can be true. I don't know. But I'm just simply stating to me what safety means and how I see it and what I'm thinking some of these things can mean. And so to me, the conversation about again, slowing down pacing, mean, maybe pacing the definition looks different. I brought back to the analogy of, okay, if you go to the store and you want to buy a firearm or you want to buy a bag of chips, you could argue that there's pacing and the purchase decision of buying a firearm. Why? Because you got to go through more checks and balances. You can still buy both. You can still use both, right? Like all three of us can go to a store today in Texas. and wall walk out with a firearm. But there's pacing in between those layers of getting that. So what does that look Rodney: right, right, right. Dre: like in AI? It could look like what you said, Ronnie, which is know your customer. That's one area where that could look like. It Rodney: Age restrictions, age restrictions. Dre: could be age restriction. It could be a lot of things. I don't know. It's endless. So that's all I'm saying. I think those are two separate things. But we've spent a lot of time on this. I don't want to. Rodney: Yeah, yeah, yeah. This this will be this is probably we we probably this is our main this is the main topic of the podcast. We we plan for other topics, so unfortunately this is probably gonna be our only one. But I will say this through the lens you're talking about, I would say this. As AI stands today, it is harmful to society, in my opinion. And so I don't so We don't even have to talk about whether or not it I think it is harmful in its current state. Take out the bioweapons. Take out the the super extreme stuff. Just look at the FCC. They can't they officially are gonna close they were requesting comment from anybody in America that they're and they're closing it in the next couple of days. Where essentially there's has been a this huge claim of ideology and the way that they tailor answers based on you know ideology and and other things that are that users are not aware of. And that reminds me of one of the biggest mistakes that Zuck made, but also Twitter made as well. which is censorship s censoring. And so I think that that I I believe censorship is dangerous and at the current moment. The even the fact that the FCC has gotten to the point where the they're asking the public for more information on that, because they already can see that every view is not tolerated. And I'm not saying extremely hateful views like sh should you you know, the like the the the the extreme hateful views. But there are clearly if you push the boundaries enough, or you say things that are outside of the status quo, These models often will try to reframe what you say, or they'll try to pull you away from a stance that you have. Or they'll try if you watch real closely, and when I saw the FCC thing, I was like, they they they see that. And that's a subtle form of censorship, which is which is, in my opinion, very dangerous. Because then what you have is you have a society of people who who only think How anthropic and open AI want you to think. And that that Dre: Mm-hmm. Rodney: in itself is extremely dangerous. And so forget dar like I agree, forget the biopic, the dario, forget all that. You guys are are controlling how people think. Go ahead and try to think of whatever three beliefs that you have that you would not say in public, but you believe it. But they they're against the status quo and go and try to talk to these models about it and see how it goes. And you will see that they'll either reject it reject it outright, they won't answer it, or they'll say, Hey, no, look at it like this, or don't think about don't think about it like this. And that is a subtle form of of censorship. And it's a it's a mind control and for young adults that will be dangerous because what I will see is is that a generation that is gonna lean more on this, they're gonna come up thinking a lot more things. in it it is dangerous. And and long story short, you know, but this is a long story short, and I guess this is gonna this is gonna be the the main topic we talked about today, w which is a good one. This is a good one actually. This was this was a good one, but Brylan: This is the conversation people need to be having. And I Rodney: Yeah. Brylan: yeah, I I really do appreciate Dre bringing like the societal perspective. Because once again, I think on this topic, if people weren't having disagreement, then we all should be like questioning, if I'm gonna be honest. Rodney: Right, right. Dre: 100 % 100 % Brylan: but nah, like this this was a dope topic for sure. Rodney: It's dangerous. It's dangerous. I I I think they shouldn't It's and it's so subtle, it's hard for me to even communicate it. Yeah. Dre: Yeah. No, I know what you mean. I know what you mean. And that's why I'm asking, what does this look like in 10 years, 20 years from now? Right. How? Rodney: That's dangerous. Dangerous. Dre: I don't even know. Right. Like, how how will these tools shape society in a way that social media is kind of shapes Rodney: N PCs. Dre: this shape society? Right. What does that do to young men? What does it do to young women? What does it do to generations? I don't know. Right. I just think these are fundamental questions that if you're going to have a billion people using your shit a week, I think it's fair to have that conversation. That's because that's a huge responsibility. That's a huge responsibility. And oftentimes in Silicon Valley, like that's not that's not the conversation. The conversation is build fast, break things, distribution, go, go, go, go, go, go fast as possible, fast, just go, go, go, go. You don't think about the shit that you knock off the shelf as you're running through the store to sort of say you don't think about that. We should think about that. Rodney: I agree. I think I agree. And so I will concede the one point, which is that I just think right now shit is dangerous. As as it currently is. No if they stopped everything. As it currently is, I believe the product is dangerous. Because I'm looking for those discrepancies. That's why I see it. But if you just go on as a user. And you just trust it? Dre: Yeah, I remember you talk about it a lot in some of the old pods about these models steering your thinking. And one of the things I think you've always stood on is Grok grounding that shit in truth and not trying Rodney: Yes, yes, yes. yeah, yeah, yeah. Dre: to steer your thinking. But a lot of these models will, like you said, literally tell you how to think. And for people that's not questioning that, and for most people, they're not questioning that. They're not Rodney: They're not, they're not. Dre: questioning that, right? So. I don't know, man. think this going to be Rodney: Is it Dre: a conversation that's going to be happening in this society. And this is the thing that's happening right now in our world. If you get on Twitter, people are having this conversation. That's why Jensen posted that thing about open models. This is the topic of the time right now, and rightfully so. Rodney: Well guys, look, I think it's perfect timing. I think I think we should have some good clips from this. by the way, Dre: 100%. Rodney: to anybody listening, we do have a new Patreon. So if you guys will go check that out, there's gonna be links to the Patreon inside of our descriptions on Spotify, YouTube, and any platform that you can find us on. And if you don't follow us on YouTube or or or or or Instagram, please do so. Find us on everything. If you do like what we're we're talking about, you want to learn more, etc., share it with your friends. And yeah, we'll be back soon next week. Dre: Alright y'all, this is a good one.