Latest / Beat The Odds with Dre, Rodney, & Brylan / Anthropic’s Safety Revolt, AI’s Bioweapon Risk & The Global Regulation Standoff

Transcript

Anthropic’s Safety Revolt, AI’s Bioweapon Risk & The Global Regulation Standoff · Beat The Odds with Dre, Rodney, & Brylan

• Dre: Yo, what's good, y'all? We back for another episode of Beat the Odds. It's just me and Rodney today. Bryan is out, man. How you doing, bro? Rodney: I'm doing great. I'm doing wonderful, man. How are you feeling? I know you just got back from a beautiful s scenario. You know, one of our Dre: Yes. Rodney: close friends got married. Dre: Yep, yep. Yeah, bro. I just got back. I literally just touched down last night, bro. It was a long travel day flying all the way back from Italy. It was like, bro, eleven hours to get back just to get to Atlanta. Then I had a two hour flight from Atlanta. So it's been a long, a long day. Rodney: Do you sleep or you be up? Dre: Bro, you know I'll Rodney: Yeah. Dre: be up. You know I'll be up, bro. Bro, cause I Rodney: You don't take yeah You don't take nothing? Dre: I bro, I did, but like I just it don't like I'd be it'd be emburst. Like I might pass out Rodney: It it didn't work. Okay. Dre: for like 20, 30 minutes, but I'm back up 20, Rodney: damn. Dre: 30 minutes. Like, you know, like I just be up, bro. I don't know. I f I just for those that don't know, I just got real bad at flight anxiety. It's gotten better over the years. But like I just I don't know, I can't really sleep that well. Plus the seats, I don't know. I just don't get the best sleep on planes. So usually not not too not too much. Yeah. That's really what it is. Rodney: The seats. The seats. The seats. I'm looking forward to a future where the seats just feel better. I I know Dre: yeah. Rodney: you can get you can get the comfort seats and you can get the first class seats, but but also it just depends. Because when you're flying international, I know that's not always that's not always an option, right? Dre: No, even I I tried. I tried to upgrade we we tried to upgrade our tickets on the way back to like first class or business. It was not available. Like it was all it was all it was all full. So they don't I don't even I don't know how many seats they have on there at that level, but not many apparently because we couldn't upgrade ours. So, you know, we were forced to, you know, sit sit sit in the in the back. But yeah, I wasn't really sleeping much. But it was all worth it, man. It was a great trip. Italy is beautiful. If if if if people have never been, I would highly recommend it. We went to Rome. Incredible history. Got a chance to see the Coliseum, which is so surreal. Like I've seen a lot of historical sites, but that was just one that I was just amazed to see. Honestly, I would go back and see it again because it was just like it's just so much hit rich history from that one place. and to just be there to see it, hear about it, do a tour of it, all of that was just incredible. So, Yeah, it was it was definitely a good trip, good food as well. So glad to be back and be like in the swing of it, man. I yeah. Is when once you get back from vacation or whatever, you just have so much shit you gotta catch up on that you'd be regretting going on vacation in the first place. You'd be like, Damn, I should have never did this shit. But, you know, it's all good. Rodney: Yeah. Right. How long were you gone? How long how long were you gone? Dre: probably ten or twelve days. Rodney: that I see. That's what it is. Dre: Yeah. Rodney: Cause w cause you could at that point you don't checked out of real real life for two weeks. So now it's so now the catch up is a little bit different. Dre: yeah, yeah, yeah. But you know what? But but you know what? People still be hitting me up about shit. That's the thing about an entrepreneur. And I I accepted that on this vacation. I was like, I'll I'll always be able to take a vacation because I work for myself. So I'll always be able Rodney: Right. Dre: to decide when I can take vacation and how I choose to do it. I can always decide that. But I'll never fully be in vacation mode. At least right now, right now, because this it's just the my business just requires me to kinda be involved. Rodney: You need to. Dre: And so yeah, that's just what it is. And so that's okay. I just accept that. My wife accepts that. And I've just told her, like, that's just what it is. Like, yo, we have a great lifestyle. These are the trade offs and the conditions for us to live this lifestyle. So we gotta be comfortable with I might have to take a call. I might have to respond to something on Slack. Like I just might have to do that. And that's okay. You know, but I'm not gonna be in full work mode. You know, I didn't have to pull my laptop out and be in full work mode, but I might have to step aside, I might have to do this, or you know, that that might just be what it is. So, you know, accept it. We're in Italy. What you're complaining about? Rodney: I I think I think I think the framing is correct. And I also think communicating it is correct. Because sometimes we don't communicate that. We just think it's automatically understood. So like somebody Dre: Yeah. Yeah, yeah, yeah. Rodney: will be feeling some type of way, and then we just like, Dre: Yeah. Rodney: well, why why do we gotta even say this? But but but communicating that Dre: Yeah, yeah, yeah. Yeah. Yeah. Rodney: is a is is a good step. Dre: Yep, yep. So yeah, man, it's been good, man. What about you, man? Rodney: So look, for me, I have been I've been slowing down a little bit. I've still my usual locked in, but maybe toned down a little bit. Maybe not four thirty. Okay. but Dre: Yeah. Rodney: five five thirty six is it's still is r where I'm at right now. yeah. Dre: Okay. A little sleep in. A little sleeping in a little bit. All right. Rodney: A a little bit, a little a little bit more, a little a little bit, a little bit. I slowed down a little bit. I I was on I I was in a mode where during the summer I had locked in crazy. Like there during the summer, I felt like the summer felt like two weeks. That's how locked in I was. It it felt like three months Dre: Hmm. Yep. Rodney: felt like two weeks for me. Like that's how focused I that had become. And I won't say I burnt out, but I think that I started to kind of get greedy and try to like try different things in the midst of being locked in. And some of those things, if you try to experiment, like I I come from the mortals culture. So we're where we believe in experimentation, trying things, different supplements, different sleep times, these different things to see how and sometimes experiments don't work. Right? So so Dre: Mm-hmm. Rodney: so you know, I had tried something And I think it it it took a bit away from my energy. And and another thing is an I I made also another crucial mistake. Don't try two things at once. So if you are let's say you are taking let's say you're taking a a supplement. It could be something as simple as a daily multivitamin. Don't take that and then also take fish all at the same time. But like because then now if you do get a benefit or a side effect, you're not going to know which one is the one that's causing and it's har it's hard to do because a lot of times what you'll see on the internet it's a prescription or a protocol To do a bunch of things. And it's probably best to go a bit slower. And I knew that. And like I said, there's a there's a greed that was involved in that. trying to really maximize it. I'm already at another mode. Everybody's like, what are you training for? Like, what do you, what do you, what do you, why are you like I I still Dre: Yeah. Rodney: go to the gym every day, but I was going to the gym. twice a day. I still go t most days I go twice. But it's like it's like what are you doing? Like and I had no I Dre: Yeah. Yeah. Rodney: had nothing. I had I wasn't preparing for nothing and and I was just trying to go to the next level. I'm like, shoot, like this is if I can maintain this, I I'll I'm gonna be I'm gonna be killing LeBron in a minute. So Dre: Yeah, yeah. Rodney: but nah that nah nah nah I went too far with that. So I had to reel that back in and cool it down a bit and realize too that That I need to take some time off. I was just telling Dre this. but speaking of actually so so there's like a a few things we're talking about, but I actually want to start it off with agents, and the reason why I want to start it off with agents is because Dre was talking about, you know, being an entrepreneur, you gotta take some calls, you gotta take some calls, maybe answer some emails on the way, etc. And before we hit record, we are talking about this new agent called Insight. And Insight is this basically, it's this new agent that came out. maybe I would say week or two ago and it is completely just inside of your the way you interact with it is directly in your messages app and it's a it's basically a personal assistant now we heard before 2026 that this this would be the year of agents but usually agents As we have come to know them, have been pretty straightforward from the standpoint of you already have to know what you want to do before like Dre: Mm-hmm. Rodney: before they can be useful to you. So if you don't have any ideas, then an agent is pretty much has been useless to people. And I think that's one of the reasons why you see people not having value from AI and agents. But what Insight was interesting, we had a friend. I think it was not a friend, Bryan. he he was he was sharing it and and honestly I had these low expectations. I I signed up for it with his code. I'm thinking, he probably gets like a benefit off of that or something like that. So let me sign up with his code. and I sign up, I connect a account or two. And I just don't I just forget about it because I just I just forgot about it. I didn't think it was no big deal. I'll another another software that I signed in with Google with, and then I'll forget about and never use. The next day it says me a it sends me a text like, hey, like I we seen this email. Like, do you want to like create a draft? Turns out for that particular email, actually I would like to see a draft. Yeah, like send me a draft real quick. It gets me right back. Draft looks perfect. Don't look like AI or anything. You want me to send it? Yeah. Go ahead, send that. Right? So then I did so then I started saying, okay, this is interesting. So I I leave it alone. I don't think about it. A couple hours go by. And I had this other task. So I had I had gotten my mom this this beamer truck, right? So and the the brakes were bad. And we had got some new brakes and rotors and all that stuff, and I had ordered the parts, but they had sent me double of the same parts. Right. So it should be, you know, rear and front brakes. Well, they sent me two pairs of front brakes and no rear brakes, essentially, right? So I had been procrastinating on fixing this issue, going back and forth with them on email. And insight was like, hey, you still have this issue going on. Like, do you wanna like do you wanna handle this? Like we can I like, you know what? I just took a picture. And then and then it took the pic it took the picture and then it drafted the the reply. It's like I need two boxes or three boxes, however many it was. Sent the email off. Boom, done. And I started to say, wait, hold on. This Dre: Mm. Rodney: is starting to make me feel this started to make like make make Dre: Hmm. Rodney: it starting to lift the load off of me. It's starting it's starting to li lift the the load Dre: Yeah. Rodney: off. And So this is the first in my in my view, this is the first Dre: Yeah. Rodney: agent that has been useful in a proactive way for me. I think Codex Dre: Mm-hmm. Rodney: is great. I was I'm a fan of Codex. I think Grockbot seems decent. I just haven't really got to really embrace it that much. It's a lot to set up. And then but instinct seems kind of like almost immediately useful. And so and it also doesn't seem Dre: Yeah. Hmm. Rodney: AI like usually when AIs create these emails or text messages or whatever, they're like obviously AI. Like you you could look at it and you're like, I don't want to send that because the receiver is gonna they're not gonna like like it's gonna they're gonna feel some type of way about that. I know I do. Dre: Mm-hmm. Rodney: When I see an AI email, I'm like, okay, like here go another AI generated email. So Dre: All right. Rodney: so yeah, that I think that's the interesting thing. Dre: Hm. I had to give it out a try. And and that I mean the product is called Instant, right? So I think whoever created this product, I don't even do you know who's behind it? instinct. Okay. Instinct. Rodney: Instinct. Instinct. And and they're I I believe they're yeah, they're from San Francisco. Their site is so simple. It's crazy simple. Like if you just go Dre: Mm. Rodney: to instinct.com, it's just a bunch of texts. Actually, a small amount, it's like a paragraph of text. They're some they're from they're Dre: Yeah, yeah, yeah. Rodney: they're from San Francisco. They're not trying like it's it's it's lightweight. I don't know who's behind this. I really don't know who's behind instinct, to be honest with you. But I can tell you that they are they're they Dre: It's a good product. Rodney: they're clearly focused on providing value as soon as possible and proactively. And and and it goes to show like it goes to show Agents can be useful, but maybe like help us help us see the value. Dre: Mm-hmm. Yeah. No, I I I love this. And I'm looking at their website. This is clearly made by engineer. Like this is Rodney: yeah, yeah, for sure. Ha, for sure. For sure. Dre: this is a straight engineer, no bullshit, like just straight to the product. Like all Rodney: Yeah. Yeah. Dre: the other sh like they just put all the effort into the product, which is what matters Rodney: Yeah. Dre: the most. Bro, this amazing. And this man, see, stuff like this is what we were just talking about, like being inspired, right? And having that that Rodney: Yes. Dre: energy. Seeing stuff like this is what makes me miss like being in the Bay Area. partly being in LA too, but really being in the in the Bay area 'cause Rodney: The Bang. Dre: I haven't been back in years, but I can't imagine what the energy is like out there right now. Like just so much exciting stuff happening, bro. So many founders working on crazy stuff. I mean, even as simple as i as as an assistant just able to draft really nice emails. You know what I'm saying? Like, Rodney: Yes. Dre: So that's dope, man. I'm I'm a check out the product. You see my involving. I actually signed up while I was on the phone. So yeah. I'ma I'm I'm I'm Rodney: No, yeah, yeah. And I think Dre: gonna give it a spin and see, you know, see what I can get out of it. Rodney: And I think it can do more than what I'm saying. I just that's just all like what like and and and it's telling me s and it's telling me stuff I don't want to hear. For example, it like I I had the IRS email me. I'm like, I don't want to hear Dre: Yeah. Rodney: about that. It's like, this looks serious. It's like I see that. I Dre: Yeah, yeah. Rodney: see I see I trust me, I saw I saw that. Leave that alone. I I said leave that alone. Yeah, leave Dre: Yeah yeah. Don't wor don't worry about that. Rodney: that leave that alone. I'll get to that. Dre: Yeah, yeah. Rodney: I'll get to that. I'ma talk to them soon. Dre: Yeah. Rodney: but the the other thing was Muse from Facebook. So people are suggesting that that so Muse is also an agent as well, and it's at Muse.ai and connects directly to your Facebook account. and I found news to be useful. Excuse me. I found news to be useful. And there's a few things you can do. There's like a feed, there's ideas, there's goals, you can create stuff. but then it I would I would say that it didn't it doesn't come off as proactive as instinct. That's what I liked about instinct the most is how proactive it is. Now some people are saying they will choose Muse over instinct. like for them I use Muse to clean up my email. It was like, yeah, we see your email has a bunch of you know, basically nonsense promotion. Like, do you want to clean that up? Yeah, yeah, let me clean that up. And and so it cleaned that up and unsubscribed from a bunch of things. That was useful to me. I was not gonna do that. That's a task that I was just not gonna do. low on my list. But then but it lean me it led me to start thinking that and I never really got a good answer from from from from Rylan or anybody yet on this, is that Why are people looking at it like I need to have one agent? Dre: Mm-hmm. Rodney: In my view, now mind you, one day if they start charging us for these agents, which they probably will at some point, then I can start to see people being a little bit more specific on the agents. But but I don't see why people want just one agent. Because I think that what's probably gonna happen is the future will likely have multiple agents. And the reason I think that is because every agent won't be good at everything. And some agents will be better at some things than other things. So depending on what you're doing. And so essentially I essentially kind of believe in Mark Zuckerberg's vision where we're gonna have these teams. Like everybody's gonna essentially turn into these CEOs essentially. And I know it's hard for to see, but I'm already seeing the first steps of that. Dre: Yeah. Yeah. Yeah, yeah. I I agree with you. I I do think people use different agents for different reasons because they're better at certain things. I I totally agree with that. 'Cause I I do I even do that now. I think for me, and maybe there is a way to do it, I'm sure there is, probably through some central platform or something. But I wish there was an easier and better way built into each other tools like Codex or or Claude or even Grox. where you can easily share context between all the all the tools. Cause that Rodney: Yeah, for sure. Memory and stuff. Dre: yeah, memory. Because I think that's the most annoying part is like starting something in one place, but then you you you can't even like really interact with it somewhere else. I don't know, like having to kind of silo it. And Rodney: Right. Dre: and I think and I think for me what happens is like it just becomes ex it it becomes mentally exhausting to try to figure out where everything is. Even in Codex. Like I use Codex a lot. But having to switch between chat and work and this and that, like sometimes I'd be forgetting like where I even where the conversation is. I'm like, damn, where where I'd be like, Rodney: I yeah, I don't like that. I don't like that. Yeah. Dre: I know I talked to Codex about this thing, but I'm like, where is it? I can't even find it. I'm like, that that is so annoying to me, bro. Or like I start Rodney: Right. Yeah. Yeah. Yeah. Yeah. Dre: something on my phone and then it don't show up on a desktop app, and like, where is this shit, bro? Rodney: D did you see they added they you could tell they ran into that issue because they've recently in the last week or two added a feature where you can say where's the chat where we talked about this and then it'll go and it'll pull it up. It's like so that that's like a brand new thing. That's a brand new thing. Brand new. no, no, no, no, no, no. No, they solved it. They solved it. They they solved it. You Dre: Okay. Okay, yeah, I miss I missed that 'cause I've just been using search or like fuck it start a new thread. I'm like fuck now I gotta start a I'm like it's annoying bro. I'm like, yeah, solve that, please. Please solve that shit. Okay. Rodney: you all you gotta do is say If you just know the gist of what you are talking about, it will it will yeah. Dre: Yeah, yeah. I will say I will say the memory on Kodak seems it is impressive. It is very impressive. Like 'cause I I have noticed where like I'll be in a chat and it'll reference it'll reference something I said a long time ago about something else. It'd be like, yeah, yeah. Because like it'll it'll draw that that that backwards connection. Rodney: Yeah, I peeped that. I peeped that connection. Dre: And I'm like, bro, I love that shit, bro. I I love that. Rodney: Yeah, I think that. Dre: Like, I love that. And I I mean I I don't I Again, I I spend a lot of time in Codex. I do use Claude a lot, but I don't I don't know if I don't feel that same way about Claude that I do about that about Codex and the memory side. I'm sure Claw has the same capabilities and all, but there's something about the way Codex like recalls something that you've had a conversation about that is like okay, you you you under you you got it. You you you you you you fully are are listening. It makes you feel like you're listening. That's what it makes it feel like. Like you're listening, you're listening. You hear Rodney: Yeah. Yeah. that's a good feeling. Dre: me. Yeah. That's that's dangerous though. I'm like shit, cause damn, Rodney: Yeah. Dre: I don't know. That could go that can go in a lot of different ways, but yeah, yeah, yeah. It's interesting. Rodney: Right. Yeah. It's funny you brought that up because I think I had that fur a moment like that for the first time, probably about two or three weeks ago. I was talking about something like personal and then it like pulled back and connected it to something that I had talked to it about like six months ago or like it was like but it made sense though. I was like, dang, I had forgotten we had talked about this. Dre: Yeah. Yeah. Yeah. Rodney: And but you remembered that though. So this is kinda Dre: Exactly. Yep. Yeah. Rodney: it's kinda it it it's kinda it it Dre: Yeah. Rodney: it's it it's pretty interesting, to say the least. it's it's interesting to say Dre: Yeah. So Rodney: the least. Dre: memory memory and like you said, that ability to be proactive. I think are two things that like yeah, agents I I that's the next sta stage I feel like to really make these things super, super useful. so yeah, bro, I I I agree with Zuck on Zuck's point on like the future is is definitely multi agents. People will be using different platforms, different tools. and even right now, I don't even see why you would just want to use one agent. Like There's so much explosive innovation happening. I feel like people should be using all of them to kinda really understand which one is good at what. Rodney: Right, right. Dre: I don't even I don't even want to get tied down to one. So I'm trying to y I'm I mean, I wanna use them all. I don't you know. Rodney: So so so that's my thing. That's my thing. And honestly, if and whenever they start charging, for the most part, I'm gonna be true. Like, yeah, of course I I I'll I'll I won't use a hundred if they start charging, but I still will be having multiple 'cause yeah, I I don't wanna I think it's too early to be like, let's switch to this one. this one came up let me this one came out, let's switch to this one. It's like I'm not let's not Why why not both? Well, why not both? Muse Dre: Yeah. Yep. Rodney: is cool. Instinct is cool. They they they operate different. It's cool. We we're in a good place. And so I think this is what I think it took a while for them to get to agents that I can see being more useful just for people who don't know what the agent is supposed to do. So they might you might tell somebody of your your family or friends, hey, agents and AI, and they're hearing you, but they're like, okay. Okay. Instinct, I've instinct, I have the feeling that they don't need to get a lot of instinct. I have the feeling that they don't need to even actually know anything. It just will find stuff. And to me that Dre: Yeah. Rodney: that they're gonna I I think that's almost like the the user experience on that. They nailed that. They nailed that. I don't know, I don't they nailed that. They nailed that. yeah, Dre: That's useful. That's useful. Rodney: it's it's immediately useful. It's it's like for example, even for example in order for me to get the referral link, I just texted I I texted, I said, give me the referral link to send to my friend. It's like, yeah, here's the link. And so I just send it to Dre. Like it's just like it's it's it it's it's just helpful to me. And I'm texting it through iMessage. So it's Dre: Yeah, yeah. That that that's a good modality too. I think iMessage because I mean, bro, we how many times do you be in iMessage a day? Like I'm always in iMass like that's probably the what the number one used app for most people. So Rodney: Yeah. Dre: that's the to me that's the that's the good interaction. I think Bryan brought it Like, how come nobody else has really done that? I think some other Rodney: Yes, said Dre: people tried to do it. I think it was what was the co man, the LinkedIn guy. He has some company. Rodney: personal intelligence? Dre: What was that? Yeah, it was like some personal intelligence. I think they tried to do it through text, but I think Rodney: He sold it though. Dre: yeah, he sold it. And I think eventually they ended up building a app the like an app experience. But I think he started off with text. But yeah, I don't know anybody that's Rodney: Texas dry broke. Dre: really pulled it off real good through text, but like to me that is the best that is the best interaction mode right there. Cause everybody got it already. Rodney: It it's real nice. It it th they it got it and it comes off responsive. You're not waiting. Dre: Yes. Yeah, yeah, yeah. But speaking of AI, bro, I think we gotta we I mean the elephant in the room is this this big text, Rodney: Yeah, yeah. Dre: this big tweet, and I I just checked it. A hundred and I think what a hundred and sixty million views already. and this is only from two days ago. So we gotta talk about it. If you haven't heard about it, this AI researcher, Jacob Coxton, he worked at AI on AI pre training at Open AI. He also briefly worked at Anthropic and on September eighth, so literally just two days ago, he resigned and accused both companies of racing towards self-improving superintelligence and essentially gambling with our lives. And he said much more than that, but you know, generally the gist of what he said is that, you know, neither company is acting responsibly. Again, they're racing towards this self improving superintelligence. and we shouldn't underestimate the power of this technology. we shouldn't underestimate the power of of of superhuman systems and their ability to hack anything, their ability to manipulate individuals. and ultimately, you know, he decided to to resign and you know he created a whole tweet thread about it. And it's essentially taken the the the internet by storm. And I would say even mainstream media, like he was doing interviews I think on CNBC, a t t Rodney: A ton, a ton of interviews, Dre: ton of interviews. So Rodney: dozens. Dre: It's not even just on the internet anymore. It has gone from just a tweet to like actual news politicians, senators are responding and and and like, yeah, we need to create these bands and fucking Bernie Sanders introducing a ban to to or law to ban super intelligence. Like so, you know, I I think there's a couple of things that we can dissect here. Obviously we can dissect the the tweet itself, then we can dissect the re the reaction. And and then even dissect the potential consequences from from the reaction, right? Cause I think again, like the he he's gotten the attention of a of a lot of a lot of people who have the ability to to put roadblocks in the way or at least create friction for some of these frontier labs. What were your what were your thoughts Rodney: Yeah. Dre: on it? Rodney: So I feel like you might be able to to predict my thoughts on on on what he th I think you might be able to predict some of my thoughts a little bit on what I how I feel Dre: Yeah. Rodney: about what he about what he said. so I think that One, I didn't expect the tweet to go that crazy. Yeah, I woke up and I saw it. It was big. But see the thing is, in our industry, just because it's big in our industry doesn't mean it's gonna get into mainstream. Right? So I saw that. Didn't think, but I checked back, you know, first I see he has fifty he has 50,000 followers. I checked back. He has a hundred thousand. I check back. He has two hundred thousand. I check back. He has two hundred and fifty thousand. so I'm saying, wait, hold on. This this is this is not this is turned into this is turned into it it almost made me want I I don't even believe that. Like, should I tweet like this just so I can get a platform real quick? Can I can I get can I Dre: Yeah. Rodney: get should I pop out with a crazy tweet and then Dre: Yeah. Rodney: get this huge 300,000 follower platform and then you know turn around. monetize it i don't think that's his goal but i guess here's the thing right one it's hard to take it seriously because there are there's becoming questions raised by elon david sass etc whether or not he this is an a truly original non-planned thing they they're they're skeptical. And their skepticism probably just comes from the fact that there are a lot of planned things that happen in this world. So they're just naturally skeptical. even Bryan himself was like, I is it is it is it entropic themselves that wanna drive this narrative? 'Cause he can see ways where it benefits them in some ways. Especially because they want regulation, they want things to slow down. They've they they've said this many times before. So, but if we were to take it as real, I still think that this is a person who one, I think he comes off naive to me. I think this is kind of something that I've been standing by from the beginning, which is that open AI is no different than anthropic. And I think that anthropic is worse because of the fact that Anthropic pretends to be different than OpenAI. And that's the that's pretty much the main thing I've I've been standing on. The reason why I'm I'm so against Dario is not really about what he's doing, it's about the fact that he's pretending that he's doing something different. And he's not. He's doing the exact same thing. And I think if this I think if this employee thought going from open AI to anthropic was gonna be different. I think that was a naive thought. And he's is not the only one, likely, who has thought the same thing, who got taken from open AI with the intention to build safety and all these different things. What I thought was interesting was how many people from Anthropic came out and agreed with them. They came out and they work. They currently work there. They're not fired or anything like that. So it's very interesting how you can come out and say these things. How you can basically say like our company is because I guess what they're getting at is that We're in a predict a predicament where we wanted to slow down, but we can't slow it down without killing our company. And and I think that I understand that, but it's hard to have it's hard to take seriously. It's hard to take seriously from outsiders, I would say, because one, you don't have to contribute. You you don't have to contribute. These are all very talented people. Most of these people are very, very wealthy. even when you look at especially this guy who came from OpenAI working there for multiple years, he's wealthy. And I mean, you could just tell by the way he tweeted. Like, if he didn't have cash, he wouldn't have tweeted that. I guarantee you. He would not have said nothing. He would just he'd be quiet. He would think and he he would think that. But The financial incentive will keep him quiet because he's not so Dre: Right. Rodney: he's wealthy. and I think that you're gonna be hard pressed to get people to kind of sympathize with you because it just comes off weird. It comes off like you're saying, okay, we can kill everybody, but we can't stop because of money. They they like that that's that's that's the message. The message is, yeah, this is gonna kill everybody, and we're worried about it, but like Just understand we can't stop because it's because of cap because because of capitalism. And like we need somebody to step in to stop us, because we can't stop. And it's and it to me, it's kind of to me giving away the power that that they have. I think leadership could stop a lot of stuff. They just don't want to because of they just don't want to. Because yeah, there would be a lot of damage to the company and they probably can't see a way. And is that an excuse? I I mean, is that an excuse w when you're talking about the death of humanity? I mean I mean when you talk about death of one person, a murderer, that's horrible. But y and you talk about you know, a serial killer, that's horrible. Well you're talking about the death of humanity and you can't stop because of money I I just comes off I think people don't really like that. And I think that anthropic employees and anthropic itself This is just my my opinion. I don't think it's a good idea to kind of like play both sides. And then I then there was another employee that went viral when he said, I'll gladly give up all my equity if we can even lower the risk by 1%. And it's like, this you're just talking. You're just talking. Like, you're not gonna give up your equity. You're not gonna give up your equity. Because no matter what solution somebody throws out there, you're gonna say, that's not gonna that's not gonna work. So therefore I gotta keep my equity. So therefore I gotta keep working on it. So it's a it's a vague fear and you can't when people are fearful, you really can't create a solution to it because the goalposts will always move. Dre: Hmm. Hmm. Yeah. I think you y you said a lot of things that are interesting. And I think similar to to Sacks and Bryan and other folks, I was skeptical at first too, right? You know, is this some coordinated thing? But then I was like, you know what? Once other people from Anthropic kinda came out and kinda agree with him to your point, then it made me kinda look at it a little bit more at face value. And I understand where he's coming from in the sense that what which I get it. It's it's very difficult to understand this if you've never really been a part of a machine that is bigger than one individual piece or really the sum of its pieces to to some degree, right? When you work at a company like Anthropic or OpenAI that clearly is now cemented as important to society. important to his customers, important to investors. These companies have raised hundreds of billions of dollars in capital. Rodney: Yes. Dre: Like enormous amounts of money. Money that we've never seen companies raise in history. I I would agree with him. You cannot just stop that. You you can't. E e even if you a as an executive attempt to stop it, you'll just be removed. There are tons of cases where Executives have tried to go against what clump company governance and board and and the board of directors and investors want to do and and and then they get pushed out. I mean, it happened to the best entrepreneurs. It happened to Callus Trav to to to Cavit Travis Kalinick over at Uber. I mean it this happens. It happened to Steve Jobs. Like, so it's not implausible. So I I I I ag I I I can understand the conditions and the position that you're in where you're forced to see the reality of what you're building and you can't slow it down. It's the same thing that happened at Meta, right? It's the only reason why Zuck can be reflective now and be like, you know what, we didn't make the right choices. But why didn't you stop it? You you could have stopped it, right? But he didn't. Rodney: Yeah. Well he was he was he was he was in the he was in the Nile. And and the see so what's interesting, yeah. Dre: But see, that's what he's but see, well, hold on, but hold on. I but I want to make the the distinction between what's publicly said and privately. I don't believe he was privately in denial. I don't believe that. He he Rodney: Okay. Okay. Okay. Dre: said that publicly. I agree with you. That was the public sentiment, which is basically what what Jacob is saying as well, which is what we hear publicly and behind scenes is different. So I'm only talking everything I'm talking about is in reference to behind scenes, not public. Behind scenes, Rodney: Gotcha. Dre: I don't believe Zuck knew was was in denial. Zuck is highly intelligent. Has the smartest people around him, has all the data, all the information there. He I believe he understood the situation. Now, what do you say publicly was different? So I'm just saying I understand that there are a set of conditions, and this is American business. This this is not anything new or novel. This is American business. When you build a machine that is no longer just dependent on the entrepreneur that built it, when you take capital, when you scale something. When you start to get massive customers, you start to create something that goes faster, that can go on without you, right? You can take Dario Rodney: Yes. Dre: away from anthropic and it's still gonna run. It's still gonna go. It's still gonna Rodney: I agree with that. I agree with that. Dre: grow. So I I so I do think that it it it is difficult to just put the brakes on and be like, okay, let's just slow down and stop. I do think it's hard to do that. Is it impossible? No, it's not impossible, but I do think it's difficult. And if you try to do that, then you risk. been kicked out, you risk being alienated. Sometimes the easiest thing to do is to comply and then try to change within the best that you can. That's really all you can do. You know, when you when you Go ahead. No, no, yeah, yeah. Yeah. Rodney: Thing I think that's what I think yeah. Sorry. Sorry about that. Dre: I I I think sometimes that's just the best kind of the best that you can do. So I I get it from that that position. I I do understand that. Rodney: Yeah. Th there's no argument that they're in a tough spot. That like we we can't we can't I can't oversimplify that. The only thing that I want to put on the table is that there are there are two choices. If the company and the employees feel the way they do, they have two choices. Continue to work on it or walk away. If enough employees walk away, Anthropic would die. Now that still would have open AI. So they might say that that's not a good solution. That's not a good solution. And it may seem unrealistic, but nonetheless, it is an actual choice. Cause at the end of the day, a a corporation to say if Dario walked away, nothing's gonna happen. Nothing's gonna happen to Dario. If Sam and Dario walk away, nothing's gonna happen. But if employees that they all are saying that they care because open AI employees, they come out and say stuff, but they're not based on the foundation of The same principle that anthropic is like people go to anthropic because they're talking about safety. And so, in a way, I think they're playing a game that might get interesting in the eyes of the American public. Because they're coming, they're not, everybody's not receiving their message as genuine as maybe they want, they think is, you know, because even though they're in a tough spot and we understand that. But at the same time, if y'all continue to work on it, and y'all are supposed to be the ones who stand for right even when it's hard, and you continue to work on it, then you're you're you're kind of you're making a choice at the end of the day. And because if eight if if Dre: Well what well two things can be true, right? Rodney: if 60 if if if 50% of the employees stopped working for anthropic, that would significantly slow down AI development. It just simply would. Now, now would they have to now would there be a lot of stuff that comes with that? Yes, there will be a lot that comes with that. There'll be a ton that comes with that. It's not like a choice that, you know, but if they're weighing that choice, I'm just talking about the how they come from the culture that they look at it as I'll get rid of all my stock if I even have 1% chance. They're saying stuff like this. This is what they're saying. They're you know, they're they're saying stuff like this. If they're saying stuff like that, people are naturally gonna say, Okay, do it. Okay, do it. There there there are Dre: Yeah, but I don't believe every but I don't believe every I I I have seen a few, but I every every anthropic and open AI employee is not saying that. They're not saying that. They're they're not saying that they're willing Rodney: I'm I'm I'm focused on my topic. Dre: to well, okay, even an anthropic. They're not saying that. They're all not coming out saying that they're willing to all give up their stock and say that. Like, so right. Well, not even well, not even Rodney: Yeah. Yeah yeah yeah. Of course they're not. Of course they're not. Cause every cause everybody realistically is not. Everybody realistically, we all know that. Dre: realistically. I just know I just haven't seen that narrative. I haven't seen that narrative be overarching around anthropic employees saying that yeah hey i i would be willing to sacrifice this for that yeah a couple of them may come out and say that but like i don't think that's reasonable i what when has that ever happen when has that ever happened so all i'm saying is all i'm saying is for people Rodney: Right. S yeah. Yeah, yeah, yeah, sure. Dre: who are listening is that yeah you can look at anthropic or you can look at open ai as different what i'm saying is that they're no different than any other corporate company not it's Rodney: Yes, facts, facts. Dre: not even just ai i'm i'm any corporate company that is Rodney: Facts. Dre: maybe Doing something that is harmful to society and they still have people that work there, come to work every day and still contribute to that mission. They're no different. So I I don't I'm not gonna hold them into some standard and say, well, hey, they should do this or that. No, but I what I do think is people should understand like, hey, this is just this is American business. So if if we don't like this way of doing business, then maybe we should change how we do business in America. Cause what I take from his tweet is that Anthropic does believe in safety. They are taking safety seriously. They do believe it's important. Like, so I think there's a difference between when you say these things, but then behind closed doors, you don't really believe that. I didn't get that from him. I actually got from him that no people at at Anthropic genuinely believe in safety. They genuinely believe that these tools are harmful. Rodney: I yeah. Yeah. Dre: They they they genuinely believe that, which I think is important. They're not just saying it. No, they actually believe it. However, There are real market conditions that have to be acknowledged that prevent them from maybe taking the necessary actions that they want. Because the reality is to even be competitive in the frontier race, you gotta have capital. You gotta push the boundaries. Nobody, Rodney: So I yeah. Yeah. Dre: nobody without the capital is actually making a serious dent right now. That's why it's not that many of them. So like Rodney: Yeah. Yeah. Dre: it's like it's like multiple things can Rodney: Yeah. Dre: be true at the same time, right? Like I think what he's saying is is is fair and reasonable. And I think to to some degree, he is leaning into your point, which is if you truly believe that, then leave. And he is doing that. He left he left both, right? So he he is leaning into Rodney: He is, he is, he is, he is, he is, Dre: that. Rodney: he is, he is, he is, he is. I can't say nothing about him. I can't say nothing about him. Dre: Yeah, so I so I think he's com so I think he's commendable, but right now he's getting killed. He's getting questioned like, is this a is this a Rodney: Well Dre: is this a market employee? No, I mean it it it doesn't seem like it, because why why leave all that bread on the table? Rodney: Well he Yeah, it's a lot of bread on the table. So so he he is a person who but see what he doesn't understand is everybody's not gonna go fall behind him. To your point about the market and the complexities and everything else, he understood his that's that's my main thing is that the choice is stay or walk away. The biggest power you have is to walk away. If you don't walk away, it's going to call into question what you really, really believe. Walking away is the biggest statement you can make. That's the biggest statement you can make no matter what situation it is. A marriage, a s a relationship, it doesn't matter what it is. You can say all you want. But if you walk away, that's when it's like everybody knows, okay, we know where this person stands. And so he him himself, he is commendable. so so for me, so yeah, so pe yeah so I I didn't originally take the skepti the skepticism. I know you know people are throwing that. I didn't really like that per se, personally. I personally didn't I didn't take that either. But I I I'm zooming out to the bigger picture with dropping itself that they need to be careful co-signing on that while they still work there, if while they're also pushing safety and everybody's not walking away. Because the American public, the average American, is not thinking like they're confused. They they may not understand all of the market stuff that we're talking about, all the business stuff that's keeping them there. They're saying, okay, so we're all gonna die. Why not just stop? Like that so when I'm watching these the like the they're saying that. So don't be the company that is marketing as the safe company. We care about safety more than everybody else. We're the good guys. And then but yeah, we think it's we think that. It's gonna kill everybody, 10% chance, or however it was phrased. But then like everybody's gonna think, so then why not stop? Like they're they're it's it's it's gonna start to call into question that that could that could and we'll see. Maybe maybe it doesn't, but from what I could tell, even just yesterday, you know, I'm hanging out with somebody and they don't like AI. They don't got no re the they're young, they and you just hear all the stuff. I'm like, okay, this is real. This is people, smart people. Gra working on graduate degrees, all types of stuff. Younger individuals don't have the best thing. And that's before this person hasn't even heard about what they just said yesterday. So like imagine you saying Dre: Yeah. Yeah. Rodney: this on the news and you already don't like AI. It's like, wait, hold on. So they saying it's gonna kill us. And then even my mom, I wake I I wake up today, my mom sending it to me, Fox News or something like that. She's sending it to me. Dre: Yeah. Rodney: I it's like it has it's getting to my mom, right? So it's it's so so I'm like, wait, hold on. Y'all want it you it's not about our nuance opinion. Cause I see their the I see where where we at with it. I'm saying, hey, look, open AI doesn't really have that trust from the public. I don't think they have that. They're th I don't think they're viewed as the the company that is the better option. Like I don't think they're viewed as that. I think they're already not trust not trusted. I think salmon's viewed as not trustworthy. I think that in general, OpenAI is viewed as just not the most positive. I think anthropic has a has a better, but my prediction is that that anthropic is gonna do more damage because they initially position themselves as different to the public. And had they not done that, then it would be the same, I think. That's just all I think. That's just my prediction. I could be wrong on that, but I just think that that if Dre: Yeah. Yeah. Rodney: you start off saying, Hey, look, we are we are the safety company. But yeah, we also might kill you, but we can't stop. It's just gonna be hard for people to understand that messaging Dre: Yeah. Rodney: w without all the context. Dre: Yeah, that's probably true. But I also believe that in addition to that, that I don't believe that there's any I I don't care what Dario says. I don't believe Dario thought that anthropic would be where it is today so quickly. So how it's so I I can't so so so I can't Rodney: No, no, no, no, no, no, no, absolutely not. He wingin' it, he winging it, he winging it. Dre: so therefore I can't fault him on point to how he started or how he 'cause I don't even 'cause I mean, what he was gone up against seem I would assume probably didn't seem plausible. I mean, you gone up against a well funded open AI that's gone up against an even more well funded Google. There's gone up, you like so like the the the thought that you because there's tons of employees that left open AI to go create stuff. We haven't heard about them. We don't their products are not as big as anthropic. They're not even as you like there's tons like of people that left and have gone off and created stuff and it didn't go nowhere. So how he started and what he had to do to get here, I I I can't even fought him for that because Maybe if he didn't take that position, anthropy may not even be where it is today. And it may not even have Rodney: Yeah. Dre: given it may not have even given the employees the option to create the wealth that they have been able to create. Because it's not that many companies that have even Rodney: I'm happy for the employees. Dre: been able to even even even be at that size. But you know, I so I I can't flaut him for how he started, but I I think there's actually three options. You said there's only two. I think there's three, which is I do think that there's a legitimate call and I and I understand, you know. The all in podcast, skepticism around regulatory capture. I understand that we've we've walked through that scenario as well. I I agree with that. That is a possible direction. But I do think there is a real option to call on the US government to decide how this technology should be regulated. I do think that's real. I and I and I I think that is that is a fair thing to ask. Just like, you know, it's not up to General Motors or for a motor company to decide what the speed limit is. It is not up to Mossburg or Smith and Weston to decide what the gun laws are. It's not up to them to decide that. They have no dick no determination on what gun laws here are in Texas. Now they can lobby for certain laws to be what they are, sure. But that's ultimately up to lawmakers, right? And so I I I do think that there is a real separation of state to sort of say where Yeah, there's there's probably need for the government to come in. And I'm not saying what should be regulated or how it should be regulated, but they do need to work with these frontier labs. And I do think there probably needs to be more conversation around the government restricting and and regulating certain aspects of the technology. So that is not necessarily on anthropic or open AI to to have these concerns. Let them go off and build the product. Let them go build a company, build it serve their customers. Do what they do best, which is create value for their shareholders, and let the government does what it does best, which is create policy for for in the conditions for corporations to work within. It happens in every other industry. It happens in airlines, happens in railroads, happens in energy, happens in every other industry. It should happen in AI as well. So I I do think that is a third option that I think these frontier labs should continue to lean into. And I think the lawmakers should take it serious. They should take it serious and really try to come up with a regulatory framework that works, like they've done with energy, like they've done with the airlines, and like they've done with many other critical industries that have huge consequences when they not regulate it. Right? Like I I don't see it any different. Rodney: Yeah, yeah, for sure. So for sure, for sure. Did you see what what Trump said? You see what Trump said? Dre: no. What did you say? Rodney: yeah, AI is yeah, it it's it's a it's a Dre: You not worry. Rodney: it's a wonderful thing. He's he's like he's like you know, we'll we'll have the we'll just have a little gear and we'll i if we don't like that rope, if we just turn it up, we switch it off. Dre: Turn it off. Rodney: You know it's no b no big deal. He he's not worried at all. So it it's it's it's I I like his Dre: Yeah. Rodney: attitude because he's kind of he's an older guy. He's he's he's you know he's he's almost eighty. He's seeing a lot of stuff. He's seeing Dre: yeah. Rodney: a lot of stuff. So for us, it's I don't know, like it's hard to say. Trump, I'm not saying that Trump is the end-all be-all, but what I am saying is he has seen a lot of doom and gloom over his 80 years of life, and he might just be looking at this like, Okay, guys, this ain't Dre: Yeah. A hundred percent. It's no different. Rodney: Yeah, this this guys, like, it's all gonna be Dre: Yeah. Rodney: good. Like, we're not gonna wipe out the c we're not gonna wipe out humanity with this, like Dre: Yeah. Rodney: he might just be laughing at this, but but also at the same time, it's it's not to say that that he's hundred percent correct. He could be wrong, he could be underestimating it. I think one of the issues, before we get to the next topic, I think one of issues that's probably happening too is is that is that is that There is national security concerns with it. So people don't people are don't have a lot of faith that we can create AI in such a way where like essentially because the the can of worms has been opened, they don't have faith that regulatory will do anything because in this particular case, this is intelligence. This is like digital intelligence that can cross state and country lines. And so that's the main difference between a lot of things. and which is probably why they compare it to the to nuclear bomb, where it's like you can you can do local damage, but you can also send it you could also send it 10,000 miles across the world, and you can do horrible damage. So I think that's how they're looking at this technology where it's like And the thing about nukes is like at some point you have to come to an agreement, but we came to an agreement after some serious stuff happened. And I think that this is kind of what I think somewhat might happen because it's be it'll be difficult for the government to regulate it without Putting the labs in a position where our adversaries get ahead. Because at the end of the day, they're not always they're not releasing the most powerful stuff to us. They're not doing that. They even now they think they have not done that. They're always testing the most powerful stuff. And that's been their issue. Their issue is that every time we test this stuff, it's doing stuff we don't even like. It's hacking into stuff, it's breaking into stuff, Dre: Yeah. Rodney: it's creating viruses. So that's their issue. But the thing is, I don't see even if they created full countrywide re regulation, they still gonna have to do all that testing in the lab for national security. So that's not gonna solve the problem from from my that that's why I I don't think it's gonna solve the problem. Like if you if you if you say, okay, cool, the American people can only have a certain amount of intelligence and certain laws around it, cool. But unless you can get and trust China, Russia, and whoever else, unless you can trust them to also not do that extra experimentation, it's like, what's the point of the conversation? Can we get them to I think that's what's happening really? The question is not really, should we regulate it? It's like, can we regulate it without basically falling behind? And that's that's tough to me. Because if China comes to the table and and they agree with us, okay, yeah, yeah, we good, you ain't gonna do it, y'all ain't gonna do it. Would you like, are we gonna are we are are we are we gonna stop? And are they gonna stop? And that's that's a little bit of the hard part. Dre: Well well we well we already know China's gonna regulate their shit. We we already know that. They do it now. They they they they restrict internet access. Like I don't for a second believe that Chinese citizens will have access to the most powerful models. I don't ever believe that. Rodney: I I I agree with that. I agree with that. But but but isn't I'm speaking more so towards the from the government standpoint. And it like so I'm separating it. So like what I'm saying is that will the Chinese stop working on it from a from a from a from a war type of standpoint and defense type of standpoint, will they s will they Dre: Yeah. Rodney: slow down that development? And that's the concern. Because even if they even if both of us agree to Okay, your citizens and our citizens don't have a certain amount of intelligence. There's still like there's still the issue of are they gonna be experimenting with super intelligence, period. Dre: Yeah. But I don't think that ma I don't think that I don't think that matters. That's just like now, like China does a bunch of st stuff, right? Like they doing the same things we're doing. They experiment with bioweapons, we do that. They experiment with these advanced satellites, we do that. They experiment with nuclear weapons, we do that. So I I don't even think I maybe. I don't know. It it's just my opinion. I don't believe that it is the US government's policy or even posture to prevent China from doing what they're doing. I I don't think it's that. Maybe it's To be equal, sure. I think the bigger, broader concern from what I what what I gather is giving these tools to everyday people. The way I think about it is like 80 to 20 rule. It's like, do you want these tools being in the hands of 80% of the population? Right? Or are you all comfortable with the idea that, hey, let's restrict it for 80%, knowing that there's gonna be probably 20% of people that will probably have access to it and we'll have to accept the collateral damage that comes with that. Just like There are people who illegally get access to guns, they illegally sell guns, right? That illegally do stuff that are outside the traditional framework. But most people are probably following the laws and probably following the rules. Most people probably are. And in society, we typically can deal with the conditions where most people are kind of living within the framework, even though we understand that not everybody is gonna live in the framework. So I I I think that's kind of the position, that's kind of what I gather. From this regulation conversation, is you're right. It doesn't stop China, it doesn't stop any of that. But it's like, do you want everybody having the most powerful hacking tool in their pocket? No. Rodney: I I got it. I I got it. It's kind of getting mixed a little bit. I I I guess we I what you're saying. It's getting mixed a bit. Because I'm looking at the conversation from a standpoint of the conversation is like they're worried about super intelligence, period. Like not like because they don't have to release it to people. They they don't have to release super intelligence to anybody. Like they like Dre: Who is they? Who is they? Rodney: enthropic open AI, SpaceX AI, they don't have to like the stuff that they're doing in labs, they don't release the most powerful stuff every time. They just don't do that. They already don't do that. They don't have to do that. What I saw what I hold on. No they don't. No no th no they don't. No bro. The current models Dre: But but they do but but they do though. They they do have to do that though. Bro, they do. If open if unless unless open AI and Anthropic agree to do it, they they have to. Rodney: bro, the current models are way behind what they have in the labs right now. Dre: Ronnie, I I I'm not disagreeing with you on that, bro. I understand product release cycles. I'm what I'm saying is that you're saying they don't have to, they do. They Rodney: Mm. Dre: will continue to make the products better. You're saying they don't have, they do. If unless OpenAI and Anthropic sit down at a table and say, together, what what are we gonna release together? Bro, at some point, some the first time you released a super intelligent model, that's the first the genie's out the bottle. Like somebody's gonna do it first. So I I actually think they do have to do it. What I'm saying is that there needs to be rules to just govern what that looks like. But bro, that's market competition. They do have to do it. They they do. like they have to make the products as best as possible, right? Rodney: Yeah, yeah. So so so it's interesting though. Okay, okay, you i i in in even if they do, even if they do, I think that they're worried about it in the labs, period. It's our in the labs they're having trouble controlling it. To me, that's what Dre: Sure, sure. Rodney: the conversation seems to be about. I I mean it could be about the people and I'm not saying I'm not saying I'm Dre: Yeah, yeah, it's yeah, it's okay, the the conversation is happening about the labs, bro, because they're concerned that eventually it goes from the lab to being released. Like eventually it's gonna come out the lab, right? Rodney: Bro bro I hear you. I hear I I hear I hear you I hear you I hear I hear you supposed stay with me. I hear you though. I hear you. I I hear you. It it's Dre: So yeah, I think that's the concern. Not not just it's not the concern to me is not about what's happening in the lab. It's about the fact that, yeah, it's happening in the lab and eventually people will have access to this. If it was only that we're only doing research, where if it was just purely research and then like this is not about product release, just pure research, I don't think we'd be having this conversation. But I think it's because no, Rodney: Yes. Dre: it's it's eventually gonna translate to a product that people will use. I think that's concerning, not just the lab part. Rodney: That is th don't mean wrong. Don't mean wrong. I'm not saying that's not concerning, but I think it's also very concerning that what they're testing in the labs is breaking out and it's affecting the public. So they're all like so even what they're already testing in the lab is breaking out. Like they they like before they can even get to fathom bringing it to the public, they're already having these these I for example, Sam came out in an interview last week and he said that the model that broke out And attacked Huggy Face was not the latest model they just came out with. And that model is something that in the testing broke out into a public environment. And then Anthropic just came out on something I think either today or yesterday, where they're talking about how they their model also did the same thing. It was supposed to be in the control environment and then it broke out and then went into the public. So from my standpoint, I'm saying they're saying it's just dangerous period to be developing this shit, just to even be developing this shit. I hear what you're saying, that superintelligence is dangerous in the hands of people. But what I'm also saying is, what I'm also saying is they seem to be concerned about the fact that even in the lab, they feel they're feeling like this is we're losing control of this in the lab. Like before we even release it, this is getting out of control in the lab. It's it's breaking out of the testing environments, the controlled environments, and then going into public environments and hacking systems and creating viruses. This is documented. Th like this is starting to happen and it started to happen more I think in the last two months. And and and so I do see what you're saying about the p the the yes that it would be much worse in the public, but I can't even see how they would even get to public super super intelligence if if it's already doing all this at this level in the lab and it's breaking into the public. Like like like cause there were cru Dre: I don't get that same concern that you see about the lap. Rodney: let me tell you why. There there was a tweet. Let me there was a tweet. that they they were worried they're worried about the labs, like the the work that's going on in the labs. They don't want it to hack into a water system on accident. Like that like that bro that like like how do you even get to the point where you offer that to like people if you e if you even worried Dre: Yeah. Rodney: about that. Dre: But but see, that's my that's but see, that's my that's my thing is I'm I believe the real concern is. I I I hear you and I understand that. And I'm not dismissing the the concern, the narrative around there, but I I'm not I'm not I'm not believing or I'm not buying into the fact that they're so concerned about what's happened in the lab. I think the fear is the outcome of what's happening in the lab and the fear of that getting into and into being released and you're putting superintelligence in the hands of Of everyday people. I think I I personally believe I per and I I do see both, but I personally believe that like from Rodney: I I I g I I guess it could be both. Yeah, I guess it could be both. Yeah. Dre: the narratives that I see, for especially from people like Jacob, to me, that seems to Rodney: Yeah, yeah, yeah. Dre: be the the the narrative. Is that no, cause cause even he he did an interview where he talked about he said, you know, I'm not even and he acknowledged your your point. He was like, I'm not worried about the models that are out right now. It's not about that. It's about five years from now. Or actually, no, he I think he said A year from now, or whatever his timeline was. He was like, I'm worried about those release models, like, because they're gonna be way more capable. At some point, you you're gonna reach a certain capability in the model where it's just we've we've like you said the usefulness is there. So what else can it do? Right. So it's only gonna get more and more capable. So to me, yeah, I I I get the concern about what's happening in the lab, but again, to me, that's less dangerous than everybody having a thing on their phone that can hack the water system. And I get you, how do you get to that point? Rodney: It is li Dre: How do you get to that point? But Rodney: it is Dre: right now, there's right now, hold on. Right now, there's nothing Rodney: Yeah. Dre: stopping the release of that. I think that's the concerning part is that you're right. You're you're leaving it up to people at OpenAI that are being conscious and deciding that we're not gonna release this super intelligent thing out there that can hack water systems and do what we're not. But there's nothing stopping them from doing it. They could do it if they wanted to. Anybody right now, anybody right now, any let's say a an open open model from China. had the same capabilities as some of these powerful models that we see in the lab. And they just released it. There's nothing stopping them. To me, I think that is more of the concern. If because they could do it. Well I mean what would happen then? I I think to me, I think that's the problem. That's the problem that I see. Rodney: Yeah, so so the thing is is that yes, definitely. definitely so yeah, yeah, I I I think it's we definitely don't want the average person to have the ability to be able to take down critical systems. Nah, I mean we we we should like that that that should be That should be if if if they need the government to step in for that piece, then okay, have the government step in for that piece. Right. but Dre: But d you don't think they should though? Like Rodney: yeah, you know, we're looking Dre: You don't think they should? I guess Rodney: No, no, no. I think they should. I I think I think I I I I think that like like cause cause here's what I'm doing. I'm understanding more that everybody's not seeing this the same. And and it could be because of the way that we think and the or the information that we read and just like just like how we think about stuff. But I I see now that understand more so that you're from your perspective, you're looking at this like the main concern is people having these super like basically like the ability to do all types of crazy shit from the palm of their their hand. And and and yeah, I don't believe that people should have that. Like I I don't believe that people should have that. And I and I don't think you know an and if and if and if and if it requires the government to step in to stop that, then I'm for that. I have Dre: Why don't you believe people should have why don't you think people should have that? Why why why? Why? The first part. Can you answer that? Why don't you believe people should have it? Yeah. Rodney: Like super intelligence? Super intelligence? Dre: Yeah. Rodney: for the same reasons that you're talking about, which is like is that is that you shouldn't have super intelligence, unrestrained super intelligence. you just it it's it's too powerful for any one person, right? It's like it's like it's like having it's like everybody having an AK forty seven. It it's like, you know, you could buy one. I guess you I guess I guess you can I guess everybody could technically okay, you could you should be able to have it under certain conditions. So so you could have a AK, right? But but under certain conditions, right? So so like test and license and you know, are you crazy? Like so under certain conditions. So if the government if they need the government to to to pop in for that piece, I'm for that. I've seen I've seen other parts of the conversation talk about just the the the development of it, period. They're worried about the development of it, period. And because they don't want the electrical grid to go down. But like that like if they're if they're running a test. a a a a pre-training test for a new model. That's not no bro, if they take down an electrical grid on accident, that's like super serious. Like that that's like like if s if people wake up and they don't got electricity because of like some random testing in San Francisco. It's like what? It it that that's like stuff like that, it it's it'd be hard for me to see how they wouldn't have it be hard for me to see how they even get to offering super intelligence if they slip up like that. To me, it looks like they're they're they're just struggling to even develop this shit. They're starting to say we're losing a grip on this shit. This shit is this shit might end up doing some shit that's catastrophic to critical systems. You know what I'm saying? So I I agree with, yeah, like at the individual level, restricting intelligence to a certain level. And if they need the government to do that, then so be it. I don't believe that they're gonna cause based on like I just don't believe that they're ever going to I don't believe they're gonna offer. individuals that much power. I just don't think they will. I and and then and then but I I'm more worried about like I think about it like how they they say that like COVID was made in a lab, right? I think about it like that, where you're making something in labs and then it gets out and then it bec it causes like a whole whole pandemic. Right. So that's what Dre: Yeah. Rodney: so that's more so when I Like I'm seeing, I'm like, or that that's the frame on which I'm looking at the whole thing. I'm like, and and and I'm sure that it wasn't China's goal to create a pandemic. They might have been trying to create some new science, but it created something that created. So I think maybe that's what their concern is, is that or part like part of the concern. but I that's interesting. I I I Dre: Yeah. But I but I think but I think that's but but but I think that's where regulation also comes into play, right? Because regulation is not just about like the end consumer. I think they also are looking for guidance on how, like you said, how to develop these things in the lab properly. I think I think there needs to be constraints and guidance around that, just like it is in in the world of the the the you know some of these bio labs, right? Like there's constraints around that. You can't just pop up an infectious disease lab. anywhere and just start testing out diseases like in a lab. Like you just can't do that. Like you know what I'm saying? Like you Rodney: You can in China, you can in China. Dre: like it doesn't work that way. So like I I I I think I I think what these labs are asking for, I think is part consumer protection, which I do think is important. Cause right now there's nothing outside of the executive teams at these labs that are deciding what gets released and what doesn't get released. Cause there's no regulatory framework in place. And then I think inside the labs to your point I think there is real fear and concern around what they're seeing. And there probably needs to be guidance and probably needs to be oversight, maybe, and probably needs to be even other things that we may have not even thought about or are unaware of to help guide that as well, because that's in dangerous environment, just as it is in the public. Right. So I think all of I I think in both stances, you probably need a little bit of guidance on both. And I think restroegulation can be helpful. Rodney: They need they need guidance, but it's just an interesting situation where essentially like the government is not that much more smarter in this scenario. In this scenario, they the the experts are really the people who created it and they're worried and I don't and I and and it's gonna get interesting. There were gonna have to be a lot of collaboration. But I'm just gonna read this tweet real quick from Polymarket where Sam Alman meets with executives from major US power companies as OpenAI pitches its AI cybersecurity technology to help defend the electrical grid. This is coming off of some attacks that recently happened last month where China was using coordinated targeted AI agents to target Taiwan. And so the the what is what is tricky in the situation is that with government guidance, yes, get that, but the government may be in a tricky spot if you're in their shoes because you don't know to what extent the adversary is Taking super intelligence too. To what level are they taking it to? And and if you're and we're starting to look because what it seems to me is that it's starting to become okay, whoever has the best intelligence is going to take over. And so and so there might be collateral damage if the government feels like, okay, well, we want to step in, but at the same time, like, and we can step in, but at the same time, it's like they might say, like, okay. Take it to take it to its height. We'll deal with the public. We'll deal with the we'll deal with the public. Take it to its highest level. You know, I don't who knows what the government actually will end up saying. Because if you are because Sam's not going to the government for no reason to say, like, hey, look, like we let's defend the electrical grid. He's seeing for he's trying to be proactive, which I agree with. Be proactive. Dre: Yeah, a hundred percent. Rodney: Don't wait. Don't wait until there's a swarm of a hundred thousand AI agents. Attacking our electrical system and then now we can't even defend ourselves. Don't wait into that. Dre: Yeah. Rodney: You know what I'm saying? But at the same time, though, the public is gonna have a hard time understanding the cost of this. Because if if open AI are entropic, they're running these tests on superintelligence that the government has permitted because of of this cold war, and it takes down, you know, grocery stores or whatever it is, the the the the public may have a hard time understanding why. AI needs to exist, but the government might know for for different classified reasons. I don't know. Whatever it is. Don't get me wrong. I think the government should be involved. They need to be involved. They need to be in collaboration. I don't know what the what the actual and and and yet maybe like the all-in podcast, especially David Sachs, maybe he's a a bit more skeptical about Dario's specific intentions, about what he wants the government to actually do. However, I do think taking back his kit some of his skepticism, the government should be collaborating and they should be involved for sure in some way. it's just tricky Dre: Mm-hmm. Yeah. Rodney: when you have this country who is already attacking. They're already attacking us with with with this new technology, and we know that's not gonna stop. So It's it's it's they're worried about it, but at the same time, they're also probably feeling like, shit, like this is actually like national security. Like we gotta we gotta build this, essentially. We gotta keep building this. You know, if we don't build it, like it's almost like I guess it's like with the nuke thing. It's like if if the if it's a race to build nukes, it's like we gotta build it. We know it's gonna kill people, we Dre: Yeah. Rodney: know it's gonna bomb, it's gonna do all this stuff, but we gotta be we gotta have it. Dre: Yeah. We we do. I mean, I I I think we do. So maybe Rodney: It's tricky, man. It there's there's a lot to it. Dre: maybe it starts with just as simple as just restricting who can even work on superintelligence. Right. Maybe it's just like 'cause I I think the part of the problem I see Rodney: Yeah, yeah. Dre: right now is anybody can work on it. Like, like literally anybody. I I I was Rodney: That's true. Dre: just what I you know, I was just I said at the beginning of the pod I was in Italy, right? One of the guys who was there works for a quant a quant hedge fund, right? One of the biggest ones, huge. Rodney: Mm-hmm. Dre: And I and then we were just shooting the shit, and he was and you know, he was like, We have just as much GPUs than any comp any company out there, excluding like the hyperscalers, right? So, like, we can build pretty much the most sophisticated models that anybody like we got the most, we got we billions of dollars we spend on on on on GPUs, got our own data centers, everything. And they're just using this shit for just trading, right? Making billions of profits. So anybody today with the capital can go buy GPUs. Go build the most sophisticated models for whatever they want, however they want, do whatever they want and release it however they want. I mean, again, is is that the right way to do it? Maybe, maybe not, right? But that's what's happening right now. So if you got it, you can go build it. And so maybe you just need to just start with only certain people, only with certain licenses or whatever can work on a certain level of intelligence. Cause right now it's not like that. Anybody can do whatever they want, is wild, wild west. And maybe it does need to be there. And maybe it will stay that way until your point, until something catastrophic happen, like some water system getting hacked and people dying or like maybe that needs to happen. Unfortunately, for the US to take it serious and then figure out, okay, how do we need to Rodney: Yeah. Dre: step in and help prevent it? Maybe. Right? But yeah, right now it's like if you got the money Rodney: Yeah. Dre: and you got the resources, anybody can build anything. And it's not just anthropic and open AI that Rodney: The the thing I'm most worried about. Dre: are built in these frontier models. There are companies, like I said. Rodney: Yeah. Dre: These quant hedge funds that are building these models that are on par with some of these frontier labs, they're just not releasing it to the public. Rodney: Mm-hmm. The thing there's almost nothing I'm worried about. Like the systems I'm not like extremely worried about. Even though they're serious, I think a lot of systems can be restored and recorrected with enough resources and intention. and they can also be defended. I'm more so worried about what I think is the the biggest weakness that w we have is and I think Enthropic agrees, open AI agrees. Like I think bioweapons are the scariest thing that can possibly like those are the like Dre: Sure. I agree. Rodney: like like those are I I think something like you you create some new molecule that you can slip into water and it it's taste it taste you can't taste it. It's it's in the water. It's like it's just it's just you you you c you Dre: Yeah. Not for real. Rodney: could go into like that's the thing. It's it's the b like Dre: Yeah, yeah, yeah. Rodney: and so we're gonna almost need I'm not gonna lie, we're starting to go into sci-fi, but this is where the world is going. We're gonna need upgrades to our immune systems. And that's part of what the Kobe that the COVID vaccines were attempting to do. we're gonna need we're gonna start to damn near be computers, bro. Like, like we're gonna need like updates. Like we're not gonna be able to because Dre: Yeah. Rodney: we're not gonna there's gonna be new threats that pop up. Basically, we can't not turn around. These these like the bioweapon part, that's the part that's gonna be that's gonna be tricky, man. Because that that's gonna be tricky, man. That's gonna be the trickiest part. Dre: Yep. Yep. No, I agree. I agree. I agree. I mean shit, we've seen what happened what when you have something like a infectious disease like COVID get out. Like, I mean, Rodney: Exactly. Dre: so like you said, you slip something in the water system, it gets streak it g it gets dangerous real quick, real quick, real quick, real quick, real quick. Rodney: Like it's tricky, real tricky. All of a sudden people start getting sick and nauseous, and like you're not understanding why a whole state is nauseous. You're like, why why everybody nauseous in the state? It's like, it's like, right, right, right. Bro, bro, bro, like, like that, Dre: Yep. Man. Whew. And that could take a long time to fight Yeah, no, you're not lying. That yeah, yeah, you're not lying. You're not lying. Yeah. Rodney: we're gonna need like work. People are not gonna need to worry. Like, people's people are worried about jobs. I don't think. I don't think there's gonna need to be a worry about jobs. I think there's gonna be plenty of work to do. I think there's gonna be bro. I think there's gonna be Dre: Yeah. Yeah, yeah. Rodney: plenty of work to do, bro. Like, Dre: Yeah. Rodney: I hear people what they say, I I get it. Don't get me wrong, I understand, but I think there's gonna need to be plenty of work. Like, because Dre: Yeah. Rodney: I I can see I can see vulnerabilities, and if I can see them, imagine what the enemy see. Like like I Dre: yeah. Rodney: I can see like I okay I'll hit this, I'll hit that, I'll hit this, I'll I'll do Dre: Yeah. Rodney: this. Like bro, like the way my mind worked, I if I had bro, if I was going Dre: Yeah. Rodney: against the US bro, bro, like would would Dre: yeah. Rodney: I know, it wouldn't even be hard. It wouldn't be hard, bro. Dre: A hundred percent bro. And we do I'm and I'm sure the US is probably doing the same thing. Like, bruh, what was it like the Venezuela attack or Iran, something shit like that? Like, bro, we literally took over Rodney: right. Dre: the whole we literally just took over the whole system, bro. Like took over the whole electrical grid, water system, all like we doing this so I mean we have the same capabilities. So yeah. Rodney: Right. Right. Control the whole thing, yeah. Dre: Yeah, war would be tricky for sure, man. For sure. Rodney: Bro, it's it get tricky. Like it and it changes the dynamic of war because this is not just brute brute force soldiers in the way that we thought about it in the past. Like Dre: Yeah. Rodney: our soldiers are gonna need yeah, we're gonna we're gonna need like brute force still. We're gonna need physical that but we're gonna have to have like these super intelligent type of people. Like that that that like Dre: Yeah. Rodney: they're but it's gonna be different, man. It's not gonna be Dre: Mm. Rodney: Man, we on we on the verge of it. But yeah, that yeah. we could look so we spent a lot of time on that topic. We could do one more or we could end it, we could put a pin on that. But I think that was a good one, Matt, 'cause cause that is all of America's talking about it. All of America's talking about it. All of America's talking about this. And and Dre: yeah, it's everywhere. It's everywhere. That tweet almost got two hundred that tweet almost got two hundred million views on just on that tweet. Rodney: It's insane. It's insane. It's insane. I'm not gonna lie. This whole AI thing is crazy to watch unfold. Dre: Mm-hmm. Rodney: It it's it's and I see I see all the sides, you know, like I I don't want people to think that I think everybody should have you know like extremely, extremely extremely super intelligence in their pocket. I don't believe necessarily believe that. do I think that there it might be tricky with open source models and and and and maybe d it depends. depends on how much control like I think you can slow down people. I'll put like that. You can slow people down. So like even if there's open source models, you can make it so that Apple and Microsoft and all these companies make it harder to run those models and stuff like that. Obviously, if you're clever, you'll be able to get around you you'll be able to get around it. But I think you could still slow people down. You'll be able to slow people down. Dre: Exactly. Yep. Yeah. And most most people wouldn't try to get around it. They'd be like they might they might try and they'd be like, All right, cool. You know. But again, it it's always gonna be that small percentage of people that will. You know? That will. And that we gotta just Rodney: Yeah. Right. Yeah. Yeah, yeah, yeah. Dre: be okay with whatever the collateral damage is, I guess, to that that subset of that subset of people. Rodney: Man, man. Yeah, it was a it was a good topic. Good topic, man. yeah, should we should we are we good? Yeah, that was good, man. Dre: Yeah, yeah, I think I think we should end it there. I think that was good. Yeah, yeah, I think we're good, man. We'll catch y'all on the next one. Rodney: Yeah.