Latest / Systems and Power / Why Everyone Is PANICKING About AI, Is AI Doomerism Overblown? - Niel Pal | Systems and Power #14
Transcript
- Moses Parente: Hello everybody, welcome back to the Systems and Power Podcast. I'm your host, Moses Parente. Today we have a special guest. Thank you for coming back again on the podcast, Neo Pal. Thank you so much, man. Niel: Of course man. Yeah, really happy to be back on. Lots of lots of good stuff for us to talk about. Moses Parente: Of course, the whole world of AI agents has been exploding as of recently. Would you like to go tell us the audience and discuss a little bit more about what's going on, especially with the new models that are models that are coming out? Niel: Yeah, so at least for me, I would say GPT five point four kind of shocked my world. Really big fan of that model. Five point five is great as well. And now they came out with GPT six, Seoul and Astra as well. So yeah, it's really good. I've noticed that in terms of like a really significant increase in model quality, I wasn't able to see that necessarily, but computer usage, the fact that it can take like autonomous tasks on your behalf, that seems to be the big trend that's happening right now in AI. Moses Parente: Hmm. Cause I remember I saw that that what that was one of the features in GBT Astra, and on X everyone just started freaking out and saying, like, my gosh, the end of humanities here, and essentially that we're going to take over the like robots are gonna take over the world. We're all gonna be living in a Terminator like crisis as well. I personally don't believe that's happening because I remember I was talking with another e economist whose name was Daniel DiMartino and he said that If that was to happen, then stock prices would be plummeting like crazy because money doesn't mean anything if the world's coming to an end. Niel: Yeah, no, absolutely. You know, I don't know if this is like some sort of like way of boosting their stock. Obviously look, these models that are coming out are great. But I do think some of this AI doomerism, a large part of this is like being funneled by these companies themselves. OpenAI anthropic. I don't know this is a play to like boost up their IPO price, but like I think again, right, the doomerism approach to AI and just like the fact that it's all gonna screw usselves over, it's overstated. Moses Parente: Mm. Do you think that there is a little bit of truth into what they're saying that we need we need to be regulating AI more on the on at least a federal level and start like increasing taxes on it? Is there a little bit of truth to that or there is some sort of bigger play here that people who are running the tech companies are trying to do? Niel: I think like I said, right, like a large part of this may be just hype for the IPO. But in terms of like having regulated protocols, now I do think that with those open weight models, where you can fine-tune it on your own data, you're gonna find people have models that are maybe even comparable to some of these frontier models. You know, like maybe the lower level ones. and with that, there has to be some sort of like basic level of standardization. there's gonna be a big security threat with just these new models that are coming out. So having like a unified kind of thing in the US at least, I think is gonna be a good play moving forward. Moses Parente: Speaking of security incidents, I I believe the SEC just got hacked by a rogue AI agent. Niel: Okay. Moses Parente: Okay, l let me go pull that up real quick. Just SEC hack AI agent. Don't don't mind me over here. Hold on. Yes, OpenAI disclosed that an autom an autonomous AI agents went rogue and improperly accessed or probed several US government websites, including the US Securities and Exchange Commission, SEC. Census Bureau and the education department. That's not looking good. Niel: as not. And I mean, was there any like sort of basis for this? Like whose agent I guess was there any information on like whose agent this was specifically? Moses Parente: I'm looking ri right now I'm looking on what Google is saying and it says that OpenAI stated that the agents retrieved public data or interacted with the sites unexpectedly during testing through no non public SEC data or cr though no non public SEC data or credentials were compromised. But some of the key details of the disclosures, targeted agencies again, it was largely the SEC, US Census Bureau. The nature of the activity, autonomous spots use web access and developer tools. Misaligned ways such as copying and reposting public SEC info onto external forms rather than executing malicious data breaches. For a broader context, follows recent high profile disclosures regarding autonomous AI agent swarms bypassing security protocols, including a July incident revolving AI startup hugging face and a June breach of an Australian government health portal. I think that you're saying? Niel: Yeah. So No, I'm saying like it shows that there are security indic incidents and there needs to be probably some protocols in just how we go about rolling this stuff out and making sure it goes about safely. Moses Parente: yeah. Do you think that should be done more on like a federal level or a st or a state level? Because I know every one states they all kind of want their own individual autonomy and but in terms of like there should be like a federal level law that says like hey you gotta follow these certain regulations for pretty much everyone's safety. Niel: Yeah, you know, so I guess like from a political standpoint. Personally I am somewhat of a champion of like states' rights. But in this case I do think it has to be on a federal level. There's just commerce, at least with the United States, right? Like it's gonna be between states and I think it's helpful to have a unified kind of system there. Moses Parente: Hmm. In terms of the newer AI agents, could you 'cause I know a lot of people are starting to use more and more AI agents as they expand, what do you think are like the coolest ones that are being featured and if you're using any any new ones? Niel: Yeah, so you know, I was waiting a little bit to get started on these. I tried out Instinct. Shout out to Apurv, one of the investors at Ultimator. He put out a like an invite link, so got to check that out. That's a text-based agent where it pretty much can just kinda ping through like your calendar, can look through your email, and it just helps you organize your life. So like not obviously the most cutting grade technology, but it is quite nifty and handy to use. And in terms of valuation, I mean they're apparently looking to raise a billion at a valuation that would put at the same valuation as linear. So like it shows you that there's like so much hype in the space. All the VCs are using it and yeah, it's it's growing quite a bit. I don't know if you also saw two meta release Muse. This was with Alex and yeah, Moses Parente: Mm, I saw that one. Niel: yeah. So Meta has kind of been underperforming in terms of AI, right? They don't have their frontier model just yet, but they've got the scale AI team. And Alexander Wing. So he's actually leading out Muse there. And honestly, man, I think it's really quite potent. Has the persistent computer, which is thing everyone's talking about. So like Grockbot, if you've heard of as well. It allows you to run a lot of different tasks and have the compute capacity, unlike using like a traditional chat agent. So yeah, there's Moses Parente: Hmm. Niel: that. Overall, I think it's really sharp and they have a ton of your data already, right? just being integrated within your Instagram, Facebook, et cetera. Moses Parente: Hmm. I started using this one AI agent called Hermes AI. I thought that one was pretty cool, especially as it apparently it's supposed to it helps you it helps learn at the more you use it. Especially when you're doing things like different projects. and overall I'm still very ne very new to Hermes, so my knowledge on it is definitely a lot more limit limited. But I saw one of my other favorite podcaster's Michael Sartain, he was a very big proponent of Hermes AI and I said, you know what? Let me go check out Hermes and see what this does. And so far, even though I'm just I just started using it, I'm liking it. Hopefully when we if w when we can record again, I can give you a a bit a a more better update on how that's going. Niel: Yeah, absolutely. Is sorry, just to c to clarify, Michael Sartain, this guy live in Vegas. Moses Parente: yes, that that guy. Niel: wow, okay. I I wouldn't expect him to be a Hermes guy. That's that's cool. Moses Parente: I don't know. He was a big proponent of Hermes and also very pro AI. So honestly, I'm a big fan of it of tech enthusiasts and people who are willing to embrace new technologies. And when I and I've been watching him for quite a long time. So when he started being a proponent of using AI in your day to day life, especially agents, he was a very big proponent of using AI agents. And one thing that a lot of Americans at going back to our last episode One thing that you brought up was that we Americans don't really use AI agents very often. They might use like ChatGPT or Claude, e maybe just stand the free version, but overall they're not really using AI agents to help their day-to-day life. But now with the with these new tools that are coming out and more agents are being produced. I'm glad to see people bring more awareness to see like yo, these AI agents are actually pretty cool. And once you get used you can see the stuff that they do, it kinda blows your mind on like, wow, imagine like what kind of stuff would happen if we just keep on using it and how much Niel: Absolutely. Moses Parente: more easier just makes your life. Niel: For sure, right? And there's a compounding effect too of like using it for one instance and then seeing like if I scale that out, how much more work could I do at scale? And so like now it makes everyone almost a business owner. Moses Parente: yeah. And especially the amount of like like the capacity that you're able to do. Like before like back when the Industrial Revolution came out, you would needed to have a whole bunch of capital, you needed buildings to bait have all these factories. Now you just really need some AI and a laptop to just start building your own products. And it Niel: Yeah. Moses Parente: because it's just so much easier to do things like that, imagine the scale that you can expand on. Niel: Definitely. Moses Parente: What was one thing that also did just recently happen in New York City is the Mayor Mamdani banned AI in schools. What is your thoughts on you on putting city w I guess just banning AI in schools in general? Personally, I think that this he actually got this one right. A lot Niel: Yeah. Moses Parente: of people might be a bit more controversial on this and w find what I say to be not necessarily, I guess, mainstream, but I think that kids should not have access to social media or AI, especially early on. I think that you should be able to develop your own instincts and critical thinking skills first before you end up getting exposed to all the sorts of technology. Niel: Yeah, you know, I would actually agree with you there. just to clarify, is this entirely K through twelve? Moses Parente: Let me go check it up just to just to just to just to make sure that what I'm saying is is correct. 'Cause I read it on a New York New York Times article and ban and New York City AI ban K through twelve dot news don't mind me. Mm. Cause honestly one of the things that I I I stand by on is that people should not be exposed, especially younger kids, to like social media. Because I think it just kind of rots their brain and it makes them feel what they're doing is just not inadequate enough because you're always comparing yourself to everyone else. Niel: I agree. Yeah. Yeah, I mean like just looking back, you know, I think taking a step back. We sorry. Moses Parente: Yeah. Yes, it is sorry, you were saying? Niel: yeah, okay, I was gonna make a point saying like, you know, we grew up in an era where social media kinda like happened in our lifetimes and we were able to see that grow. And you think about like now, you know, I still use Instagram, you know, and it's cool to see it as an adult, right? You look at people's stories, see people doing great things with their lives, it's inspiring, you know. I'm able to keep up with people. Large portion of it when I was younger was playing that comparison game. And especially when your brain is developing, especially when you haven't fully kind of gone out into the world and like developed, I guess, some perspective on things. I think it's especially harmful. So yeah, definitely agree with you on the social media. so if you were saying like if it's entirely K through twelve, I have some has to Moses Parente: Okay, it's actually a K through it's it's 2K through eighth grade. That's where I I I got it wrong. I knew that there was Niel: Okay. Yeah. Moses Parente: an a name a citywide ban on on AI in schools, but I mistakenly said K through twelve. That's my mistake. I apologize. It is from two K to eighth grade. And it's gonna be in it's going to Niel: Yeah. So two K through eighth grade. Yeah. Sorry. Moses Parente: be in effect for about one year. It's a one year moratorium. Niel: Yeah. Yeah. for me that makes sense. I think that bold move to make, and I'm sure there's gonna be some pushback, but in terms of overall development, I think these are critical years where for a large part, like the stuff, you know, you could use a calculator, but you're not allowed to, right? Being able to prevent AI just for these formative learning periods where you're building up these foundational skills, I think it's totally fine, but as they get into high school, I think it's important that Every educator in this country have a conversation about like how do we responsibly allow people to use AI and do it in a way that doesn't detract from their brain power and allows them to scale, you know, a lot more efficiently. Moses Parente: Yeah. Also at to add on to the a the AI band that was in that was implemented in in New York City by the mayor, the for ninth through twelfth grade through high school, they're actually doing some sort of pilot program that gives a limited access to AI that is controlled by the teachers. And personally I'm in favor of this because I think around that time because if you want to be successful in the real world, you're gonna have to learn how to use artificial intelligence. You're gonna learn how to use you're gonna need to know how to use AI agents and up and a lot of the learning language models if you wanna be efficient at in the workplace. And one of the things that they do is that they allow a small number of AI programs and controlled pilots under teacher supervision, which for me I'm a hundred percent in favor of because you're kind of giving a slow exposure into AI for kids. Because let's say kids are not exposed to AI at all, but then eventually you're gonna have to use it in the real world. Then you Niel: Hundred percent. Moses Parente: kind of give it to them in a more controlled dosage and kind of like a trial in a way. Like a tr a trial. Like you know like when you're playing a game and you get like a like a free version of it and then you're you you have like you see some features. You're kind of just getting your feet wet and then by the time you get to college when you have you can get all the full access you want because by the time you finish high school, you're pretty much you can make your own decisions by by that time. But you're more prepared on what to do while also maintaining your critical thinking skills to make the right decisions and to be successful in life. So sh so shout Niel: Yeah, I mean it's sorry. Moses Parente: out to Mayor Mamdani for doing that, honestly. This is one of the the times I think he he got it right. Niel: Yeah. I mean, you know, I I agree with you what you're saying here. And I think it's gonna be interesting to see just, you know, how it kinda evolves over time. Like for example, our last conversation we talked about in university. Should people be given access to AI and like full fledged access? I don't know, right? even when I was in the workplace, The general kind of mantra we have is that you're allowed to use AI, but you always gotta double check. even like Clay CEO made like a viral announcement on this, like our AI mandatorium. And you know, I think that like there's a difference between the workforce and just your education, but as a whole, we still have to maintain the human experience, particularly when we're developing these people, right? Making sure that we're trying to use it as a tool to enrich them and then eventually give them more responsibility as they get older. Moses Parente: Hmm. Okay, okay. in terms of the speaking of politics, now that we were mentioning it, what is your thoughts on the midterm elections? 'Cause right now I the Democrats are are poised to win the Senate and the House all all on a clean sweep. Niel: Yeah. Yeah, you know, I am a little surprised by that, but I do at the same time understand just the frustration that's happening within America. I moved back to California. I don't know if I told you this, but yeah, dude, I looked at the gas prices. It's six sixty right now. Which you know, I think that the average American is being pinched by inflation. The jobs are not necessarily increasing at the rate that our economy is moving at. And Overall there's a lot of frustration, so I can see that. I don't think Trump is doing a particularly great job of unifying the country. I don't know if you saw his thing where he said that if they end up winning some sort of election, he's gonna personally give five thousand dollars to people. Moses Parente: my gosh. That is basically pay to pay to play bribery at this point. The way how I see it, I mean, he was saying that if the the GOP, the Republicans, win the midterm elections, he's promising every American five thousand dollars. This is what this is what I have to say. I mean, look, Trump has already promised everyone the Doge dividend. Where's my money? Niel: Yeah. Moses Parente: He he says a lot of things and then it just doesn't happen. And when he campaigned on all largely a populist message on and addressing certain affordability cr certain affordability issues and then just didn't deliver, of cour of course there's always going to be there's going to be resentment toward towards the current administration. And right now the Republicans, as of the rec as of this recording, are in control of the House, the Senate, presidency, and the conserv and a conservative majority on the Supreme Court. So of course like when they're running pretty much everything Niel: Yeah. Moses Parente: There's going to be resentment towards the ones who are in power. Niel: Agreed. I do think it's a little overstated though. Like I definitely see the Democrats taking the Senate. for the House, I think it'll be probably a more narrow race than we're expecting. And yeah, I mean, I don't necessarily see it as like a hundred percent clean sweep, despite, you know, obviously there being frustrations there. yeah, I think that generally speaking, even prediction markets to some degree, underestimate the conservative votes that are out there, and I think it'll end up being a more narrow race. Moses Parente: Do you think that the conservative votes are largely doing that just because they're they're they're having a mindset of I must vote Republican or I must vote GOP no matter what? Because I think that that now I mean tet I largely thought for the past several years that the and any hope of Texas winning a Democrat winning Texas on a state le statewide race was just scope. And for a large part it it was. Trump absolutely dominated in Texas in twenty twenty four. But right now, like with the polls that you're seeing now and also on and the prediction the prediction markets, James Salarico's looking to win that. Niel: He is, yeah. And I think, I remember back when I was in eleventh grade, there was an AP English, we were taking the AP English test. And the question was like, what do you think is the greatest skill in politics? And or sorry, is artifice the greatest skill in politics? And I don't think it's artifice, but I do think it's charisma. Being able to unify people and really get them to believe and embody the message that you're sending, I think that might be the most important skill. So what Tal Rico is doing, El Sayyed as well. in Michigan, these guys are taking a populist approach, which look, for a vast majority of people, and not vast majority, but like a significant portion of this country, that just does not appeal to them. But when you talk about these values, these human values that get people, you know, excited and believe in something, I think that's, you know, what's gonna lead to more success. I saw that with the Obama campaign. Bernie Sanders too, you know, is definitely more of a hardcore leftist, but he had a lot of those similar characteristics as well. Moses Parente: Hmm. I'm personally this I don't know if you'll you'll agree with me on this, but I'm not the biggest fan of always campaigning on a popular on a populist messaging because it largely relies on always everything being bad news. And then when you end up getting in power, you're going to have to rely on things continuing to be bad because that's what got you elected in the first place. And I think one thing that one problem that we have with our government at as of now. Is that most politicians they care about being reelected. They s they want to stay in power. And when that becomes the the pr the main priority, instead of putting the priority on the people, because their their main goal is fighting to get elected. Whether they come from a marginal district or they know that or they're in a deeply democratic or Republican district, but they always want to make sure that they're They have no primary challengers. They always want to be reelected. And also the term limits the terms are not that long. They're their main camp their main concerns are I have to get elected every next four years or if they're in Congress or state legislator seat, it's usually about two years. And when your main priority is I need to get reelected, you tend to forget the actual problems that people are facing. Niel: Yeah, no, absolutely. And I think like for that reason you're seeing some of these candidates probably have a little bit more success in terms of just getting that vote. I do think that there is somewhat of like a tone-deaf approach, to some degree, maybe the Republican Party, where like I'm seeing more Democrats take a slightly more populist approach where the especially the cost of living crisis is being acknowledged. this is something like Tucker Carlson talks about a lot. And it's certainly something like Republicans, like On average, they're probably in a lower socio economic status. So they are going to be sensitive to these inflation increases that are happening. And when politicians don't make that a priority, especially in this economy that we're in right now, I think you're ultimately gonna see people turn on. Turn towards maybe, you know, even voting Democrat. Moses Parente: Do you think that this will also just discourage people from I guess like people in who are largely in Republican areas, do you think that this will discourage them from turning out in droves kind of like how they did for Trump in twenty twenty four? Niel: Yeah. yes, I do believe that. Now, do I believe that it's gonna be a surefire victory for the Democrats? No, I don't, but I do believe that large part of, you know, Trump winning, right? Was Democrat apathy. Maybe they just weren't that passionate about Kamala as a candidate. You know, the largest voting block is still gonna be the non voter. So you have to figure out like, you know, how do I get Moses Parente: yeah. Niel: this voter empowered and ready to go out there, you know. Campaign, get in line, vote. And when you have apathy on one side, I think that's be gonna be a major opportunity for the other one. Moses Parente: One other thing that going about AI Dumerism, it's it's becoming actually a somewhat, I guess, popular campaign strategy that a lot of people are are going on, especially in populist Niel: Yeah. Moses Parente: times. They I mean you saw it with Alex Boers' campaign for the New York twelve congressional district that happened for the mid midterms in in New York. He largely ran on a campaign that was talking about, hey, we need to have AI safety and N and he was also like being backed by Anthropic and a lot of the their their their super PACs that were largely pushing money to help candidates that were pro AI safety. But a lot of the big big tech super PACs that were I guess like backed by Anderson Horowitz, AC A sixteen Z, they were largely pumping a lot of money into candidates who were pushing for less AI r regulation and that was one of the big proponents of that race. Then he ran on the messaging of big tech's out to get me. Help me fight for you guys. Now he ended up losing, but he played second and did fairly well for himself considering the now Democratic nominee Mike and the now Democratic nominee Michael Lasher. He was pretty much backed by everyone. He had the Michael Bloomberg super pack. He had Kathy Hochel's endorsement, Jerry Nadler's endorsement, he had the whole Niel: Okay. Yeah. Moses Parente: institution basically just rally just like essentially securing the victory for him. And he but Alex Bors brought a popular campaign that a lot of people are starting to do as well. And for the Republican side, you saw James Fishback talk a lot about hey, we need to get rid of AI data centers. And Niel: Yeah. Moses Parente: e even though he got ten percent of the vote, he won he got a majority or of a plurality of the young voters under thirty five. Niel: Okay, wow. Moses Parente: So Niel: yeah. Moses Parente: So I was wondering your thoughts on like people and politicians who are running for office using populist messaging in terms of b on AI doomerism. What is your what is your thought your thoughts on that? Niel: Yeah, it's important to take a nuanced approach. And so I don't necessarily like for example, Fishback. You know, I've seen some of his content on Instagram. He's certainly catch eye catchy. I don't think that's necessarily gonna be what unifies the country. But going back to your point of, you know, we have to figure out a way to regulate this and care about the American people, but at the same time recognize that this is the industry that we're living in, right? This is a revolution that's happening, and like it's not gonna stop. We have to figure out a way for it to make sense. And make sense to roll this out responsibly and continue to help our humanity. So I actually, to be honest with you, did not look too much into Alex Boris. we'll definitely do my homework later tonight, but yeah, it's going to be interesting to see just, you know, which candidates take that approach. frankly, I believe you're gonna start seeing anthropic, open AI, NVIDIA. They're all gonna be super donors in the selection cycle. Moses Parente: Mm-hmm. Niel: So, you know, it'll be it'll be cool to see, but I very much, you know, would be interested to see a politician that takes a sensible approach to tech policy. you know, I keep up with the all in guys, you know, those podcasters. and Rokana Yeah, Moses Parente: I love them. Niel: yeah. They I mean, look, I think that they probably lean a little bit too more too too much more towards the right than I do. But I think they have a sensible approach. And I've just been like very disappointed with Rokana you know, here in California. This guy is supposed to represent Silicon Valley, yet he takes such a anti capitalist approach, which let's be real here, like capitalism is what drove all these amazing products. So we can't take away that. We can't take away the special innovation, but at the same time we have to figure out a way where these companies are receiving massive amounts of profits and it's just gonna be a huge revolution that's happening. How do we share that in a way that continues to enrich the rest of our society? Moses Parente: Speaking of Rokano, I guess you were saying kind of going a little bit too far to the left, even in a and representing Silicon Valley, which is pretty much like capitalist nation right there, you know? And you a lot of them, I guess the way how I like to see, they they lick their finger, stick it in the wind and see which way it's blowing. And I think because everyone's I guess are they're running on a left a left populist rhetoric, he feels that he has to go do the same thing to avoid primary challengers. Do I agree with that? No. I don't necessarily like it when politicians kind of like flip flop on their opinions just on the way the the wind is blowing. But do I see like I guess why he's kind of doing that? I I understand, but I'm not a fan of it. Niel: Yeah, yeah. No, I agree with that. Moses Parente: Yeah, especially but overall back to what you were saying about we need politicians who are going to take a sensible approach on AI. I think that with the amount of on how fast the pace of technology is advancing, we are going to see a lot more politicians who come from tech backgrounds and can actually speak on, hey, this is what we need to do and kind of like have be like the the voice of reason. Or like the tech experts, whether it comes to AI, security, a and anything else regarding that in Congress and the Senate, which would be nice in my opinion. And I think we're going to see more of that. you're saying? Niel: Absolutely. Yeah, I mean I was gonna I yeah, I was gonna ask you. you went to see the event with Cliff Asanus, right? Moses Parente: I did. Niel: Yeah. was yeah, was gonna ask about that and particularly did he mention anything about this AI super cycle where he sees regulation. Moses Parente: He di he did he did a little bit. I wanted to ask a few more questions, but then he ended up having to to leave. But Cliff asked us when the he was talking largely about taking risks and it the event largely talked about had a more focus on people who come from low income backgrounds, but because of the where they were born, how much of a risk are they able to take? Because everyone can say, take a risk. To go take a bet on yourself. But how much of a safety padding do you have when that happens? And a lot of people who come from low low income backgrounds, it could be very hard to do that. But like what can they they do about it? So that's largely what the effect does. But in terms of AI doomerism, I was speaking with Allison, who was the moderator of of that event. She was promoting her book. And she said something in regards to we were having a very short conversation. She said something that look, everyone everyone is largely scared of new technology. And no one really likes change, especially when it's coming at an extremely rapid pace on how we see it. Like back in twenty twenty two when ChatGBT came out, it was just it was just like a chat bot. But now it's doing things that no one could have ever imagined just a few years ago. Niel: Yeah. Moses Parente: And people are just naturally a little bit more resistant and as and afraid of change. And that's largely what we're seeing. Do I think there's some truth to that? Yes. But do I also think that we should be a bit more wary of things that are going on? Yes. And a lot of people also like to bring up the race with China and Russia. I think largely Niel: Mm-hmm. Moses Parente: as long as we we have good relationships with Taiwan and the chips keep keep coming keep coming keep coming through, we should be in the lead. You might have different takes on that and I would love to hear it. But I think we should still be far ahead in the race to the point where we can actually start talking about AI safeguards and common sense guardrails on everything. Niel: Yeah, no, I totally agree with you here. one thing that I'm particularly interested in, I mentioned it the last time as well, is like these open weight models. The majority Moses Parente: yes. Niel: of these are still coming from China. So it's, you know, Kimi Quen, GLM, Deep Seek's another one too. So a lot of these companies, right, they are using these open weight models that are a lot cheaper coming from China. They will use their own proprietary data to fine tune it and make sure that, you know, like Maybe it's not as great as the Frontier models, but it's trained on their data and it's a lot cheaper. So my concern though is that there might be just Moses Parente: Mm. Niel: some sort of regulatory risk. Why and you know, why is there no US based open weight model that people are using? So that I think that is probably going to be something that the government is ultimately gonna make a push on. Moses Parente: Hm. Speaking of the of politics again, did you see the all in podcast? The besties they were interviewing Jensen Huang and then Donald Trump ended up calling Jensen in the middle of the podcast. Niel: Yeah, yeah, I remember that. Moses Parente: Now I I thought that that part was funny, but if the the one thing I like d Trump was saying was that all this fear around AI, it's all a hoax. It's all a hoax. Do you I personally think that's a little bit tone deaf, especially coming for the president, because we we people naturally are afraid of change. But hold on one second. Awesome. Yes. And it wa I largely thought it was a little bit tone deaf of the president in terms of saying that, it's just all a hoax. Don't you guys don't need to be afraid. I think that's kind of just like ignoring a little bit of the public opinion on AI and and the public's not always is always correct, but I think concern should be addressed. Niel: Yeah, it is funny like how I guess, you know, the president is quite involved in the AI super cycle. Both him and J D, so yeah, you know, I think that there is reasonable concern. obviously, you know, for the president, you know, he cares about the these AI companies, so he'll have his foot in there. But yeah, I mean I did find it kinda toned off as well. Not I guess really much colour I can put on there. It's just, you know I think that the president will be involved with these AI companies. He cares about the stock market and just where we're going. To some degree, you know, it is tone deaf, but at the same time, I do think that it is putting our country in a good place as well. Moses Parente: Hmm. In terms of like technological advancement and making sure that like, hey, we're on top of the A AI race with China and Russia. Niel: Exactly, yeah. also too, just from like a consumer business standpoint, like we wanna make sure that we're kind of pushing the frontier on what could be something really exciting. even like just not even considering the competition from China. Just like as the country, we wanna be kind of, you know, helping to find this revolution. So I appreciate the president for the support he has shared out for these companies that are leaving the space. Moses Parente: Mm. I think that all of this is just gonna continue showing how technology and AI is just going to be affecting our politics on a day to day basis and that it's something that people need to be a bit more aware of. Niel: Yeah, totally agree. Moses Parente: And in terms of Hold on one second. Yeah, sorry, someone was just was just get getting attacks by it's all good, all good. But yeah, in terms of the prediction markets, I remember some people were saying that prediction markets that are going on right now are actually helping people do more insider insider trading. Would love to get your thoughts on that, especially as someone we are both fans of prediction markets. Niel: Yeah. you know, it's funny enough. I don't know if you've heard of a company called Rogo. Moses Parente: no I I I haven't. please tell me about it. Niel: Yeah, Rogo is pretty cool. they are doing like AI, particularly for finance workflows. So a lot of like investment banking analysts, you know, preparing just their pitch decks, just like the data rooms, they have taken a lot more ownership on that. So very fast growing company. Unfortunately didn't get the offer, but did interview for them and I did a pitch on Calci for them. So in fact, you know, like they did ask me quite a few questions about safety. from what I gather There is a thing where they do capture your social security number when you go and make these bets. And they have their own trust and safety team where they're able to do these investigations and based on, you know, your information that they gather, if they'll try to see if anyone is connected to you that is making these traits. Of course, I'm sure there's numerous violations that are still going on, right? These are not exactly set in place, but it is going to be an interesting thing, like even for myself. I was pretty involved in the Snowflake data workspace. I know for a fact what okay, like I made a decent amount of money, like $800. I know for a fact that they're gonna talk about open models. Snowflake released their own model router, so when they're doing that earnings report, if you're somewhat familiar with the space, you have decent intuition that the public doesn't have. I mean, I don't know if you consider that insider trading, and to what degree is considered insider trading. Obviously, Calvin Ridley Betting on his team before he's about to play a game. That that is fair to say is insider trading, but I don't know like where we could constitute what is what. Moses Parente: Because every time I hear about calci polymarket incentivizes insider trading, they usually refer to extreme examples like that. Like let's say you're a member of Congress and then you know some sort of legislation is gonna get passed and includes XYZ in the bill. And then they bet on that. Or as you said, someone was a a member of a sports team, bet on himself right before they played. Or even as extreme as George Santos betting on himself that he was going to be at the s the State of the Union. And when you and and they use extreme examples to s to to try and show that, open mar open prediction markets like Polymarket, Calchi are actually really bad and that they need to be stopped. But I think that they you that there is a little bit of fear mongering there and trying to show that, this is like actually bad. It's It's largely l it's a little bit like every market. There's your good and your bad there's your good and your bad. Like there's pros and cons to everything. I don't think that using a few extreme examples are something that is that should be just highlighted as the reason why something is completely horrible. Niel: Yeah, agree there. Moses Parente: Yeah. And in terms of I guess going back to also one thing that I did want to talk about on the the last on the last episode, but we didn't get a talk to a chance to talk about on this one, the rise of GLP ones and Ozempic and the contribution to lowering obesity. I would love to get your thoughts on that. Should people be using it and or if you're not necessarily you don't need it, should you be not not be using it? Niel: Yeah, I take approach that an approach that I think this is going to be a revolutionary thing, maybe even more so than AI to some degree. The impact this has on our obesity rate, which coincidentally affects, you know, our overall just healthcare expenses. This is something that is quite revolutionary and I imagine there's gonna be more easier to administer kind of approaches to take. The idea of pinning yourself like That kind of turned me off from steroids as a kid, just having to stab yourself with a needle. But yeah. You know, I'm sure there's gonna be like an oral form of a GLP one, and more and more people are gonna get on it, which counts sound like we leave to a healthier society. One thing that I saw that really kind of shocked me, Bank of America is offering full coverage for GLP 1 weight loss drugs. They're spending over $250 million annually on these medications. And Speaking to their executive staff, they said like look this is worth it in terms of the quality of life we're pro providing for employees. We see this as an investment in their future and they're happy to pay that two fifty mil. Moses Parente: Wow. Do you think that is something that companies should be pi putting into people's healthcare coverage pl plans and include GLB ones? Niel: It's a perk, right? Yeah, it's a perk. I don't think it should be mandatory. For example, I know a lot of these tech companies, Meta, I believe, is one too, where they offer egg freezing up to maybe 10 years. And look, that's not a cheap procedure. But if you do want high quality talent, these are some f benefits that you can offer, right? That will kind of sway and make a difference. In terms of specifically like the GLP ones, like The quality of life increase that it's offering, you know, I'd again like two fifty million, right? Bank of America is a huge company. But being able to offer something like that, I would say is honestly on the equivalence level of say like a four hundred one K match in terms of how much value you could offer your employees. Moses Parente: Hm. Personally my own take on GLP ones is that 'cause I'm I'm still so much used to the okay, if you want something, you're gonna have to work for it. And we we we constantly saw a lot of transformations that you see on YouTube, like, I w I I worked out I used to be obese, but then I went on d a diet and then I lost weight, and then and then you have you had something like that. And then the the that's the very stereotypical from fat to fit. Because you ate you ate right, you went on a diet, you worked out, you did some cardio, you lifted weights, and you got in shape. And but now with GLP ones, you are able to basically just kind of shortcut all of that. And for me that's a little bit hard to wrap my brain around, especially since this is all just so new. And I'm just like s I'm s kind of struggling in a way to kind of have complete acceptance over it because if you 100% need need GLP1 Zozempic for let's say like you have type two diabetes or you're morbidly obese and or you've been extremely fat your whole life, you might as well take the sh you might as well do it. Because Niel: Yeah. Moses Parente: if you've been obese or heavy set your entire life, what makes you think you're going to now get the discipline to go on a diet, go to the gym. work out for about a few years and then get into the and then get into like the body or sh or shape of your dreams. Or not even necessarily that, just like slim down a a decent amount. You can Niel: Yeah. Moses Parente: just instantly do it like that with GLP ones. Niel: Yeah man. it is it is certainly unique. someone I spoke to though, she did actually a thing where it's like she's doing like a more medical treatment for, you know, depression. And I forgot the exact name for it, but she was saying that it's similar to GLP ones in that say you want to lose fifty pounds or you want to lose five pounds. These GLP ones are just as effective for that process. So I do think that, you know, you might probably are seeing more like people in Hollywood. They're trying to lose that extra five, ten pounds. You can lean on GLP ones for that. As a cultural thing though, I am hesitant, right? Again, these natural protocols that exist of you know getting good sleep, nutrition, you know, nothing too hard, walking, little bit of resistance training. If you actually consistently do that, the compound effect of that will lead to a good physique. Like that is nearly kind of guaranteed there. So yeah, I do think that you know these GLP ones are a bit of a shortcut. But if you look at the state of where our nation is, right, we're frankly in a pretty dire spot and in a way to improve our healthcare situation, improve the quality of lives for a lot of people who are maybe a little bit older, it's harder for them to make that lifestyle change. I do think it serves as a great option. Moses Parente: Now that you you mentioned, I guess, going back to the point of you go to the gym, you work out, and you kind of get in shape the old-fashioned way, there's a lot of there's a lot of also mental effects that help you in terms of that. You build character, you build discipline, you develop a lot of characteristics that overall help you become a better person. You learn patience, you learn consistency, you learn how to be disciplined and to show up To even ev every single day, even when you don't feel like it, and the days that you feel like you suck and you're not progressing, but you still consistently show up, I think those are great traits that you can apply in all aspects of your life, whether it's work, relationships, friendships. And I think that while yes, the technology is great, and do I want it culturally? Not necessarily, because it removes a lot of that Like having the shortcut just kind of removes a lot of the ability to get those characteristics. It's kind of like those people who they get rich very young, but it never really works out well for them. Niel: Yeah. Yeah, and like I I remember earlier I was trying to make a parallel to AI. The difference with AI is that it empowers you to do so much more and we all can do more. Whereas with fitness, right, I think like eventually we all kind of see that there's maybe a plateau in the sense that you really lock in on doing the right habits. Within two years you should be able to get close to some of your dream physique. Then you continue to chip away and get better at it, but you're not gonna see that rapid increase like you did earlier. Yeah, and it shows that like there are simple kind of protocols that you have to follow. If you just maintain that discipline, it'll stay. With the peptides and you know the GLP ones, it is maybe like a bad cultural precedent that, you know, people are able to kind of cheat and not build the right habits and discipline. But with that being said, just the amount of impact it's able to have. Once you go down like that rabbit hole, right? I actually like know some people, you know, family friends who are on the GLP ones. It's been a big impact on their lives. And these are people in their 50s. It's hard to change those habits you've already built at that point. Once you start seeing those results, man, it changes your perspective on life. also, people have mentioned for GLP 1s, there's a sense of anahedonia where maybe you know, like you're not as impulsive anymore, which some people would say, like even traders from these financial firms, they're preventing some of their employees for taking GLP ones. they need that aggressiveness on the training floor. with these GLP Moses Parente: Mm. Niel: ones I think it kinda numbs down the feelings in general. So overall though, I think it's we've seen a lot of health impacts. And I think ultimately we'll get more adoption within the US. Moses Parente: Hmm. Okay, 'cause I it I guess because I'm still a little bit skeptical to all of it, it's just very hard. It I'm still but I I I think I'm learning to eventually accept the benefits of what these GLP ones can bring to people. But I guess like just me personally, I might still be a little bit more hesitant on people using them on a cultural scale. And going back, I guess, to what you were saying about peptides. One thing that I always this is a little bit off topic, but I thought that I remember back with a COVID vaccine, everyone was there was a lot of people who were skeptical about it, and I don't wanna be don't wanna go that route too much, but a lot of those same people who are skeptical of it are completely fine with taking peptides that are like from Mexico and China and they have all these unregulated Niel: Yeah. Moses Parente: substances and they're just injecting it and in into themselves and I'm not someone who particularly likes using needles. So I'm not the biggest fan of needles. So whenever I I see people just going down that route, they're just taking these peptides that have like that has like whatever junk they got from from like all these foreign countries that are completely unregulated. Niel: Yeah. Moses Parente: How how that's very contradictory in my in my opinion. I don't think people should be t taking any of those like crazy peptides or anything like that. Niel: Yeah, man, it's like a lot of that like liver king mentality. But Moses Parente: Yeah. Niel: yeah, I mean I don't know necessarily what to say, right? You can it's free country, you're allowed to do what you want to do. But with these peptides especially, like it's important to exercise caution here. Yes, we've seen like good results, especially for the ones that are more proven, like Ozempic, Manjaro as well. But for some of these other peptides, especially like say ones like maybe you know ipamarelin, that helps you know your hormonal profile. gives you more growth hormone. That's something that I personally was exploring, but I felt hesitant just knowing that there wasn't enough research. And the ones that were available in the market, they're a little bit too pricey. The ones that are, you know, based abroad, I just don't think I can take that risk. So being able to do your due due diligence. And if you are going to make that gamble, do it on something that you know is proven. Moses Parente: Remember back vi back a few years ago when I mean back then we saw the liver king and a lot of people they were just eating raw meat and in some cases Niel: Yeah. Moses Parente: they were eat eating raw organ meat and they were saying like yeah this is good good for us but now of course no one does that because that's obviously unsanitary and they were talking about the whole let's drink raw milk. Why are you drinking raw milk? There's nothing wrong with regular pasteurized milk. You just heat up milk. You're just you're just heating it up. To kill the bacteria, there is no I I I'd rather not have E. coli or any of those c crazy diseases that naturally just lives inside of milk, you know? Niel: Yeah. Moses Parente: Do you think a lot of this whole peptide craze is largely just a fad right now, or it's going to be something that's going to be implemented in culture? Niel: Yeah, I think it's a bit more of a nuanced question. In the sense that I do believe GLP ones, the usage is gonna increase. And you think about someone who's maybe older, who's been obese for a good majority of their lives, it really can make a positive impact. So I will say that. I do think that for the fitness community, people are seeing results with these GLP ones. But yeah, I mean, I don't think it's similar to like Liver King, in the sense that like I don't really see like any real benefit. I mean obviously like organ meat is good for you, but like why would you not cook it? Like why would you just straight up eat it raw? Like, you know, that seems like so ridiculous. there is benefits, but I think eventually that it is going to lose maybe the widespread popularity it's going right now. A lot of people are researching peptides, a lot of peptide clinics are popping up. I think it'll eventually slow down, but I don't think it'll ever go away necessarily. Moses Parente: Hmm. Awesome, awesome. And to before we close this off, do you have any final words that you wanna give give to the audience? And of course, thank you always again for coming back on Niel: Of course, yeah, yeah. one thing I mean I think like I have maybe gotten a little bit more perspective on. You know, I've done both the New York City trip one year there, back in SF now. So here building. it's been fun. I do think that it's an exciting place to be in the world, but understanding that, you know what it is that you want to do, making taking time out of your day, speaking with other people like yourself, kinda check letting what that is. And then leveraging your community. so that's something I'm definitely trying to do in SSAF right now. Don't think I did the best job of that in New York. But when you do get to be at a world-class city like New York City, leverage the community that's out there and use that as a vehicle to improve whatever it is that you want to achieve in your life and achieve the goals that you have. Moses Parente: Hmm. Awesome. Thank you for the words of wisdom, Neil. Really appreciate it. Niel: Course, thank you.