AI Layoffs, Compute Costs & Agents | Naveen Rao & Alex Finn on This Week in AI Episode 16

22:24Not uncapped.

22:25Alex Finn:No, not uncapped. Let me do that again. A research lab that involves perhaps longer timelines than a quarterly result there we go correct uh so just ballpark me am i am i gonna see something from you guys that i can that i can use or test or buy in the next three years five years three months i just don't have a good vibe for what your pace is going to be yeah i would say when we have our full-blown products uh we're probably three years away from that right now which is faster than i anticipated i had actually said it would be five years uh initially things are going faster and i do see that those will accelerate because of AI.

23:01Alex Finn:So I'm kind of pricing in that acceleration, if you will, on the timeline. So I do think we'll probably be in about three years. Now, that being said, we are building artifacts right now. So we'll be releasing a new model that works on top of what we call dynamical systems, which is a key non-Von Neumann concept. And so that's going to be released in several weeks, actually. I want to get over to Alex, but tell me, explain for people who don't know Von Neumann in the way that you're using it. I think people have mostly heard of Von Neumann probes as a sci-fi concept. moving. So just break that down for us.

23:30Alex Finn:Yeah, von Neumann is really the concept on how we build computers today, where we have kind of siloed memory and compute. And there's basically a connection between them. It's a memory interface, if you will. So basically, you have to you pull some some data from memory, you do something to it, like an arithmetic operation, and you write it back. Every computer that we know of that we built works this way. Non von Neumann means that we don't have a strict separation between compute and memory and actually dynamical systems or physical systems like your brain work in this way. They actually do both together.

24:04Alex Finn:And that's a concept that we're, we're, we're going to be exploiting. So just to double click on that, essentially my brain is full of, of neuron soup and the, the computer is the software. The software is the computer is the just there. That's right. And also the memory is the compute and the compute is the memory. All right. Alex, I want you to weigh in on what we're talking about here and kind of where you're seeing TCO as one of the leading lights of kind of the personal agent space. Interestingly enough, my first job at a startup, we had a load balancer in a closet and we were so happy we could afford a load balancer.

24:37We were very proud of ourselves. And we've gone all the way back now to look at my hardware in the closet, amazingly enough. And you're politely saying, hey, I got some serious hardware, but it's weird to me that we're talking about that as an edge in a very software focused line of technology.

24:55Alex Finn:Yeah, I, you know, it's funny. Every time I talk about hardware, local models on Twitter, I immediately get 90 % replies going, oh, those models are stupid. Why would you ever use them? Opus is so much better. Well, there's advantages to being able to have unlimited, dumber intelligence, right? It's not as smart as Opus 4.8, but it's still, you know, six months behind. It's still Sonnet 4.5 level, the local intelligence. And there's an edge to that, right? Because now you have unlimited intelligence that can run 24 seven, three 65, and that unlocks a new level of use cases. So one thing I'm doing right now is I have a Hermes agent powered by Quinn three seven on my multiple Mac studios and spark.

25:39Alex Finn:And it's constantly all day scraping Twitter, Reddit, other social media sites for opportunities. So my new startup I'm building, Henry, it's around the idea of finding autonomously opportunities online and exploiting those opportunities for the user. I'm basically running the prototype for that right now on my local models, on my computers, and they're able to go all day, find data, find information, and then ping me when it finds specific opportunities I can exploit. And this is something people don't really understand is that when you have unlimited intelligence, when you have local models running, you can unlock whole new things you can do, things that can watch your computer all day.

26:23Can we talk about the opportunities that you're unlocking here? I think that's a really interesting point that I want to better understand. You did a video recently on Hermes use cases, and I think getting from what it can do in broad to what it can do in micro is going to be useful for folks that are listening.

26:37Alex Finn:Yeah. So I am a strong believer that the reason why the sentiment is so low about AI in America is that most people are not getting value out of AI. They go on chadgbt.com, they say, tell me a fart joke, and then it tells a joke, and then you're like, oh, this is stupid, I'm never using this again, right? They just don't know how to use it. They don't know how to get value out of it. And so I'm trying to figure out how to fill that gap. So to be more specific about what I'm doing, I'm trying to build a system where you can go in, Alex, it learns about you, your skills, your interests, learns you got kids, you're an excellent father.

27:12Alex Finn:And then it goes and it finds opportunities online. Okay. Here's Alex's skillset. Here's a challenge someone's having on Reddit r slash parenting that Alex can solve. And then autonomously builds the system for you to exploit that challenge. So you to build the solution for that challenge based on your skillsets. This is only possible if you have kind of this autonomous intelligence, always watching. Right. And so I think if we built a system for people where an AI got to know them deeply, their skill sets, what they're doing, and constantly looked for opportunities to use those skill sets to create value, that could turn a lot of people onto AI in America.

27:53Alex Finn:And so that's why I think the advantage of local hardware is you can have intelligence doing things like that for you at all times. It's actually really funny that you bring this up right now because there's a company that announced funding just in the last day or two called Townie. I think Andreessen may have backed it. I don't have this pulled up, but that's close enough. And it's literally what you're describing. It's like an AI that's supposed to be able to kind of go out there and find all your personal contacts and help you. Naveen, based on what Alex just said, I'm curious, what is your personal agent setup?

28:22Because on one hand, you can just hire an assistant to go forth and do your schedule for you. Or you could spend the time setting up your own automation. So what have you chosen to do as a startup CEO today?

28:32Alex Finn:You mean in terms of like how I do my scheduling, that sort of thing? I mean, are you an open-claw guy? Are you a Hermes guy? Or are you just going to your Google Calendar yourself and looking at your day like some sort of troglodyte Neanderthal? I'm still a troglodyte. Me too. Yeah, I look at my Google Calendar. I sort of like, I have an assistant who's a human who still does this. I encourage everyone to use AI to make their job easier and faster. But yeah, I'm still a troglodyte. I use Apple Notes for everything. Um, you know, I, I have like 10 plus years of Apple notes of all of my thoughts.

29:05Alex Finn:I write them in there religiously. Um, yeah, that's Google docs for me. Like I, I, my last two jobs ago at crunch base, I got totally Google docs pilled, uh, before we even use that phrase. And now it's how I think, but here at the twist, we use notion and I've had to completely relearn how to do work. Yeah. And after being dragged by my non-existent hair, kicking and screaming for two years, I've come around to it. But yeah, it's interesting how we've become wedded to tools. Like, do you guys remember when they changed Office to have the ribbon up top versus the traditional bar? My parents had a small business and they were, it was months of complaining about that change.

29:42Alex Finn:I remember that too. It's like, why is this thing here? They had the little thing where you could disappear it also. Yeah, but I mean, like it was better. Having all those horrible dropdowns was not good. Alex, how are we going to get people like, you know, not my parents per se, but let's say people that are less AI forward that have gone to chadgpd.com, asked for a, I'm going to change your example, a sonnet about flowers and getting them to go, hey, that was not what we're talking about. When you see headlines about data centers, headlines about AI, headlines about people trying to do more with this technology, how do we get them into this world that we all think is progressing quickly and will unlock, hopefully, a lot of GDP growth?

30:20Alex Finn:Yeah, I think there's many things to it. I think you look at countries that are doing it well. One of those countries that are doing it well is China. In China, when Open Claw was released, they literally had festivals in public spaces where grandmas were lining up to get Open Claw installed. Right. But here everyone thinks it's like the worst and it's drinking all our water or something. And like that water thing. Yeah. Yeah. I'm not even going. It's so crazy. I might even go into it. But it's I think there's multiple sides. I think the biggest issue is the marketing. I think we have absolutely horrendous marketing right now.

31:01Alex Finn:And I think it's a lot of people's faults. I think one of the biggest people's faults is the CEOs that are laying off people right now. Right. In 2020, we had 0 % interest rates. I was in middle management at a public tech company, MongoDB, at that time. Not to out anyone, but the command to all middle managers at that time was hire as many people as you possibly can and pay them literally whatever they want. I'm not even, that was the command in 2020. It was a beautiful time. Well, it lasted. It was. I wish I was an IC at that point, not a middle manager. I was asked for anything. But point being is, now we're paying for those mistakes because we've had elevated interest rates for years now.

31:47Alex Finn:And what are we doing? Instead of saying, hey, we've been really irresponsible last five years, we're going, oh, well, AI efficiencies, everyone's fired, right? AI is incredible. Everyone's fired. Not that we weren't prepared for market conditions. Not that we overspent like crazy. Every CEO is putting out this canned tweet. That's like, oh yeah, AI is amazing. We have product manager shipping code. It's over, right? But we talked about flat structures, small teams. And I don't think you can have a company of 50 ,000 people if you're going to have this flat series of small teams to Devin's point about how he's building his startup.

32:21So how do we mesh this idea that to take advantage of these tools intelligently at the corporate level, we have to be super flat and not that wide with the idea that we all also agree with, that a lot of these layoffs have nothing to do with actual AI automation, but they're just using that as a false flag to wave over the top of them to avoid saying, we screwed up at MongoDB, we hired Bob for 400K a year, and Bob's worth 40.

32:43Alex Finn:Yeah. Yeah. I mean, I think there's different things to it, right? Yes, the structure of a company changes with AI implemented, but at the same time, if the company was responsible the last several years, they're not laying off 30 % of the company to make it so there's no middle managers. Why can't middle managers just become ICs? Why do we just have to fire literally every single middle manager in the company? Right. And so the point being to come back around to it is we have a marketing problem in America. People think AI is the devil because the CEOs, because the politicians that are profiting off of saying AI bad, here's a bill to ban AI.

33:22Alex Finn:Right. We have we have politicians going on X going, yeah, we have to stop all AI research. Stop it all. It's done. You know, I'm not going to mention Bernie Sanders name, but we have politicians going out there doing that. right profiting so we we just like we just have such a major marketing issue that we we have to kind of fix here in this country i was actually in dc this week uh and i was talking with people from the government i'm not gonna call out who but you know all the big all the big people that were that they would that way you think about involved in this like commerce and stuff like that sure and uh this is actually exactly what i said i said we we have a perceptional a perceptual problem.

33:59Alex Finn:And people right now are starting to protest data centers being built near them. And a lot of it, almost all of it is based on misinformation. The water thing is completely false. You know, that cancer rates are higher is completely false. I think when you have energy generation, that may be true, but with data centers, it's not. And I think this actually hits at a more fundamental part of the problem is that, so yes, you have this thing that feeds into that narrative of like, oh, CEOs are saying they're firing people. But I don't think those people are, the middle America people who are protesting data centers aren't the ones who are getting laid off.

34:39Alex Finn:So that's not where they're feeling it. I think the perception is that there's Silicon Valley that's running away and the billionaires are being created and we're not part of it. And so I actually put it back on myself and our industry a bit more to be a bit more thoughtful of how do we involve the public? Like weird things have happened. You know, my previous employer is, is guilty of this. They've stayed private for very long. I've told, I have brought this up ad nauseum with the CEO of Databricks, as I'm sure you've been told, but like, I, yeah, you care more than I do, but I'm still pissed off about it.

35:16But your, your point is that companies stay private so long, all the values extracted before they list. And then the upside is essentially precluded from the normies.

35:23Alex Finn:Exactly. And I think people are feeling that like, hey, this thing is happening. It's running away from me. All the quote unquote elites in the coasts in New York or in California are getting all the value and we're getting fucked. Right. That's basically what people are feeling. And I think the sentiment of like, I hate AI is more about that than it is about like all the things that people are talking about overtly. So what do we do? Because I talk to founders, I mean, three to four or five a week for Twist, and everyone's just excited, optimistic, full of energy. The future to them is bright.

35:59The trees are green. The sun is out. And then I go on the local Providence, Rhode Island subreddit, and people are talking about their energy bills. And they do have a boogeyman in mind, correct or not. It is precisely and exactly AI and data centers. And this is just me reading the normie room, if you will. Yeah.

36:18Alex Finn:But what's interesting is before 2020, when I was at Intel, I used to go to China quite a lot. They were my customers when I was at Intel, and I spoke to people there. And there was an energy there. People were hungry. They wanted to win. And the younger people there were super focused on AI. So I think what we see in China is a different sentiment where the normie feels that um ai is a superpower and it is a benefit to give give them competitiveness we see it as a boogeyman because somehow we've screwed up the narrative i i actually do blame the the doomers like and i specifically will call out anthropic on this like they have sown this narrative and over and over again it's going to destroy the earth and this and that it doesn't help it really doesn't freaking help they build amazing models great but stop it stop so i don't disagree with that, Naveen.

37:11But I do have a question about how you think about Anthropic. Do you think they're being sincere?

37:15Alex Finn:So, you know, I had this conversation with Jason, actually, Calacanis, last time I was on the show. And he seems to think, yes, they are sincere, that they believe that they are building, you know, the next deity or whatever. That's the way he put it. Maybe they are. From my experience, I think a lot of people there are sincere in that they think AI could destroy the world, but I'm like, why are you working on it then? Right. If you, if you seriously believe this, why do you feel righteous in what you're doing? And that's, what's weird to me. I'm like, then maybe work on a way to make it better or safer or something.

37:53Alex Finn:And they claim they are like, yeah, they would argue that they are with project last wing and, and, you know, making the world secure with mythos and so forth. But I don't see a way to fix the PR problem that AI has in the near term. And I'm really worried that it's going to become such a political issue that we're going to see this block progress on the hardware and power generation level. The two things, Nibian, that you're talking about that matter so much. And we're going to end up with Alex talking to a smaller room of people about how to get going with AI, how to build cool things because everyone's going, absolutely not, I don't want that.

38:24Get that AI slob out of my face. And it could retard our own growth, our own progress. I mean, it's just, it's an overall net negative, but I don't have a pitch. I don't have a solution or anything to bring to the table. So I don't want to complain too much, but we got to do something.

38:37Alex Finn:Well, one of three things are going to happen. One of three things will happen. Either there's the best case scenario, which is the leadership decides, hey, we're putting this country in the wrong path. We're going to have chaos and lose to China if we continue on this path and telling everyone AI is the devil. Let's we don't need to exactly work together, but let's like make our messaging what the truth is, which is this is an incredible technology that empowers everyone. That's number one. That'd be the best case scenario. Then there's the medium case scenario, which is there's chaos in the streets and people get angry in these protests around data centers drinking water.

39:12Alex Finn:It just gets to the point where the hands are forced by these companies or they'll keep getting rocks thrown through their windows. That's number two. And then there's the worst case scenario of all these, which is the Bernie Sanders of the world win and are able to legislate. These people have no idea how AI works or really capitalism or the economy works. They end up winning because people are so upset and they end up legislating this stuff, which kills our economy. And we have to give 50 percent of every A.I. company to the government, which would be the stupidest thing to ever happen in history.

39:42Alex Finn:Those are the three ways. All right. All right. I want to I didn't mean to get political. That's not political. That's just like, you know, guys, I'm going to get political. I got to go political. Naveen didn't have to. He used to go to D.C. He went to China. Now he has to go to D.C. to figure out what the hell is going on. This is inherently political. Let's spend one minute on this and then we'll get back to fun technology whiz bangs for the audience. But the Bernie Sanders idea of putting a chunk of equity from these companies into the national trust, to me, is not that far off from the government buying a stake in all the quantum companies, backing individual foundry firms.

40:14Agreed. So I feel like, ironically, the theoretical party of free market has already taken us into state capitalism as an acceptable way to approach industrial policy in the AI era. So why are we so opposed to Bernie's idea? Other than that, 50 % is way too high. But, you know, I mean, I don't know. I wouldn't mind if 5 % of Naveen's company and Alex's company were both in the National Trust. And that could save Social Security as birth rates fall. Maybe we need something radical to solve this.

40:42Alex Finn:So I actually don't disagree. I don't think it's a crazy idea. I think, like you said, 50 % is crazy high. And you can't destroy the incentives we have right now. That's the key. We cannot do that. But we do need to involve the public somehow in the upside. We need to show the public that AI is a win, is a competitive advantage, and somehow let them share in the benefit. it. Yeah, that has to happen. And the problem with Bernie's thing is that I think you go and own it. And let's say, you know, the value goes up and basically gives you more revenue to the government and then it feeds into bureaucrats who do nothing with it.

41:21Alex Finn:And that sucks, right? You're not you're not putting money where it's going to be used to actually drive stuff forward. I think that's what I have a problem with. But I do almost feel like, hey, if there's some way to, you know, provide training for people to use AI more effectively. To Alex Finn's point, somehow we use that money in a way that continues to compound the advantage. Just one point I'll make here is the reason why taking 50 % or 25 % or even 1 % of every company and putting into some trust, seizing private property is fundamentally un-American. We should never have policy where seizing private property is the fix to things.

42:01Alex Finn:So there just has to be solutions other than seizing private property. The moment you start seizing private property, you're messing with all of the incentives, which fundamentally just changes how the country works, right? So as Naveen just said, instead of seizing private property, why don't we incentivize people getting value out of AI? Why don't we give everyone in the country a$20 Chad GPT plan, Plus, we have an initiative where we spin up educational services where everyone gets taught how to get value out of AI. That's how everyone gets value out of it. Why is everyone getting value out of it?

42:39Alex Finn:It's us just going around taking things from people and redistributing it. That doesn't make any sense. Well, and actually the contrast, the trust idea versus like the Lutnik idea. And I'm not necessarily saying the Lutnik idea is great. It's weird. It is state ownership. Spell it out for us if you don't mind. Well, basically, he already did this with Intel. He bought 10 % of Intel with the government. The government owns 10 % of Intel. Strange, right? But the way he's looking at it is like, all right, well, if we figure out policies that make Intel more successful, that 10 % goes up in value.

43:14Alex Finn:Now the government can invest more money into other companies and compound it. It's kind of like being a VC, right? I've met people who do that job for a living, but this takes me back to Microsoft. I used to be a Microsoft beat reporter back in the day. And I would always press them. Why don't you guys have a venture fund? Look at what Google is doing. Look at what these other companies are doing. And they always would kind of look down their nose and be like, well, you know, it's not really accretive to our balance sheet, which is a polite way of saying we're too rich for a poor man's games like VC investing.

43:42But it provided signal later on. So they got into it in 12 and so forth. And so maybe, maybe, but I'm not convinced that that's something that we could get everyone to cheer behind. We have to find something that doesn't just appeal to us, but that someone on Main Street actually goes, hey, they're doing something for me. They're giving something in my domain. A government stake in Intel is a highly technical, well, for the average person, a highly technical transaction they don't really understand. I thought it was a public company. What does public ownership mean? That's the level they're at.

44:15Alex Finn:Also, it's not clear what the government, even though it appears their investment was a really good one. And I don't think it's clear to 99.9 % of Americans how they benefit from that. Right. The average Americans. Oh, Trump, you, you, you know, you have a five X, your Intel investment, but they just don't see the connection between that and their life getting better. But they can see their life getting like that might be true as well. Let's be honest. Intel going up in value since since the U.S. bought a chunk of it has zero impact on my life. It doesn't even really impact that month's increase in the national debt.

44:49It's it's it's small potatoes. Welcome potatoes, but small potatoes.

44:54Alex Finn:It's like a corporate VC kind of in the government. But I think, like, let's look at examples where at least perceptually there's a technology that helps people. You know, like people, for whatever reason, have gotten behind like auto manufacturing, right? They're like, oh, I can see it because there are jobs created and they know what a car is, right? And somehow they see it as like, well, you know, when there's government intervention here, I benefit from it. So we kind of need to come up with a narrative where we show that, right? We show that the everyday person is getting a benefit. Their cost of something is coming down.

45:28Alex Finn:Their access to something is going up. Like we need something tangible because right now we're all living in our Silicon Valley bubble. Every one of us, myself included. And talking about agents and this and that, the person in Nebraska has no freaking idea what an agent is, nor do they care. It doesn't impact their life. I mean, I hope Alex gets them to care. Yes, I hope you do. I'm working on it. And shout out to you. I think you're actually, no BS. I think you're actually a very important AI educator and evangelist. And I think it does matter. But to Naveen's point, the guy in Nebraska cares about the Cornhuskers football record and the price of fertilizer, neither of which are currently being assisted by AI.

46:07That's right. Yeah.

46:08Alex Finn:So we need to at least start to show like, hey, fertilizer came down 20 % over the last year. You know, this is why. like create those narratives right and and we do this i think then we'll have to me this is this is incredibly important to national security because otherwise we lose this race right if the if the if the public is not supportive of ai no matter how much silicon valley cares it's it's not going to matter that's gonna be what about so another idea that i've seen people talk about and by the way i didn't mean for us to end up so much on the how to save the public conversation about ai but i think also all the startups founders that are watching this care about this so we'll stay with it for for one more question.

46:46What about when you build a data center in a community, let's say it's near your house and it does plug into your grid and does use some of your water. I'm not gonna say that very loudly, but what if there was something that came to a voluntary tax by companies that are building these data centers or operating them or getting value from them to directly pay into that municipality, county's coffers, and then people could see directly in their community that bench, that bench came from AI because we put a data center in us And then people could compete to have access to data centers in their community.

47:16It could become a net positive versus a, my electricity bills are going up. And honestly, I don't think it would be that much of a cost compared to the, the CapEx and so forth of building at one gigawatt data center, which is like, what, 50 today?

47:31Alex Finn:Something like that. Yeah. Yeah. I don't know, Alex, what do you think about my, my brilliant voluntary tax idea? It depends how it's executed. I just my gut fundamentally, as opposed to any sort of seizing or taxation, because I, you know, and again, not to get political, I just don't believe increasing taxes all of a sudden creates things. I think if there was a very specific system put in place where the money taken from the data center with no middlemen taking a dollar out of that goes into building something productive, right? What if you built a data center, tax it at a small percent, and then literally 100 percent of that money went to giving AI accounts to everyone in that area?

48:11Alex Finn:I think that is not bad. The issue is our system overall in this country, Republican, Democrat, doesn't matter. when you take money from one place and put it somewhere else, somehow along the way, 80 % of that money disappears and we don't know where it went. Right now, as long as that's the way, I just don't think taxation is the solution. But if we had a system in place where it was like very clear, this money goes here and every single one of those dollars is back in helping the people, then I think that's pretty good. This actually goes to an old school model, right? If you look at old companies, like old automotive companies, or pharmaceutical companies, any manufacturing company in a local community, they did exactly this.

48:48Alex Finn:They improved the community and they did it voluntarily because they know they learned through iteration that the backlash is worse. And I think we in Silicon Valley haven't done that. And I actually blame SaaS inadvertently because I think SaaS is just sort of divorced from the physical world, right? Very much so. You don't really have to interact with the physical world, but now it's coming back hard. We're interacting with the physical world in real ways. and you can't go in and just be like the guy that's going to be the wrecking ball that kind of puts in a data center and screw all you guys, whatever you think.

49:21Alex Finn:You got to think holistically. Like, all right, well, yeah, we're putting the data center in, but you know what? We're also building a better rec center. We're improving road access. We're doing things that the public will feel. Yeah, if I was doing this, if I was going to bring a data center near Providence, I would be like, I'm going to pay for the buses for five years. Just like buses are free now in Providence. And you know why? know why? Because your community supports AI and AI supports you. It's a good sales pitch. It totally is. Okay. Bringing us to a conclusion, I want to end that somehow we ended up in kind of a bummer there for a bit.

49:52I think we're all a little worried about politics and the backlash. So let's have a little fun to wrap up. I'm so glad I have you both here because I think we'll all have different views about this. But Anthropic, finally, filed to go public. Everyone's very excited about this. It's a private IPO filing, so we don't have the numbers yet. We're still waiting. But the fact that they've gotten to that point, I think, shows quite a lot of maturity and an imminent IPO, maybe Q3 this year. So Alex, if you had to guess, what is the IPO valuation for Anthropic? And are you a buyer at that price?

50:21Alex Finn:It feels like it's probably going to land somewhere around 1.2 to 1.5. Am I a buyer? I'm actually not going to be a buyer of any of these major IPOs coming up, OpenAI, Anthropic, or SpaceX. I'm a big believer in all three companies. I'm a big believer in all their leaders. I'm a big believer in all their tech. I think they are cornerstones to the country and our economy and will be for a very long time. I just can't remember a time where there was an extremely hyped IPO. And then a few months later, the price was higher than where it was when it came out. And there's also some interesting things happening on where rules are being bent for these companies as well.

50:57Alex Finn:SpaceX is going into like the S &P and the Nasdaq on like day one, when typically it takes like a year to do that. So there's just too much weird things, too much hype, too much weird things. I'd rather let price discovery happen and then make up my mind from there. It just seems like it typically when anything is money based, you want to zig when other people zag. That's just kind of how it works. And right now it feels like 99 % of people are zigging. And so I'm just going to kind of play it safe and zag. Naveen, same question to you being a brat, guess of price and also interest in purchasing.

51:33Alex Finn:Yeah, I think Anthropics, yeah, around the 1.2 range. I mean, okay, positives. I'll answer your question in a second, I promise. Positives are we do involve the public and the retail investor, which is great. I think SpaceX has bent some rules. Some of it has been around trying to involve the public more. So they actually want to bring thresholds down for people who can get into the IPO price, stuff like that, which I think are all good things. I think retail investors tend to distort value and price discovery quite a lot because they're not super analytical. Yeah. Yeah. So I don't know if it's a good thing or a bad thing, but at least we involve the public, which is sort of the intent of the market.

52:15Alex Finn:Whether I'll buy or not, I mean, I don't know. I mean, I agree with Alex that the leaders are all, I believe in all of them. Elon is hard to bet against, as we've all seen. But man, the market, the actual revenue of SpaceX just doesn't feel that it justifies the valuation. If this was like$500 billion IPO, 100%, I'd be there in a heartbeat. But at 1.75, I'm like, oh, crap. I saw this one as well. This is pretty funny. The Elon magic multiple of 10. That feels too high to me. I probably will not buy into that one. and I probably am wrong. You know, but... This is why all of my money is in zero cost index funds because no offense to the world, I've done a lot of fantasy stock trading and let me tell you, I'm worse than the masses.

53:11So I'm fine with that. Guys, a real trade. I want to give you some shout outs before we go. One, Naveen, unconventional AI is unconf.ai. Anything else you want to shout out? Roles you're looking to fill? Things you want to stress before we go?

53:24Alex Finn:Yeah, I mean, people who are interested in nonlinear dynamics and how that works with AI. Absolutely, we're looking for you. People who are interested in new kinds of hardware circuits, weird, wacky stuff, physical design, any of that stuff, come talk to us, even like modeling these kinds of things. So yeah, we have some really interesting roles going on right now. Awesome. And then Alex, meethenry.ai is where Henry Intelligent Machines PVC lives. Anything else you want to shout out? No, if you are looking to get into AI agents, which I really think you should, check out my YouTube. but meethenry.ai is my next startup I'm building, building an AI agent for everyone to create value.

54:00Alex Finn:So check it out. Hell yeah. I appreciate y 'all. This has been This Week in AI. We're back with more awesome founders from the world of AI here on your podcast feed. My name is Alex. We'll see you next time. Bye.