Raoul Pal: The Journey Man
00:36:56 · Micky Malka, Raoul Pal
[00:36:56] S1: Thank you. I hope we're... We are right on it. [00:36:59] S0: I don't see I don't see how not. I don't see... I honestly don't see how not. Because we're seeing marketplaces already for some of this stuff. I know we've written about it at Global Macro Investor that this seems to be the way for everything, that everything gets tokenized both in a monetary token attached to an information packet token, because it's the only way that we can allow agents to operate in an economy and collect the information it needs to do its stuff. [00:37:26] S1: I... Do you wanna start a company around it? We're happy to talk to you on it because you're spot on. And I think that's where the agents are gonna be amazing because in Web two point it was really... It's still really hard to organize your personal data to know how to get a benefit out of it. But imagine if you had your own agent that can actually do that for you. And it organizes your your data to figure out what's the right... How how to create the tokens that are needed for somebody else to... For you to exchange value. And I think that's where it's gonna get super interesting. [00:37:59] S0: And how do you think about agents... Autonomous agents as investable entities? Are you starting to think, well, there's clearly just going to be a RIA or wealth adviser agent, and these are not necessarily companies per se. How are you thinking about that transition?
00:38:20 · Micky Malka
[00:38:20] S1: I think that I was I was throwing a controversial statement a couple nights ago talking about, you know, I got three kids They're in... One of them is in college. The other two are gonna be in college in the next few years. How hard is is the labor force gonna be for them when they graduate? And I I I had three hardcore statements. The first one is, in this era, you only go to college if you either get a scholarship and you end up... You can graduate debt free or you have some trust or family that can afford it or grandparents that can help you afford it. But you should not go to college if you're going finish with debt right now. The second big statement was a bunch of people are going to have a hard time finding jobs, but we... I think there'll be the entry workforce may be that we work for agents. An agent will hire young people to do tasks, which is very different than a company hiring them. And I think that's that's coming sooner than we expect. So I think that's how agents are gonna agents are gonna enter the economy. It will take a while. We're not there. The agents are very buggy. The user interfaces are not great. I don't find the phone is the right element to connect with agents. The phone is quite clumsy to do that right now. I find... I don't know you, but I find myself using more AI on my laptop than I am on my phone. Yeah. I can I can put more features, more applications? I can have more access to to ins... To enhance the experience of my hard drive, my voice, my notes, my video than I can on my phone. So I think we're still missing a new interface. Maybe it's the new glasses. Maybe it's something around the glasses. I don't know yet. But, we need all of us to make these agents really take off the way they should.
00:40:31 · Raoul Pal, Micky Malka
[00:40:31] S0: It was interesting. Fiverr, the platform that, you know, you can hire people to do stuff, made an announcement they're cutting, like, 30% of their workforce because of AI. I'm like Yeah. Why don't they just become an agent platform? Yeah. And I think of agents as bundled tokens because they they use different tokens for different outputs, and they're like, those baskets of of tokens being used for different applications feels like there's a Fiverr or something like that could completely take over a marketplace because that's all it does anyway. [00:41:05] S1: A 100%. I agree with you. It's gonna take them a little bit of time. It's it's always the innovator's dilemma. And what... If if they don't do it, there'll be a 21, 22 year old kid who, think about it this way, there's already two full grades of college students, almost three, that since they had to apply for college, ChatGPT was around to help them write the essays. And in a few years, no one's gonna know how it was before that. And that generation is gonna build applications and agentic experiences because they're building agents at school already for their classes, and it's gonna be much more native. So I'm very bullish on that generation coming up. [00:41:48] S0: Yeah. You said this to me the other day when we were together. You said I'm spending all of my time with 21, 22, 23 year olds because they're native to this, and they're the only people who are gonna get it.
00:41:48 · Raoul Pal, Micky Malka
[00:41:48] S0: Yeah. You said this to me the other day when we were together. You said I'm spending all of my time with 21, 22, 23 year olds because they're native to this, and they're the only people who are gonna get it. [00:42:01] S1: Yes. It is like going back to 1996, '95, 9798, and trying to talk to somebody about the Internet that is not actually playing with it twenty four seven. What could they tell you? How would they describe a page view or a HTML website or a browser or a SMTP? Like, they they they had no con... It's... We could try it, but it was not the same as the guys that were building We were lucky as a generation that we grew up with the infrastructure of the Internet showing, beginning built. So we understood what was possible. We understood that, hey. The packets on TCP IP can actually be compressed more and you can send video. And you will understand, yeah, I can see that happening. I can see how packet because I understand how TCP IP works. So you can understand, yeah, voice, video, yes, it's cool, but it's it's possible and and better and better. These kids are the same for AI. They understand foundational models. They understand the application layer. They are they are they are... They don't take notes on anything but on an AI agent. They do the homework there. They're just different. It's the same way as the kids of the nineties were.
00:42:01 · Micky Malka, Raoul Pal
[00:42:01] S1: Yes. It is like going back to 1996, '95, 9798, and trying to talk to somebody about the Internet that is not actually playing with it twenty four seven. What could they tell you? How would they describe a page view or a HTML website or a browser or a SMTP? Like, they they they had no con... It's... We could try it, but it was not the same as the guys that were building We were lucky as a generation that we grew up with the infrastructure of the Internet showing, beginning built. So we understood what was possible. We understood that, hey. The packets on TCP IP can actually be compressed more and you can send video. And you will understand, yeah, I can see that happening. I can see how packet because I understand how TCP IP works. So you can understand, yeah, voice, video, yes, it's cool, but it's it's possible and and better and better. These kids are the same for AI. They understand foundational models. They understand the application layer. They are they are they are... They don't take notes on anything but on an AI agent. They do the homework there. They're just different. It's the same way as the kids of the nineties were. [00:43:12] S0: It's fascinating. And do you think the businesses they build will start falling in the category of one employee style... You know, you can build a billion dollar business with one employee idea. Are we getting there yet?
00:43:12 · Raoul Pal, Micky Malka
[00:43:12] S0: It's fascinating. And do you think the businesses they build will start falling in the category of one employee style... You know, you can build a billion dollar business with one employee idea. Are we getting there yet? [00:43:24] S1: I think we're on our way faster than we think. I think, using just in historic context, you know, when, a hundred years ago, a hundred and ten years ago, the the vanguard of entrepreneurship in America, the barons, when they... When their companies went public or when they got really, really big, they still own 70, 80% of their companies. And their friends own 10%, and their merchant banks own 10%. And capitalism and society moved that equation from a founder owning 80 to the founder owning 20 and the financial investors owning 80. I think we're gonna start to see the move back where companies have less employees, hence, less dilution. They're more capital efficient because of all these agentic experiences, and the founders will probably own more of their companies. So we'll see a proliferation of new companies. [00:44:17] S0: That's fascinating. We've seen crypto go the other way that they tried to force as much out of the hands of of the original founders, and it loses alignment of interest.