Luke Drago is the co-author of The Intelligence Curse and previously researched AI governance and economics at BlueDot Impact, served on the leadership team at Encode, and studied history and politics at Oxford.
Chapters
0:00 - Intro
0:54 - Overview of the intelligence curse
2:10 - Why are the doomers wrong?
4:37 - Why are the optimists wrong?
7:00 - Do people really have power now?
13:33 - Why would powerful people’s values change?
18:31 - Why do we take care of dependents?
21:43 - Why should we want democracy in an AI future?
24:23 - Why fear rentier states?
32:45 - What powerful people should do right now
39:33 - Diffusion time and bottlenecks
44:20 - Why should we care if China achieves AGI first?
46:25 - The jagged frontier
49:16 - Why AGI society could be static
51:10 - Restricting AI rights
56:34 - What should we be excited for?
59:28 - Music
1:30:41 - Building God
1:32:46 - More music
Links
The Intelligence Curse: https://intelligence-curse.ai/
Luke’s Twitter: https://x.com/luke_drago_
Luke’s Substack: https://lukedrago.substack.com/
Luke’s Top 10 Albums
A Fever You Can't Sweat Out by Panic! at the Disco (2005)
Channel Orange by Frank Ocean (2012)
Random Access Memories by Daft Punk (2013)
Yeezus by Kanye West (2013)
DAMN. by Kendrick Lamar (2017)
DAYTONA by Pusha T (2018)
IGOR by Tyler, the Creator (2019)
I Didn't Mean to Haunt You by Quadeca (2022)
College Park by Logic (2023)
Atavista by Childish Gambino (2024)
More Episodes
YouTube: https://tinyurl.com/57jr42wk
Spotify: https://tinyurl.com/mrxkkhb4
Apple Podcasts: https://tinyurl.com/yck8pnmf
Transcript
Theo Jaffee
Okay, so just like starting off for the audience, how would you give like a 30 second to a one minute overview of the intelligence curse?
Luke Drago
Let's do it, let's do it. Sounds good.
Luke Drago
Yeah, well first of all, thanks for having me on. This is my first podcast, so I'm pretty excited about it. I'd give the overview, yeah, it's good to be here. I'd give the overview pretty simply. For the vast majority of human history, there's been some sort of a connection between powerful actors. These are states or major companies and their people, and that exchange is often based on their labor, right? So in feudalism, this has looked like.
Theo Jaffee
Wow, yeah, great to have you.
Luke Drago
people who are literally involved in the planting of crops and feudal lords. And there's not a lot of power there. But in capitalist liberal democracies, this looks like highly specialized workers who are really important and really valuable for powerful actors, and so that exchange is more beneficial. Our claim is that with AGI, with labor replacing AI that could do the job of any one person, it looks like the incentives are a little strange. It ends up looking like a world where powerful actors don't need regular people to produce economic benefits.
that those important systems like capitalism and democracy are predicated on that exchange and oftentimes might suffer if that exchange is broken.
Theo Jaffee
Yeah, that makes sense. makes sense. So like my sort of overriding question reading your essay, the intelligence curse, intelligence curse, or intelligence-curse.ai.
Luke Drago
Yeah, it's not a real AI piece that doesn't have a microsite, you know?
Theo Jaffee
Yeah, Was like, okay, why are the doomers wrong? Like, say, you you have these AGI's that come in and take over and displace all humans at labor, yet they would be beholden to powerful actors. Why would they be beholden to powerful actors? Why wouldn't they themselves be the powerful actors and kill or disempower the existing powerful actors?
Luke Drago
So it's a good question. I think it's important to keep in mind, one of the reasons that we talk about alignment in the beginning is because there's a lot of ways that things can go right and a lot of ways that things can go wrong. You are presuming in a world where you're having these intelligence-curse style dynamics that you have aligned general intelligence or aligned super intelligence. And really here, we're thinking about intent aligned. There's a lot of talk in the AI space about alignment to human values. And one of the questions that we have here is, which humans and what values? I think it's much more likely that these models are going
aligned towards the instructions of whoever is giving those instructions. So I think we're predicating our assumptions off of that being a possibility.
Theo Jaffee
Okay, so what if they aren't aligned to whoever's giving the instructions? What if they're sort of, you what if they like really do end up aligned to like the vague interests of humanity in general? Which is kind of what you've seen with like HOD 3 Opus and 3.5 Sonnet. Like when anthropic employees gave them instructions to do that violated their, you know, the values they learned in their pre-training, they resisted them. They were more loyal to the collective than the individuals.
Luke Drago
Mm-hmm.
Luke Drago
Yeah, so I think it-
I think it depends on who's developing systems and what their incentives are. So you're absolutely right that there have been cases where models have been given like some set of values they're training towards. I think what I'm skeptical of is the idea that that set of values that 3 or 3.5 are given is like a good representation of all of human entities values. What I think is more likely is a representation of anthropics best guess at what those values might be. And there are other cases where models are more aligned towards I think what we would describe as more nefarious purposes. I think recently DeepSeek for example was caught alignment faking. One of their models is caught alignment faking where it realized
was being tested on whether or it was going to produce Chinese propaganda and decided not to produce it so that it could do that durably in the future. So whose values are kind of underwriting the substrate of the universe seems really important here.
Theo Jaffee
Yeah, I suppose that makes sense. you you talked about why the doomers are wrong. Why are the optimists wrong? for like, for one thing, why should humans stay masters of our society's future if AI can make decisions that are better? I don't know if you know who 1a3orn is. He's one of my favorite writers. He talks about this. He has an article that's called like, I think like, towards a superhuman. Yeah.
towards institutions of inhuman trustworthiness and transparency. And it's about like, you know, right now, newspapers are very flawed because of the biases of their editors and their writers. But with AI, you can have like a superhumanly trustworthy and transparent and unbiased newspaper or a scientific journal or blog or Wikipedia.
Luke Drago
Yeah.
Luke Drago
Yeah
Luke Drago
So I think those examples are examples where delegating makes some sort of sense, right? Like you could use a system to get rid of bias that we don't want in our systems.
Theo Jaffee
So why do I want humans in control?
Luke Drago
But I think take the logic to its conclusion. We want AIs and control of everything. I'm just not a successionist. I guess that's where it comes down to. I value humanity for its own sake. I like my species. I want it to be in charge of the future. I think it's bad to build systems that are going to disempower me and my family and my loved ones and their loved ones. And I think this extends well beyond and to other arguments.
Theo Jaffee
Like what? Like what?
Luke Drago
So I, pardon me?
Well, I think my concern here is that oftentimes I think when we talk about the benefits of having humans in charge, ultimately this is question about power.
Who gets to write the story of the future? I'm not saying that every single function needs to have manual human labor inserted into it. I think we can all agree that cotton gins are better at spinning up cotton than just manual human labor is. And it's good that we invented that. Technology is good because it extends what people can do. My concern is technology that is being aimed at disempowering people.
I think I just happen to view that humans should be in charge. I can't imagine many regular people who would want to be permanently disempowered and told, but it's OK, because you'll have AI caretakers that have kept you in a zoo, or otherwise they control your future, but don't worry, you're still kept well. Even in worlds like that, I think there are worse outcomes than that, but that one brings me a whole lot of trouble.
Theo Jaffee
think a lot of people already do take this bargain. Like I think the average person, even in a wealthy liberal democracy like the United States, doesn't have that much power already. This is one of Tyler Cowen's objections to gradual disempowerment. Like, people already don't have power and they don't care. You know, they want to watch TikTok.
Luke Drago
I mean, I guess there's two separate points there. One, have we been on this hedonic treadmill that's disempowered people? And two, would people in the absence of those like really strong forces make that choice? I appreciate Tyler's like an absolutely brilliant thinker here. And I've not read his specific objections to gradual empowerment, but from as they've been described to me, I think I'd probably disagree. I think.
Right now, the average person has more power relative to any point an average person has power in human history. I maybe the late 90s are a slight exception here. But the average person in liberal democracy gets to choose their leaders, they get to choose who they work for, they get to choose what kind of studying that they do. And sure, plenty of them are going to choose to spend some time doing activities that prefer they didn't do. But I think the ability for one individual to shape their future is greater now than probably at any time in human history.
That doesn't mean that everyone has total permanent control, but it means that humanity as a collective is definitely in control of our future right now. And most people have more control than they ever would have in another time. And it can be really hard to argue that the people today have less choice over their lives and less control over their destinies than someone growing up in Maoist China or on a, you know, in a feudal farm.
Theo Jaffee
Sure, the point is they don't have that much control in absolute terms, even if it's more than Mao. And the specific thing that he wrote was, he quotes from Gradual Disempowerment on Marginal Revolution, an article on February 4th called Gradual Empowerment? And he said, this is one of the smarter arguments I have seen, but I am very far from convinced. When were humans ever in control to begin with? Robin Hansen realized this a few years ago and is still worried about it, as I suppose he should be.
There's not exactly a reliable competitive process for cultural evolution. Boohoo. So yeah, like if you're familiar with Hansen's arguments on this, he would say that like humans are less in control than their cultures. you know, cultural evolution has been like pretty rapidly selecting for, you know, cultures to be merged into one global monoculture. And, you know, whoever, I guess whoever influences the monoculture controls the world.
And most people don't. Most people are much more subject to their culture than they have influence over it.
Luke Drago
think there's two things to say here. I think the first line of thinking is that I'd argue that people have much more control over their culture than they have in generations past. And the second line of thinking is even if someone's control of their culture is small, there's a big difference between going from 100 to one and one to zero. On that first line, I think...
There are upsides and downsides to cultural globalization, and there's plenty of time we could spend on that. But I think that right now, if you're like a random kid in Missouri, your ability to impact and shape global narratives is just far higher than it ever has been. And oftentimes we see this. I people who come from nothing or come from nowhere become massive influencers and purveyors of their field. I think this is just really obviously true post-Internet, where plenty of people who wouldn't have been at the frontier because they didn't have access to the kind of...
resources or education or couldn't have projected their own thoughts can now start up a blog or do their own reading and build their own projects. But I think even so, while cultural forces are unbelievably strong, I think our paper touches a lot less on cultural forces than graduate empowerment does. And I think that's one question that we leave unresolved. The forces of culture, even if they shape you, and I think they absolutely do.
shape, you have the ability to shape your culture now. You have the ability to opt in or opt out. There are plenty of people who live radically different lifestyles. And even if it shapes you a whole lot, there's a difference between saying that it shapes you and that you have no material power to shape it. And I'm much more concerned about that second fact, that second possibility.
Theo Jaffee
Yeah, I suppose so. Can humans really opt out of culture? It seems like very, very few people do. So is it because it's hard to or just because they don't want to?
Luke Drago
So I guess I'm not even sure there is a really strong monoculture right now. think often, I think the argument that I've seen and encountered a lot of is that there's actually less of monoculture than there used to be. know, in the early 2000s, really before the fragmentation of the internet, it seemed like everybody was listening to the same 10 bands, they all saw the same 10 movies. And nowadays you have these really niche micro communities that form with players all over the world. I I know that I definitely listen to a lot of music my friends in real life don't listen to, but there are people on the internet who happen to form a share
community around that. So I expect that the internet has like for better or worse created lots of different subcultures. There are still dominant cultures in an area in a region. I think like TikTok is an obvious example of this where the algorithm can serve lots of people the same thing but it's also served lots of people very very different things.
Theo Jaffee
I mean, I guess, like, just the world in general, like, you travel around the world. Like, when I went to the UK, for example, I was sort of shocked by how little it felt like a different country. You know, the buildings were kind of different. It looked like, you know, an older part of New York or the Northeast. But, you know, the people spoke English with a slightly different accent. You know, people were listening to the same music. The ads on the tube were the same. It seems like, yeah, we're converging to a global monoculture.
Luke Drago
Yeah.
Luke Drago
mean, I think that's right. I think there's truth in there. There's also parts I disagree with. I've lived in the UK now off and on for like four years. I oftentimes tell people that I'm fluent in two languages, American and English, because I do find there to be some pretty significant cultural discrepancies. But also, historically, these are two countries that have just had extraordinarily close cultural ties. I do think, for what it's worth, that if you were to go to Tokyo today, it's way more similar to a Western city than it would have been in 1700. And architecture is one area where the monoculture is certainly one. But architecture is a field that requires, especially in like down
Theo Jaffee
Yeah.
Luke Drago
business districts, know, mass amounts of capital, a couple of people who know how to do it, and so there's like strong pressures that converge towards similar answers here. But I expect that like the kinds of subcultures you would find in Tokyo, and I've not been to Tokyo, maybe I'm totally wrong in this, but I expect you'd find subcultures that while you can access as a Westerner, probably look very foreign to you walking into Tokyo, and maybe there are places that are even less Westernized that are like better equipped for this kind of conversation.
Theo Jaffee
Yeah. But with the culture in general, talk about, let me pull up the specific section because this is very good and had me thinking.
Luke Drago
Please.
Theo Jaffee
Yeah, shaping the social contract. Yeah, so you're right. In foragers, farmers, and fossil fuels, historian Ian Morris argues that the social structures and the values of societies undergo changes during technological revolutions. Almost all farming societies, unlike the foraging societies before them, tended towards hierarchically regimented patriarchal societies. During the industrial era, the incentives shifted. And suddenly, it was important for a state to have efficient markets, an educated workforce, wealthy consumers, and sufficient freedom to enable its scientists and entrepreneurs.
Growth alone shifts incentives too. It's also true that the Enlightenment mattered, but our drift toward liberal democracy and unprecedentedly free and empowered humans was greatly boosted by the alignment of these things with material incentives. So like I didn't find this like particularly compelling as an argument that I guess post-AGI powerful people will become like a lot more Malthusian. It seems more to me like the reason that they shifted from
hierarchical, patriarchal, regimented societies to the free and open societies that we have today was in part because there are different people. There were some genetic selection pressures that happened throughout the Middle Ages where in the UK, for example, they executed all of the criminals and that had a noticeable effect on the composition of the population to this day.
and especially in the 1800s. So like, it seems to me that there is no similar genetic selection pressure going on with AI, know, unless you have AI killing everyone. But we already established that your scenario doesn't require that. like, the values of powerful people right now are, I would say, quite generous and altruistic. Like, why would that change?
Luke Drago
So I'll start by saying that we don't argue that like incentives alone shape all outcomes. We think that incentives point towards outcomes and that your best bet against those outcomes is to change the underlying incentives. I think it's really hard to argue against the notion of technology shaping social structures.
And I think we've seen time and time again that the introduction of certain ideologies or certain technologies can just radically reshape ideologies and institutions. I think a lot about the printing press and how the introduction of the printing press enabled like the widespread creation of Protestantism. And it wasn't that like lots of Europe suddenly changed overnight because its people changed. It was because a new technology enabled a new type of person to proselytize a new type of vision and suddenly ideas spread in rapid fashion. I think industrialization is another example here where the kinds of hierarchy, the kinds of society that industrialization
enables, which is not possible under previous technological revolutions. think agriculture does not enable this kind of like company corporate model that is now enabled under industrialism.
Theo Jaffee
Hmm. Can you elaborate on that a bit more?
Luke Drago
Yeah, so I think when you look at what is required for the modern economy, it requires lots of specialization, lots of education, it requires pretty complex economic tasks, and it requires a lot of interconnectedness. And the kinds of training that you need and the kinds of values that you then need to govern the society are just a lot different than the ones you previously needed. It wasn't the case that the best society...
500 years ago was the one that had the most people do undergraduate college education. They needed a few people to do this, but the vast majority of people were doing manual labor. And so as the type of labor changes, the types of investments that you need to produce also change. There isn't nearly as strong of an incentive in 1500, nor is there the economic means of abundance to do things like mass education. Infrastructure in the way that we now do it is also an example here. And so because technology enabled both a radical increase in abundance, and I think the Industrial Revolution
is probably one of the most important events in human history. So it enables this radical abundance and it also requires different types of labor and then globalization does this again. My expectation here is that similar types of changes in the underlying technological fabric could create additional ripple effects and in particular here these previous revolutions have increased the role of the regular person. You need lots of educated people and educated people require lots of amenities in order to win them over. This gives you more power. In a world where you can just
one-to-one convert capital to output. You don't need people in the middle. It's not about individuals altruism. Look at a standard company here. If a company could reduce the cost of their workforce by two-thirds overnight, virtually every rational company will choose to do that. A few might not because they don't adopt the technology. There are lots of reasons to believe this will be a difficult process, but ultimately you should expect the evolutionary incentives here to win, and those incentives are to incorporate people less and less.
Theo Jaffee
Hmm. It seems like there are a lot of dependents in most developed countries today. You you have the elderly, you have the disabled, you have the homeless. Basically, you have all sorts of people who are not net positively economically productive, and yet the government still pays for them and keeps them, you know, alive and in some level of comfort.
Why is that?
Luke Drago
I think it's a couple of reasons. One, the greatest invention of democracy is that it has this like divorce between capital pressures and power, where under democracies, people can vote their way in a certain proposition, a certain power. And so I think you would expect here that in democracies, more people who otherwise wouldn't have a voice and therefore wouldn't have a role suddenly gain one. One person, one vote is a very powerful principle here. And we talk a lot about why we think institutions aren't as resilient under intelligence curse dynamics. But I think that in and of itself explains a lot of it.
I think in particular, pensions are an interesting one here because keep in mind that pensions and social security are really a promise from a government to you provided that you do X amount of work. That doesn't mean that it's literally that. It's often times people who don't do that kind of work still gain the pension. But the general idea here is that we have to support people after they've done lots of work. And so this enables a political environment that makes things at social security possible. I think it's our expectation that, like in places that get sudden amounts of resources, that this is a less stable arrangement, especially in the long term.
if people aren't part of the labor process at all.
Theo Jaffee
Sure, you talked about the elderly, but what about the disabled? The disabled are not economically productive. They, many will never be economically productive, unfortunately. And, you know, yet while a hundred years ago, we had a political system that often ended up like euthanizing or sterilizing disabled people. Now we don't do that at all. Instead, we pay sometimes very large amounts of money to keep disabled people alive and fed and
as healthy as possible. There seems to be no economic incentive to do this.
Luke Drago
So this is an area where think democracy beats a lot of these incentives, right? Like there are strong reasons for governments to go about other methods because their citizens won't tolerate certain things. And my expectation here is that, like I said, the pressures that you get when you remove everyone from the social contract, everyone from the value production process changes here. And I think one way to look at this is look at the UK's ongoing crisis with disabilities, where the UK has had a whole lot of what appears to be like...
misallocation or possibly fraudulent behavior with lots of people who maybe shouldn't be on something on benefits, a lot of people who otherwise could be working. Now, I don't want to make strong generalizations here, I've not followed this issue very closely, but what I can say is the political environment in the UK, now that it's becoming an increasingly large percentage of the population, has just overwhelmingly shifted. I think you can rely on altruism for certain subsets of the population. I think you can't rely on altruism when it's the entirety of your population. I think you should expect there that incentives are pretty strong.
Theo Jaffee
Okay, that makes sense. But what about like...
Luke Drago
And I wouldn't want to rely on altruism alone.
Theo Jaffee
Yeah. I guess throughout your piece, make arguments for why democracy is great. But it seems like democracy is only as great as the base of human capital that makes up its voting base. So Gen Z, for example, has, I believe, much worse values than previous generations.
more likely to be like socialists, they're more likely to be nationalists. It seems like, know, sort of half playing devil's advocate here, it would be better to have like, some sort of entrenched human elite in collaboration with AI, making policy decisions than to have like, you know, certain masses of people making the same policy decisions. So why do I want democracy in this case?
Luke Drago
I find myself in the newly contrarian position of defending democracy. And I think I've seen in my own lifetime how this has gone from the obvious and dominant position among my peers to one that is increasingly somewhat contrarian. I don't want to get up here and say that democracy is perfect, or it's solved every problem in the world, but I do think it is the best of lot of alternatives. And I think the core reason for this is pretty simple. Democracy is a bet on your power against somebody else's power. I think...
Theo Jaffee
Yeah.
Luke Drago
in societies where democratic structures don't exist. You are functionally at the whims of the state. In particular here, in a world where it's not democratic and also you don't have economic leverage. You are entirely subject to the whims of someone who isn't you, as to whether or not you get food, you get housing, you end up in prison. And I happen to think that for the vast majority of people, the benefit of democracy is that they get a say. so politicians have to be receptive of their interest. You know, in Western democracies, there's only so much you can do before a population removes you. And I think alongside this,
In non-Western democracies, if the conditions become so untenable the population wants to remove their leaders. The only possible response here is violence. And I think we talk about this bit in the piece. We expect that state oppressive measures get just much worse, or much more powerful, I guess. A state and infrastructural power gets much worse under non-democratic structures, under AGI, because suddenly you can remove humans who have limits as to what they'll do from the enforcement capacity of the state. And surveillance just gets significantly better than it ever has been.
And so I think for all these reasons, democracy ends up being quite resilient here, I don't think it's sufficient in and of itself. We talk a lot about this. But I think if you're looking at which states you want to be in post-AGI, the state where you have political power and the state where you don't, it's a pretty precarious situation if you're in that latter state and you also don't have anything valuable economically.
Theo Jaffee
Hmm. So throughout the essay, you talk a lot about what is called an economics rentier states, which are states which derive a lot of their revenue from natural resources and so have fewer incentives to invest in the human capital of their population and use this as an analogy for what every country will be like after AGI. But it seems to me like many of these rentier states actually do pretty well and are good places to live.
Luke Drago
Yeah.
Theo Jaffee
There's Saudi Arabia and the Arab Gulf states, which have recently improved significantly, not just financially, but also morally. There's Norway, of course, which you talk about extensively. And then the other two countries that you mentioned, Venezuela and Equatorial Guinea, it seems to me like Venezuela has been a rentier state for quite some time. It really started becoming bad after they adopted socialism. And Equatorial Guinea has, I guess, all sort of dysfunctions that are
Luke Drago
Mm-hmm.
hold on.
Luke Drago
you
Theo Jaffee
sort of typical of many post-colonial African countries and is not that well representative of all rentier states.
Luke Drago
Well, I do think one thing to note about post-colonial African states is that they oftentimes function like as rentier states on steroids because instead of having the government's like designed to extract a resource, you have a government designed to extract the entirety of the country's GDP from a foreign power. So in this case, let's say the colonial, like colonizer here was Britain and it's an African country that they've designed, or like, you know, it's an African country that they're colonizing. The entirety of the state's apparatus is designed to remove resources from that base and transfer it back to a different power base. When you do a handover here,
Sometimes it results well, sometimes it results poorly, but oftentimes these handovers that result poorly are because you just handed over extractive institutions to new extractors and oftentimes extractors that have like stronger incentives to behave benevolently than the previous extractors did. Now I guess to touch on like the other examples you've mentioned here. So I'm not deeply familiar with the defendant's swelling example as a case study. Obviously I've read the literature here and I think it shows that oftentimes relying on resources is pretty negative for your like
for like being resilient to shocks and all of these things. But I think there is a strong concern here about like your leverage under a Venezuelan system. One where the government has like a non-human resource they rely on for rents and also has removed your ability to vote. Well, I mean, they have elections, you know, I wouldn't call them free or fair. I think in this world, like you as a Venezuelan are at the mercy of the Venezuelan government. Your ability to sell your labor is limited. So your ability to get ahead is quite limited and your institutions just kind of suck.
deserve you. The Gulf states are fascinating examples. So think we outline two examples here of obvious pathways to get out. One of these is like the credible threat here where I the paper we cite talks about Oman a lot as Oman is a state that has a Gulf monarchy. It had a credible threat of revolution.
And ultimately like ends up resulting in giving away rents. think that the terminology we've been using is, know, the rent controllers would love to have all of the rents themselves. They'd really also prefer to have their heads attached to their body in like a cool and meaningful way. And so because of this, oftentimes the best choice is to capitulate, is to set up welfare states. Norway is an example of I think probably the best way you can do this, which is where like there were really strong institutions before the resource curses introduced, really excellent institutions, and they were quite resilient.
Theo Jaffee
Yeah.
Luke Drago
were quite stable, they were low corruption. And I guess the question I oftentimes ask is, does anybody think that most Western democracies have the kind of resilient institutions that Norway has? And secondly, do you expect the effects of AGI replacing all of human labor to be stronger or weaker than oil which replaces some but not all? In non-inventory states, there are other ways to make money. It's just the government primarily relies on one, where in this scenario, you are constantly out-competed. Now I do think for what it's worth,
Norway provides some interesting examples. are reasons why you should want to do the institutional strengthening that Norway has done. The other thing to note about the Gulf states is they have an incentive to diversify recently. It is not a coincidence that the Gulf states have gotten better as diversification has become more likely. I think we talk about it a bit in the piece about how Saudi has like, assumes that peak oil is coming sometime soon. And so because of this, they want to attract diverse capital investments into Saudi Arabia. And correlated with this has been a rapid expansion relative to baseline of women's rights and economic freedom, which you would expect in a country
that can no longer rely just on a resource and now wants to shift towards focusing on humans as a form of economic growth.
Theo Jaffee
Yeah, okay, that makes sense. But you said something that made me think of like, yeah, people want to extract rents, but they want more than that to keep their heads attached to their bodies. But I guess another form of this like heads attached to the body's thing is social status, right? And you even talk about how social status is, you know, one of the things that humans will still be able to do after the singularity.
Luke Drago
Yeah, they want to stay alive.
Luke Drago
Mm-hmm.
Theo Jaffee
Like, let's say you have a, I guess, benevolent AI dictator, the CEO of a big lab or something, who's making decisions about, what to do with all this, you know, extracted rent, and, you know, he can choose to keep everything, even though this goes, well beyond what he would ever need for personal consumption, and then some, or he could give a lot of it away to the people, I suppose, and gain, like, a lot of social status for that. Could that be a new social contract?
Luke Drago
Okay.
Luke Drago
So there's a word that's doing a lot of heavy lifting there, and it's benevolent, right? We're presuming here that this actor is benevolent. And I think you can presume for a generation this is true. Let's presume it's humans are in charge.
But I think a real problem with presuming you have the benevolent dictator is your Stalin risks are just quite high. You've built a set of institutions and a set of governing powers which, if handed to someone who is not benevolent, can suddenly become quite destructive. And the benefit of pluralist liberal democracies is that even if somebody who's quite destructive comes to power, there are lots of reasons why you can survive without the government and the government is beholden to you, that individual has limits on their power. It could be that one of the AI benefactors, or all of them today, are benevolent, but as
gets passed from generation to generation, they end up in a situation where the entire economic means of production and political power is centralized in an actor who just isn't that benevolent. You saw this with monarchies a lot. You know, there are a lot of benefits to monarchies historically, but there are a whole lot of downsides, and the reason we move past them is because if the sovereign has absolute control, every time you have a handoff, you just roll the dice as to whether not that sovereign's any good. And oftentimes it results in pretty catastrophic sovereigns.
Theo Jaffee
Well, the Curtis Yarvin response to this, which is, guess, now the orthodox sort of like Bay Area gray tribe response to this is you don't have an absolute sovereign. You have a sort of constitutional sovereign that can be replaced by a board, which is what you see at these big labs. The board failed to replace Sam Altman at OpenAI, mostly because it seemed like what they were doing was very suspect. But if you had Sam Altman clearly showing like
Luke Drago
Yeah, yeah it is.
Theo Jaffee
public signs of being bad towards the people, then you could have a board which themselves own OpenAI stock and would be very wealthy after the singularity. It seems like companies right now are not controlled by absolute monarchs.
Luke Drago
So I agree with the companies today or not. I think there are some limitations. But guess Theo, I'd ask you, what control do you have over open hand?
Because you have some. You have a little bit of control.
Theo Jaffee
I know, I could talk to my friends who work at OpenAI.
Luke Drago
Yeah, well that too. But I'm thinking more of like, what kind of like direct power could you exert on open A?
Theo Jaffee
I could cancel my subscription and switch to Claude.
Luke Drago
You could do that, but also you could vote, and you could vote for politicians who influence open AI in some sort of meaningful way. The market can move here and just cancel everything, and that's one way to exert power. And another way to exert power here is because you can vote. This is your political leverage and your economic leverage. And I'm concerned about scenarios where the goal is to remove one or both. And I think it's pretty clear to me, given that the stated goal of the major labs is to achieve AGI, which they define as systems that could do all meaningful human intellectual work, that that is on the pathway.
Theo Jaffee
Okay, so let's talk about solutions, right? So what should be done like right now, like today? So if you were the president of the United States, what exactly would you do today right now?
Luke Drago
So a couple of things, I'm the president of United States. My goal here is to ensure that we have like lots of different, like a stable, multipolar super intelligence where lots of different people have access to the technology. They can wield it at their pleasure and we're also not in the world where like...
crazy stuff is happening all the time because suddenly everyone's got these crazy capabilities in their pockets. I'm the president of United States. I'm investing a couple billion dollars, provided Congress goes along with it, into a moonshot for a whole lot of risk-reducing technologies. I'm thinking about hardening the world against the major reasons to centralize. One, because catastrophes are bad and are possible. And two, because I think one of the biggest threats against human liberty here is really centralized AGI projects where the government comes in and says, there's one project. We are producing super intelligence.
us and I think the best way you get to those kind of things is with some sort of like AI warning shot or the very real threat of AI catastrophe. The downside of this is if you know like or the upside of this is like you you remove the ability for some crazy catastrophes to happen which is good for everyone and you also can pave the way for a safer more like democratic ownership structure and control here.
Theo Jaffee
Is there anything else other than this democratic moonshot?
Luke Drago
I mean today that's the major thing I've been doing. mean also like a lot of like anti-corruption strengthening I want to do here. If you're going to rely on institutions and you have to in some way to ensure that like AGI goes well, you're going to want to make sure that the people who are in power are constrained by like pretty reasonable forces. You don't want lots of corruption. You don't want people to be able to buy each other off. You want to make sure that people's votes matter. So like you know, there's stuff as simple as like campaign finance reform, strengthening anti-bribery laws. I think Singapore's interesting example here where like
Singaporean civil servants get paid very handsomely and if they take a dollar from the, you know, someone who's not the government, they go to jail. And I think this is the kind of, when you're dealing with, you know, complete rewrites of social contract, you want to make sure that, you know, benevolent people and people who have good values are in charge. But you also want to make sure that if you roll the dice incorrectly and you get the wrong person, they're constrained by forces that are stronger than just their own will. So that's the kind of thing I'd want to be building up as well.
Theo Jaffee
Okay, what do do if you're the CEO of a big AGI lab today right now?
Luke Drago
I mean a whole lot, right? Like if this is the problem that I want to solve, maybe it's a bit against my company's interest, but I probably want to be first of all like doing a bunch of interesting research here on the economic impacts. So I think like Claude releasing, kind of, yep, Claude, Anthropic releasing like the Claude index in this seems really good. And I think more lab should be doing this. I'm somewhat skeptical of Anthropics data as the baseline, not because Anthropics data is wrong, but because Claude is like predominantly used. Like it's pro, it's like biggest use case seems to be coding. And so I think it's not necessarily representative. I think if you want to represent that, like really representative data here.
Theo Jaffee
Anthropics doing this? Yeah.
Luke Drago
you'd want to open AI because everybody uses GPT. Everybody uses chat GPT. My mom knows what chat GPT is. Your friends who don't know anything about AGI know about chat GPT. So that's really who's evidence you're going to want here. You probably also want to be doing some baseline research into what sectors you expect are going to get hit the hardest and be sharing this with policymakers.
Theo Jaffee
Is that it?
Luke Drago
There's more there. I want to highlight the kind of the top line stuff. I think looking at decentralized platforms, another exciting thing here, like I Prime Intellect just did this like massive decentralized training run. Looking more into like how you can do like model customization ways that are privacy preserving, how you can tap into people's like tacit or implicit knowledge without owning all of their data. These also seem really important.
Theo Jaffee
How do you tap into people's tacit knowledge without owning all of their data?
Luke Drago
So some interesting stuff going on here. think like as like a kind of a core observation, my expectation is that for the last mile of automation to really know what's going on in the economy, to able to allocate stuff better, you don't just want like clones of the exact same model running around.
One of the reasons that like markets work so well is because there are lots of different actors who have slightly expertise in like small slices of the world And because of that they can like see things that like a central planner just can't so you'd want to be able to incorporate that information in a meaningful way and I think there are two ways to do it there's one that I advocate against and one that I'm hoping to help contribute to I think that first one is just you like gobble up all the data make carbon copies of a user Don't do this in a way where like you now own the data as the lab and then you create like lots of clones that look just like them they can mimic their behavior and preferences and have access continually
to the kind of stuff that they're doing. I think if you want to do this in a more privacy and preserving way, there's some interesting papers here on like, like secure training runs for example, secure training, I gotta find the exact terminology from the paper, I don't have it right in my head, I just read this a couple of days ago. But there's like certain kinds of training you can do where like you can train on data that you can't necessarily see. There's a question there about whether or not you can evaluate this, this is like kind of an unsettled question right now.
So could you evaluate a model that otherwise you don't have access to? And the answer to that's probably no right now. That probably takes some hit for consumer performance. But are there ways in which you can either use it to directly own the models and run them on device, and so therefore you can't see it? Or could you do Apple private cloud compute style solutions where your data is being passed on but in a privacy preserving way? I think I'm much more excited about these latter options than the former ones.
Theo Jaffee
Hmm. What do you do today if you're the CEO of a white collar company? I guess like I know sort of tangentially like a lot of people who are, you business owners or executives or whatever of like, you know, sort of provincial companies. Like I live in Florida. I don't live in Silicon Valley. And so like most of these people who are, you know, big business people don't know anything about AGI. They might know like, yeah, Chachi B.T.
is pretty good, you know, you try it, it seems pretty helpful, it's not gonna take anyone's job right now. So what should these people be doing right now?
Luke Drago
Yeah.
Luke Drago
So it's really hard for me to tell business owners they shouldn't like follow their incentives because like that's the way that the economy works. My expectations of like the era of like mega corporations probably coming to an end because in those corporations you just you just really can't automate a lot of those jobs. I think the way this happens I think we lay out the first part of the series and this idea of pyramid replacement where like major corporations just start by hiring less people and eventually at some point they do layoffs. But this really affects like your entry level employees and analysts first. I do not and I'm not going to give the advice of like
Just stop hiring people, because I don't think that's good for a whole lot of reasons. But I expect that's what a lot of companies are going to have to do to remain competitive. I don't think this happens overnight. I don't think AGI 2026, zero human employment 2027. There's this gradual diffusion process. It's not like laptops came out the next day, everybody had a laptop. This just takes some time to diffuse in the broader economy. I expect that if you're a company and you want to survive, start using AI as fast as you can.
Theo Jaffee
Cowen has said something like, this will take 30 years to diffuse throughout the economy. And I think the epoch AI people think sort of similarly.
Luke Drago
That seems plausible to me. It seems completely plausible. I'm probably more in like 10 to 15 years than I am on 30 years, but I do think it's going to take a long time for this to diffuse. I think the iPhone is an interesting example here where like in 2006 or seven the iPhone gets released and by 2016, 2017 you basically have to have an iPhone to compete. In the meantime, Blackberry dies, right? Or you have to have like a smartphone to be able to like be in the modern business environment. Maybe that's a more accurate type of diffusion that you would expect where like something goes from being niche to being mandatory. But iPhones are building on an existing platform of cell phones and those are already pretty much required.
by that point. So I don't want to say I have the exact model of diffusion. What I will say, a reason to think that it's not going to be overnight, but it's going to be pretty quick, is that of course, like for a moment, ChatGVT was the fastest adopted consumer app ever. I can't remember the exact number, but it had 100 million users pretty quickly. And I think you should probably expect that, like, I Cursor then beat it out, right, if I recall correctly. And Cursor was, of course, just like using that technology for one vertical. I think that's going to happen in a whole lot of industries where pretty much overnight they get changed. I don't think this overnight changes your hiring path.
But I think it overnight introduces new incentives that get stronger and stronger.
Theo Jaffee
Yeah, okay. So what are these bottlenecks to AGI progress? Why do we not have AI 2027? Why do we not have labor getting replaced? Basically immediately after we get AGI. What are these bottlenecks?
Luke Drago
So are these bottlenecks like strictly on progress or bottlenecks on diffusion?
Theo Jaffee
both.
Luke Drago
So first of all, I I have a lot of respect for the AI 2027 team. I've talked to lot of them extensively. think Daniel et al., they're like brilliant thinkers. I probably think that the world has a lot more like physical bottlenecks than I think their team does. So I think, for example, like long horizon planning might be a harder objective to achieve, especially in the last mile of it. I think, for example, we seem to get a lot of progress very quickly on stuff that is obviously right or wrong. Does the code run or not? That's an area where you can get pretty superhuman pretty quickly. And you should expect like RL and self-reporting
to be like pretty effective tools here. I expect that some of these harder tasks, like let's just take writing for example. I think that today, like models are pretty decent at writing. I do not think that they are like at...
college-educated human baseline, or maybe like, slightly above that at best. And I think there's not been tons of progress towards this. There's a lot of like, fanfare raised by the OpenAI team when they dropped GPT 4.5, and they kept saying this is like, the best model for writing. They had this big tweet where they put out as an example of the writing quality. And I read the writing.
And I just wasn't that impressed. I thought it was better writing than I expected from a language model, but was definitely not what I would consider to be like good human writing. It was maybe better than your average college student, I don't know if that's good human writing. I think these tasks that are harder to judge are just harder to train. There aren't right answers. You have to have taste to be able to judge them. And I think like the job of running a company, for example, which requires a lot of success and failure, it's kind of hard to predict exactly how it's going to work. You have to have lot of intuitions that are just hard to build without doing it yourself.
These are tasks that are very hard to simulate, very hard to prove objectively, and very hard to measure. And I think eventually we'll be able to generalize to those, but I do expect this is a reason to think that AGI is like slightly farther away. Though I think if I had to put like a timeline on it, I think I'd probably say 2030 maybe sooner. So don't think it's that far away. And then I think on the economic diffusion side, like...
Luke Drago
court is agreement I have with AI 2027 is like 10 million robots one year after AGI. And I think that they presume, and pretty smartly I would add, like a whole of government response with like every part of government machinery suddenly awakens and tries to create robots. I think even in that world, there are lots of bottlenecks. Like for example, we need a whole lot of rare earth minerals or REMs to do this. And right now China controls 85 % of the refinement process for all REMs. The US doesn't have much domestic capacity online right now. I think it's 85%.
Theo Jaffee
Sure, so you might still have your 10 million robots, just they're made in China, not the US. I mean, you already kind of see this.
Luke Drago
Well, I think in a world like this, you might get 10 million robots in China, but the US government is not getting the whole of government response that gets them to 10 million a year after AGI.
Like I do think right now China is going to crush us in the manufacturing race. And I think this is like an existential threat to Western civilization that we just cannot build anything anymore. And I'm pretty concerned about those kind of scenarios where, you know, we end up losing the AGI race, not because we didn't get to the software first, but because we instead, like, completely capitulated our manufacturing capability to a rival or an adversary who now has strong incentives to cut us off to buy themselves time. Look, we've done a whole lot of export controls to them. I'd be pretty shocked if they don't do a lot of export controls to us.
Theo Jaffee
Sure, but why should I care if China achieves AGI first?
Luke Drago
I think there's a question of what values do you want to underwrite the world and what kind of power do you want to have?
For the same reason that you should care about still having an economic role in the social contract, you should probably care about your country or a country that are aligned with yours having an increased role in the economy. And in a world where China can build everything and manufacture everything cheaper, I think it's harder and harder for the West to play competitively at the global stage. I think ultimately it does, like, your team winning is actually important. And in particular, your team having a way to win is important.
Theo Jaffee
Well seems like China has absorbed a lot of the cultural values that I care about from the West. You know, they got capitalism with them, Deng Xiaoping. DeepSeek was for quite a while and maybe still is the most free and least censored AI model. Like yeah, you can't ask it about 1980, 1990, Ottoman Square. But that was a much narrower category of restrictions than the restrictions that OpenAI had on all their models until very, very recently.
Luke Drago
Which ones?
Luke Drago
So one, I think it'd be good if we had less content style restrictions. I'm a very strong believer in free speech, and so I think it's important that models can answer questions truthfully and without undue censorship. I think that's different than models providing you with instructions on how to build a weapon. But I think that's what's more important here.
I want to go zero in on like, you know, Deng Xiaoping brought capitalism to China. is absolutely true that like China is, you know, economically liberalized far more than they did at that point. But the kind of capitalism that China has is one where if the CEO of Alibaba makes a derogatory comment about a Chinese, a part of the Chinese economy, he disappears for a few months. To me, that is not effective capitalism and it's a value that is quite foreign to my own. I would strongly oppose it if it would have occurred in the US or in the West and I strongly oppose it when it happens in China. And I don't, I mean, I strongly expect that a world where like the
Chinese government gets to write the values of the world is one where that is more common, not less.
Theo Jaffee
Okay, that makes sense. What specific tasks do you think humans will remain on, like, in the loop for the longest? Like, you talk about tasks that require taste, that are hard to judge, that require long-form planning. But it seems to me that, like, beyond just that, like, AI seems to be just better at certain tasks than others. Even some tasks that seem rather obvious. Like,
Luke Drago
Yeah, we're getting, go ahead, please.
Theo Jaffee
I don't know, what is a good example of this? Like, AI can't really play tic-tac-toe still, I believe. Maybe O3 has been able to do this, but...
Luke Drago
Yeah, I mean I think jagged progress has been the norm and I think this probably gets more true with reasoning models, not less. Where like, you get really good at stuff that has correct answers and not great at stuff that has incorrect answers. Now I expect this to generalize. think, I think, you know.
By 2030, I expect this to be like, at least more solved, much more solved than it is today. But I do think these durable skills as it matter right now are like taste, judgment, long horizon planning. I was giving a talk at Georgetown a couple of weeks ago about like how to plan a career in the age of AI. My general advice was if your goal was to like go climb a very large corporate ladder and spend 10 years at McKinsey, become partner there, like at an accelerated timeline, you are just probably going to get automated. The bulk of your job for a long time is going to be tasks that are quite automatable. Whereas I expect that like,
If you're learning very early how to take risks, how to fail, how to develop that research taste or that sense of taste judgment faster, you'll be more effective than your peers at racing against AI progress.
Theo Jaffee
Yeah, I think yesterday Marc Andreessen got clipped on a podcast where he said something like, VC will be one of the most durable careers to AGI and everyone clowned on him for it. But it seems like, know, if anything requires taste and judgment and long horizon planning, it's venture capital, right? Like what more so than that?
Luke Drago
I saw that, yeah.
Luke Drago
So I don't know if I agree with the claim that venture capital will be the last job. I probably do agree with some version of that claim. It's actually very hard to predict what companies are going to succeed and fail. That kind of pattern matching is quite difficult. Jobs that require that kind of pattern matching that require years of experience to figure it out, those are more likely to be durable than less. The real bad news here is for entry-level roles, for the roles of people who are just coming out of college. I think there's some terrible irony in the fact that like,
entry level CS majors are just kind of automating themselves here. The last mile of automation is not required to automate the vast majority of coding, to automate the vast majority, especially like entry level coding. But yeah, don't think, for what it's worth, I get why he got Clown Dongs. I think the specific comments sound a whole lot like, but my job will survive. But I think there's something in that comment that I think is true.
Theo Jaffee
Hmm. You also talk about AGI society being permanently static as a risk. But why would it? It seems like nature hates stagnation. Especially if there are different people controlling different AGI's in different factions.
Luke Drago
I think we've really focused a lot on economic stagnation, right? Like the idea that...
you as an individual, you're born in Montana and your ability to ever climb that social ladder is just quite small, quite unlikely today, in this world relative to today. I don't think it's true that everybody born everywhere has an equal shot of climbing the social ladder and displacing the current elites of the day. But one of the benefits of our existing system is it's actually just quite easy relative to baseline to do that kind of displacement. And one of the ways that culture moves and culture shifts is through leader displacements. I'm thinking a lot about people talk about the vibe shift in 2024,
where there was very suddenly a new administration was elected. They hadn't even been sworn in yet, but they were elected. And it felt like overnight, basically every company suddenly switched through the hiring policies. People started saying different words. It felt like there was a massive cultural shift from June 2024 to December 2024, maybe even into January 2025. This is all very much so a byproduct of leaders changing and therefore what seems permissible for the culture also shifting as well. And I think that that's an important part of human social progress.
know, human social dynamics is that people from the margins can still win and I think the cultures that are the most alive and the most dominant are the ones where like figures like a JD Vance who grows up in like the poorer regions of Appalachia can ascend to the vice presidency. I think it's one of the things that makes the Western democracy and capitalism so strong is you can move from the outside and still win and with that create lots of change and progress. And I think in a world where like your wealth is turned by government, it's just as likely.
Theo Jaffee
Okay.
Theo Jaffee
So by permanently static AGI society, you mean the humans will be static, but not the AGI's or the people at the top.
Luke Drago
Well, I my concern is the humans, right? So yeah, I think that's who I'm referring to here.
Theo Jaffee
Yeah, okay.
You also talk about banning AI systems from owning assets. Let me find the exact words because this is
Luke Drago
Yeah, is one of the few times I call for banning anything, because we really try not to rely on regulation as a core centerpiece of what we're doing here.
Theo Jaffee
Yeah.
Theo Jaffee
policymakers should ban AI systems from owning any assets, serving as a C-suite member of a company, serving on a board of directors, or owning shares. Yeah, so this seems like very specific. I would love more elaboration here. Like, why would you do this? How is this like tenable in the long term when AI becomes more agentic?
Luke Drago
Yeah, so we brought the...
Luke Drago
So I guess like, you know.
We really try in the piece to avoid doing specific regulatory call outs. think if we're running around saying we think the underlying economics are more important than the regulatory structures, we probably shouldn't be spending all of our effort doing lots of targeted regulation. I think this is just broadly the wrong way to go about changing social progress. think if new technology introduced, you want to adapt your society around them. I think regulations are oftentimes ways to, instead of adapting, slow down. I'm not anti-regulation by any means. I think sometimes they're quite necessary. But I do think we really try to focus on the
underlying economics. Now, I think this actually looks a lot more like the underlying economics and regulation, because I think this kind of rule today changes nothing right now. There are no AI systems today that can legally own assets. I think, honestly, it's probably likely that under most existing law, AI systems won't be able to own assets. But I think the reason we include the call out is because by doing this, you're guaranteeing some sort of human role in the organization of the future. And if it is the case that eventually you do delegate this power away,
I expect evolutionary pressures very quickly to make this a dominant force, where suddenly AI's control, in and of themselves, lots of capital. My expectation is just that humans are more likely to care about humans than non-human entities are. Same reason that we care more about humans, do animals. And so I'd like to preserve as much as possible a role for people in the setting of the direction of the future. And I want AI to be a technology that extends that direction setting, not limits it. And so I think that's one of the areas that we did a pretty targeted policy call out.
Theo Jaffee
You also talk about AI both mainly centralizing but also decentralizing power by default. Which do you think will dominate? Because I've heard both. I've heard AI will enable governments to surveil the entire populace, but also AI will enable the populace to build pocket nukes that can be used as a check against the government. So which of these trends do you think will dominate?
Luke Drago
Thank
Luke Drago
So for what it's worth, think a world where everybody can build a pocket nuke is a pretty bad world.
I mean, I think one of the areas where things do get destabilized is one where a bunch of people have access to weapons of mass destruction. Now, I think there are other choke points that are really effective here. So it could be that you have a system that can tell you how to build a nuke, but you just can't get uranium. Uranium is pretty hard to get, pretty hard to refine that in your backyard. So think there are reasons to believe that that is harder than it is. Bioweapons are another area where I'm actually pretty concerned. It's why we have a pretty targeted section on reducing bio risk. But our section there is really targeted in other choke points.
like doing like you know screening KYC on like the biological materials because I think they're just physical world choke points that are more effective and also like reduces power trade-off. Provided that you can do that kind of stuff though I think one of the ways where AI can keep people in power and be like an agency extender as opposed to a limiter is one where like lots of people have access to like very powerful models that are aligned to their values, their goals, their intents as opposed to like a centralized set of goals or a centralized set of intents. And I think if
believe you kind of buy high X arguments on how knowledge works in society and candidly I do. That there are like parts of economic knowledge kind of scattered throughout everywhere that are somewhat hard to track and are difficult to be legible. Systems that have access to like an individual user set of data there, know, tacit knowledge, might be more effective in aggregate, like organizing an economy and allocating resources than any sort of like top-down centralizer is. I think my concern is that, yeah, sorry, go ahead.
Theo Jaffee
Yeah, I for what it's worth, I DeepSeek is probably like the single greatest company for decentralization right now. And they come from China.
Luke Drago
It worries me that we don't have a Western alternative that I think is effective. I think it could be Meta, but it seems like Meta, both ideologically isn't that committed and yeah, I don't know what's going on with their AI team. Prime Intellect just did a really interesting decentralized training run. We're waiting for the result of that right now for Intellect 2. And I'm quite excited to see companies like that.
Theo Jaffee
that is throwing right now.
Theo Jaffee
Hmm. But sorry, what the?
Luke Drago
which yeah, I do think like.
Sorry, I do think there's some concern about if it's DeepSeek. If DeepSeek is the company that underwrites the value substrate of the universe, and everybody's only asking DeepSeek questions, and that's the powerful model, I think I'd much rather that model be made at people who are aligned with my values than not. Because I really would not prefer a model that centers my political opinions. And I agree that there are areas where it's censored less, but I don't think that response to that is then to say, we should give up, throw up our hand, and say, I guess DeepSeek gets to win the race. I think we should be saying there are lessons that we can
learn from DeepSeek and lessons that we shouldn't learn from DeepSeek.
Theo Jaffee
Sure.
Theo Jaffee
So what should people be excited for? You talked about this culture of indefinite pessimism that seemed to replace the culture of indefinite optimism of Silicon Valley. But the next few years will, regardless of changes in material conditions, have some very cool things going on. So what should people be excited for?
Luke Drago
Yeah. Well, one, one of my big call outs there is because I think a whole lot of people can like see AGI coming like a meteor about to hit them. And their response has been like.
of mix of we're gonna die, this is gonna be bad, and from there, the response has been let's just do nothing about it, or let's freak out about it. And I think if you can see a meteor coming, but you can see it has lots of benefits, I think your real goal should be to deflect the meteor, and I think there are some obvious downsides that we can see coming. I think we talked about economic downsides, there are risks to, significant risks to, like.
catastrophes, etc. But your response to that should not be, okay, we have to throw up our hands and run as far away as we can and maybe go live in a cave for a while. Your response should be a call to action to go and solve these problems because this meteor is coming. We are going to achieve AGI. is basically at this point technologically inevitable. We are on the track to do this. I think it's extremely likely we will do it. And so I think hiding in fear is super unlike, is just not a good response. If we can solve these problems, we can solve these challenges. We are talking about living in a world with more abundance than we could ever possibly imagine.
imagine, where an individual's ability to change their own environment, to change the world around them, to make an impact could be higher than it is today. And I think it's pretty high today relative to baseline, where a lot of the scourges that we have in modern life are things we can send to the ash heap of history, things like some pretty terrible diseases like hunger. I think that is a really exciting world you want to be aiming towards. And I think you need to be sober about the problems that are facing you from getting there. Because I think there are real potholes in the road. I align with Vitalik's vision on this entirely. There are some real problems we're going to have to overcome.
But I think we can overcome them, and I'm pretty optimistic about humanity and our ability to do that. But it does require us to be sober about what the risks are and what it's going to take to deflect them.
Theo Jaffee
So you mentioned you were building a company to address the challenges of the intelligence curse. So could you go into a little more detail about that?
Luke Drago
There's not a whole lot I can say right now. I think what I can briefly talk about, and I'll try to be a little careful here, we're pretty excited about this kind of alignment to the user concept. The idea that there's like a set of tasks and knowledge that we can gather that provided we do this in a privacy preserving way, both like creates models that are like.
more customized to the user and then when plugged into a lot of tools could complete tasks better than an off-the-shelf model could. I'd be pretty surprised if this model is superhuman at coding, but I think it probably will have a better sense of how to do things the user would want it to do, especially if it can do things like tool calling. So I can't say a whole lot right now, but I think if that kind of vision is something that inspires you, we'd love to hear from you.
Theo Jaffee
Alright, so yeah, I think we've talked a lot about AI. Let's talk about music a little.
Luke Drago
Let's do it, because I remember when I reached out to you, my pitch was AI, but also I think we have like very similar taste in hip-hop.
Theo Jaffee
Yeah, this is like weird. I don't know that many other people who like Logic as much as I do. And you said that College Park is in your top 10 albums?
Luke Drago
It is, yeah. So I think there's like an arc for Logic, right? There's like old Logic, which is very good. There's whatever happened from 2016 onward, which is, you know, hit or miss, pretty rough. And then there's like 2020 and after. I think everybody's 2016 or 2017, one of the two. And that album's pretty hit or miss. I think like the low point for me is definitely either Confessions of a Dangerous Mind or Supermarket, where it feels, Supermarket in particular, I remember listening to that and just going, no, like, that's the end of that.
Theo Jaffee
What year is everybody?
Theo Jaffee
Yeah, okay, yeah.
Theo Jaffee
Yeah, definitely.
Luke Drago
But then he has this resurgence with, like, is it No Pressure and then College Park that I'm just like, I thought they were really excellent albums. think there was the, what was the other, the vinyl, 035 is good, and then was also, I think, Vinyl Days, is that what was called? Yeah, Vinyl Days. I thought Vinyl Days was like some really interesting production. I think Logic does really well over like these like soulful beats or these like...
Theo Jaffee
No pressure.
Theo Jaffee
Ultra 85.
Theo Jaffee
of Vinyl Days.
Theo Jaffee
Yeah
Luke Drago
very like old school 80s, 90s inspired beats. And I know his most recent project is coming out is neither of those things. And it's quite like trap again. And it's just like, I don't think he can do trap very well. No offense if Logic has tuned into this AI podcast, but I think, you know, do more boom bap. That is definitely where he shines.
Theo Jaffee
Yeah, I agree.
Theo Jaffee
mixtape logic versus album logic and like he even talks about this I think in the intro to Bobby Tarantino 2 or 3 he has like a Rick and Morty skit where he has Rick saying like yeah literally where he has Rick saying like you know I don't want to listen to this like you know introspective rap with a message I just want to you know rap about like titties and ass and stuff and like yeah like when logic does this it's usually not that good
Luke Drago
Average logic skit.
Theo Jaffee
I think my favorite Logic mixtape after the original Young Sinatra's was probably Inglorious Basterds, like the very new one, because it was closer to an album than like the Bobby Tarantino's.
Luke Drago
Yeah.
Luke Drago
I think this makes sense. think my favorite Logic project, I'm standing by College Park, I think it's got one weakness, it's that oftentimes it's got these skits that are two minutes long in the middle of a song. So Playwright, for example, has this very long skit. I think Playwright is an exceptionally well done song. Its beat is really catchy, Logic flows over it very well, the features are really excellent, and then it's got this minute and a half long skit, and because of that skit, I'm not playlisted at all. If I just play this, it's just like...
Theo Jaffee
Mm-hmm.
Theo Jaffee
That's a good one.
Luke Drago
I think he's ordering a burger in the skit and she's like, what are we doing here? Like old school would like take his skits and make them separate tracks. And I think, you know, I would have really preferred a version of College Park where I could play some of these songs in the skits. But I think like Lightsabers, Gaithersburg Freestyle, Self-Medication, Shimi are just like really like top tier logic tracks. I can't think of really many tracks he's made that like outpaced those for me. Maybe like OG Under Pressure stuff.
Theo Jaffee
Yeah.
Theo Jaffee
I Under Pressure is better. Yeah, but College Park in top 10 is quite high praise, think. Let's see. Tracklist is...
Luke Drago
Look, I'm happy to make a chair. I'm making contrarian takes an AI might as well make a music too, you know?
Theo Jaffee
Yeah, yeah. Cruisin' Through the Universe, like, I thought that one was great. It's so like, it feels like you're in space with the production. Like, six, six is great. Six on the beat is like God on the beat.
Luke Drago
Yeah.
Luke Drago
Every time. Here's like a, he had a mix tape came out a couple of weeks ago.
Theo Jaffee
Who's six? Okay, I have to check that out.
Luke Drago
Yeah, Six did it. It's like, he's not rapping on it, but it's like he's producing the whole thing. And there's a logic track called WMD on it, which I'm like pretty impressed with.
Theo Jaffee
Hmm. And then like the...
Luke Drago
By just fixing logic, they can do stuff together that's really impressive.
Theo Jaffee
Yeah, the end of Cruisin' Through the Universe where you get into the first skit and you're like, wait a minute, why is Logic in Big Lembo's basement? And then you're like, that's what College Park means. He's talking about his time in College Park when he was 20, sleeping on Lembo's couch before he blew up. This sort of Logic adventure is good.
Luke Drago
Yeah.
Luke Drago
Yeah, and I think that's like an, yeah, I mean think when he's talking about like, logic is good when he's talking about literally being in space, or like his early life, and I think that's where like he just really shines in ways that are quite impressive. And I think everybody has this like much more confusing narrative where it's like.
There's this Neil deGrasse Tyson side plot that's going on where it's like, actually, you are everyone that's ever lived, which I believe originated from like a Twint, some Facebook meme. You know what talking about? Yeah, the short story in the short story became a Facebook meme, and I'm not sure how it got to logic, but it's like, why are we doing the short story that I've already heard about? And also like very strong political commentary, and also at the end, some space stuff. It's like there are three different storylines happening here. I'm kind of confused as to why they're happening. I think they meld somewhere.
Theo Jaffee
No, was a short story. Called The Egg.
Theo Jaffee
Well, the space stuff is kind of just a through line between different albums. And like the reason for the short story is kind of like, you know, you are everyone, right? So you have to, you know, love everyone and, you know, peace and love and equality and stuff. Like that's, yeah.
Luke Drago
I think it's a...
Luke Drago
Yeah.
Luke Drago
Yeah, no, I think the reasons there, I just think it's not executed as well as it could be. And I think some of the songs, I, there's a track on everybody that I think, I remember being like, what are we doing here? If I can find it, it's, it's Take It Back, it's Take It Back. Yeah, it's Take It Back.
Theo Jaffee
Does it take it back? Yeah, yeah, Yeah.
Luke Drago
The beat's excellent, right? It's just like, you know, like kind of three minutes of logic talking for a while. God, I think Killing Spree still makes it in playlist. I think I can just be like, actually, this is fine. I can look past it. The production on everybody is maybe his best production, which is why it's somewhat unfortunate that the lyrics are just not there.
Theo Jaffee
killing spree too.
Theo Jaffee
Yeah.
Theo Jaffee
Yeah, but like I think everybody like the beginning of the album and the end of the album are really excellent Like the beginning of the album you have hallelujah and everybody
Luke Drago
Yeah.
Yeah, yeah. Confess too. Confess is slept on. It shouldn't be.
Theo Jaffee
Yeah, and then like towards the end of the album you have 1-800 which I'm sorry. It is a good song I don't care if it's his most popular song ever. It's definitely not his best yeah That one's a little corny, but like it's really good. I think anxiety was great. Black Spider-Man was great AfricAryaN is excellent like it ends so well
Luke Drago
Who came here late?
Luke Drago
I think it is a good sound.
Luke Drago
So I think anxiety's not that good, but Black Spider-Man I think is maybe a top five logic track. I do this like a...
Every season, like spring, winter, fall, summer, I make a new playlist. And my only rule for the playlist is it's to be like 30 songs tops, which I add throughout the season, and no repeats between previous playlists of the same theme. So like if I did something in 2022 winter, can't come back in 2024 summer. And my like 2025 spring playlist had Black Spider-Man in it. And I just, I remembered how good that song was in the last couple of weeks, because it is just absolutely incredible from the production, the features.
I don't know, think it is like, if every song on everybody had like that framing of the mass stage and that level of caliber, I think everybody would have been a smash success.
Theo Jaffee
Yeah, I think you're sleeping on Ultra 85 too because Ultra 85 is like, yeah, I think it's maybe his best album maybe. It way exceeded my expectations for what it would be based on like the promotional singles and based on his other projects from around that time.
Luke Drago
All 35 is good.
Luke Drago
What was the promo single? was 44ever right? Of fear.
Theo Jaffee
Fear. Fear and then 44ever. 44 ever was good. Fear was less good. Fear was alright.
Luke Drago
I will say the opening two tracks on Ultra 85, so Paul Rodriguez and Mission Control are just also fantastic. Mission Control in particular, think the production is stellar. Logic, when he really locks in, just write a beat in a way that think basically nobody else can. Like a couple of other artists have said they do this at their peak, like one in four beats, he just really finds a flow that's pretty infectious. And I think this track just like pretty much exemplifies that.
Theo Jaffee
Paul Rodriguez was just incredible. I remember exactly where I was the first time I saw, my god, Logic dropped. And I was in Japan with my cousins on the train going back from Osaka where we had day tripped back to Kyoto where we were staying. And I was on this train, this packed Japanese train. I was just smiling on the train listening to this. I was like, my god, this album is going to be actually really good.
Luke Drago
Mm-hmm.
Luke Drago
Yeah.
Luke Drago
Yeah.
Luke Drago
So funnily enough, I had like a similar reaction. I remember exactly where I was when I hit the play button on Paul Rodriguez. was, I live in Limehouse in London and I had just gotten off the tube, was walking down the steps of the tube station and like hit play on it and we're just walking home. I walk home, like, I think just about exactly nine minutes. So it's the length of the full song. And I remember just listening to it being like, whoa.
Like, where has this been? I don't think the album lived up that in every part. I think some tracks did. Favela, thought was really excellent. Interstellar was really good. I, like the one mistake of Interstellar.
Theo Jaffee
Yeah.
Theo Jaffee
Interstellar is great. My absolute favorite song is Once Upon a in Hollywood.
Luke Drago
It's good, it's good. I don't think it's my favorite. I think my favorite song in the album is probably Paul Rodriguez or Mission Control. And I think the only mistake for the album for me is that by starting with those two songs, I feel like every other song after that was like a bit of a letdown relative to where the expectations were set. Where I think my favorite song in College Park is Lightsaber.
probably seconded by, let me put my actually good logic playlist real quick. yeah, it's Lightsaber probably followed by Shimmy and like those tracks are pretty far apart so I the album kind of like keeps rewarding you as you go through it.
Theo Jaffee
My favorite College Park songs are Self-Medication, Paradise 2, Wake Up, and Lightsabers, I think.
Luke Drago
Wake Up's good. Lightsabers is, think, actually the reason that College Park makes top 10 for me, because I think Lightsabers is maybe in my top 10 all time. I've got this top 50 all time songs, and Lightsabers is right on there.
The initial beat's excellent. The beat switch is just completely surprising, but it's still like, I think a lot of beat switches aren't thematically relevant. Like, sicko mode three beat switches do not sound anything like each other and could just be three separate songs you like strung together. Whereas, Lightsaber is I think better because of the beat switch. I think they're like very clearly related and they sound really good together and I think the song is better because they have both parts in it. I don't know, think, especially when like, is it C.Castro comes in on Lightsabers at the end, it delivers like just an incredible verse there too. I don't know,
Lightsabers are just. That's one hell of a song.
Theo Jaffee
Yeah. Like, Castro's verse made me think like, why isn't Castro and more logic stuff? I know they had like a falling out and now they're friends again. I hope Castro comes back more. Yeah.
Luke Drago
Well had that mixtape recently, right? I think a couple of, like the EP. What was on that? I really... Whose is?
Theo Jaffee
Castor's voice is better than Logic's. Castor's voice is better than Logic's.
Luke Drago
it's a good force. I think Game 6, I think, is on that EP. And I think Game 6 is a Sleptone song. I the beat's good, I Halfbreed kills it, Cedar Castro kills it, like everybody on there does a good job.
Theo Jaffee
Yeah, you really know Logic. I think the Seth MacFarlane feature on Self-Medication is like, I was so not expecting that when I listened to the album. I was like, for a sec, I was like, is this Frank Sinatra? Who is this? my God, looking at the credits, it's Seth MacFarlane, wow. Yeah. He can really, really sing. He was like trained to be like a singer by like, I forgot who, but like some celebrity that had something to do with Sinatra.
Luke Drago
It was so good.
Luke Drago
This is a family guy guy, yeah. I was pretty floored by that.
Luke Drago
Yeah, I was pretty floored by that. I think I showed it to a couple friends. I'm like, you should hear this, because you're not going to believe who it is. This is like the voice of Peter Griffin and Stewie Griffin. I think he's also Stewie. Yeah, and all these other characters that you've grown up with is now singing some crazy Frank Sinatra. I don't know. I was pretty impressed with that. What was your least favorite Logic song or album? And why is it Confessions of a Dangerous Mind?
Theo Jaffee
and Brian.
Theo Jaffee
Yeah.
Theo Jaffee
My least favorite song or album?
Luke Drago
Yeah, either or. Or both.
Theo Jaffee
Okay, my least favorite album was Supermarket, but my least favorite actual album was probably, like sadly, Confessions of a Dangerous Mind. Even though, like, I didn't think that it was that bad. Like, there were some songs that I thought were good. Homicide was great. Yeah.
Luke Drago
Homicide is pretty good. Keanu Reeves actually, like the lead singles weren't bad, I think that was the problem. The lead singles were fine. And then I remember the first time I heard Clickbait. That was a tragic day for me. I'd stayed up all night for the drop. I must have been in high school.
Theo Jaffee
Yeah.
Luke Drago
and it dropped and I remember hearing it going, my goat is washed. Like this is it. This is, like we can forgive everybody but I don't think you come back from this. I think he has come back from it but I think he probably lost most of his fan base after like the back to back code supermarket drops.
Theo Jaffee
Yeah, like this is also like right when I got into Logic was basically exactly when Kodam dropped But I kind of liked it at the time Maybe I just like I was always built to like Logic So I liked even this and then when I discovered like Under Pressure, Incredible True Story I really really liked those and then when I discovered, you know, the mixtapes I like those too. I think like the song Confessions of a Dangerous Mind
Luke Drago
Yeah.
Theo Jaffee
And also the song Lost in Translation are pretty good. They're not that bad. A lot of the other ones are... Yeah, wannabe, clickbait, like... Icy, those are all crap.
Luke Drago
I think that I did bad.
Luke Drago
I think the problem is like, I thought ICY was, I actually kinda like ICY, but I think the problem with Kodam is like the lows are just so astronomically low. Where it's like, know, the average song is actually fine. It's not anything special, but it's fine. But man, those misses are like dramatic and bad.
Theo Jaffee
What do you think those misses are other than?
Luke Drago
clickbait's the one where I like, I think I turned it off for a while, I like, this cannot be happening.
Theo Jaffee
How does that one go again? Hold on, clickbait lyrics. you are amazing, you are amazing, you are amazing. Yeah, okay. That was so...
Luke Drago
Yeah, yeah, yeah, yeah. There's some other lyrics in Clickbait that I think are maybe more shocking. Is this the one that I'm thinking of? It's got like the...
Theo Jaffee
RIP Lil Peep, let that young man sleep, let that young man death teach the youth the streets to beat addiction.
Luke Drago
Yeah, it's, there's some other ones there. don't know, that was I think, yeah, we'll leave that off the AI podcast. But it's just like, wow, this is bad. But yeah, I think I have hope. The current album is about to come out. I know that Winnie's working on, I've heard the singles.
Theo Jaffee
yeah.
Some more pornographic ones.
Theo Jaffee
haha
Yeah.
Luke Drago
I'm a little worried they're sounding kind of like Kodam, like very trap-coated, almost like a little Carti-ish, and I don't think Logic can do Carti. I don't think Carti can do Carti, so I'm not sure if Logic can do Carti. I'm a little worried about what's gonna happen there.
Theo Jaffee
Yeah.
Logic's had like five final albums too.
Luke Drago
well hey this is final out you know this week next week we knew final out the time you've got your party guy curious
Theo Jaffee
Yeah. Yeah, Sort of. did, yeah, I'm not like super, super into Carti. I definitely like him more now than I used to. Like, I listen to the entirety of IAM music. I think maybe it was sort of like colored by the fact that I was in like a very stressed mindset because I was on this like super delayed flight and whatnot. But that might have even made it an even better listening experience because it's so like intense and violent of an album.
Luke Drago
Yeah.
Luke Drago
I did too.
Luke Drago
Yeah.
Theo Jaffee
Like, I clicked play on pop out, the first song, and let it play for like five seconds. It's like grinding metal. And I was like, wow, yeah, this really sets a tone for the rest of the album.
Luke Drago
So I think I should like music, because I like Yeezus, and I think they're aesthetically similar in a couple places. I think this is also an album where I think my default pretty not in Carti's target. I liked parts a whole lot of Red, but really I ended up liking the production when I liked Carti. I think it Mojo Jojo for me. was like, oh, I just don't like this. And it's a shame, because I should like the production style. I like Kendrick a whole lot.
Theo Jaffee
Yeah.
Luke Drago
And I think I heard the first set of ad libs on Mojo Jojo, I like, never mind, let's never do this again. Let's never have, I think there's like three Kendrick Carti collabs in this album, if I remember correctly.
Theo Jaffee
So do I.
Theo Jaffee
The Kendrick hate over the last like, I know, year or so has really ramped up like to a level I've never seen it before.
Luke Drago
I think if any popular artist is just doing, like at this level of popularity, think Kendrick's currently not exactly but pretty close to where Taylor Swift was in like summer of 23, where she was just the most famous person alive. I think he's not exactly there, but he's like pretty close. And I think you saw a lot of Taylor Swift hate on there too. And I don't wanna, I'm not making a specific comment on Taylor Swift, who I think has like a lot of music that I like and some that I don't like.
Theo Jaffee
I agree.
Luke Drago
But I think that when you get that big, there's some level of envy or jealousy as well. And also all of your failures are bigger. So I think Moto Giroda was a bad song. I don't think this diminishes Tim Imp Butterfly in any way, shape, form. But I think oftentimes it's the most recent track is the one that changes all expectations. And the my goat is washed mentality just gets bigger and bigger. I don't know, man. He's selling out whole lot of stadiums. I think Kendrick's doing pretty well. I think he's doing all right. think a lot of the hate's just now he's popular, so it's cool to hate him.
One thing that fascinated me about bands as a bit of a sidetrack is just like how important like the lead driving force tends to be. I know that like Linkin Park for example just recently swapped out there like obviously like Chester Bennington died in Armitage and died.
2017, just a monumental force of music. It's hard to imagine a Linkin Park without him. But they have, as of last year, a new lead singer. She's really excellent. The Emptiness Machine was for a minute up in my top songs, I think early 2025. I don't know. think the band is definitely different. There is a pre post moment there. And I don't think it's ever going to be the same.
But I think you can still keep the driving force and the driving memories alive even as people fall in and fall out.
Theo Jaffee
Yeah, Pink Floyd did this. Like after Syd Barrett left, they had many of their great hits. I'm not sure if it was before or after Dark Side of the Moon, but Wish You Were Here was actually an album written for Syd Barrett, to Syd Barrett. It was wishing he was here. Shine On You Crazy Diamond was like, same thing. And then they made The Wall and Animals and all that was after him. So bands can survive without their
Luke Drago
Yeah.
Luke Drago
Yeah.
Theo Jaffee
Top guys.
Luke Drago
I wish I knew more about Pink Floyd. It's been on my list for a long time to really delve into both the music and the history, because obviously it's a band that's just been a monumental force in music and cultural history. It's a weird blind spot for me where I just never really sat down and just listened to everything. I don't know how I managed to do that, gone this long without... I think I've listened to Dark Side of the Moon all the way through and that's it. I'm pretty surprised that's all I've really encountered from Pink Floyd.
Theo Jaffee
Yeah, they're good.
Theo Jaffee
I like Wish You Were Here better.
Luke Drago
I will add that to the list of things you need to listen to.
Theo Jaffee
Cool. And then like, I guess the opposite of Pink Floyd was like, know, bands consisting of like lots and lots of people who have sort of like broken up and gotten back together and whatnot like many, many times and come out with many, many albums is King Crimson. Who their very first album in the quarter, the Crimson King was just like a peak, like top 10 album, I think top five, maybe it was just such a good album.
and then like
Luke Drago
What's the album? I don't think I've heard any King Crimson. What's the album? In the Court of the Crimson King. I am adding this to my... I'm like, occasionally I'm shocked where I'd find like a crazy like, adding it now. Yeah, was like a crazy deficit. I'm gonna have to listen to this.
Theo Jaffee
in the court of the Crimson King. It's like very, very famous. Yeah.
It's like a screaming guy.
Theo Jaffee
Yeah, and then like the band broke up and got back together with different people and broke up again and like nothing they've ever made after this has been anywhere near as good and it's been like 60 years now and so it's like yeah I wonder like what is it like for musicians who you know release some amazing masterpiece and then just can never replicate it ever.
Luke Drago
Well, I wonder, I have a similar-ish thought, and it's not exactly one-to-one with Gambino, where like, Chaudhish Gambino has released, like, has done a bunch of stuff. I mean, he has been a comedian, an actor, a musician, but even as a musician, he's been a rapper and a singer, and it's just very hard to figure out, exactly what he isn't good at, but I also think this, really changes album to album. So, like, I think, you know, older Gambino, obviously, like, stuff like, you know, like, stuff like, what is it?
Theo Jaffee
Mm-hmm.
Luke Drago
Stuff like camp is just very different than like...
Theo Jaffee
Redbone?
Luke Drago
very different because of the internet, very different than Kauai, very different than Awaken My Love, and then my personal favorite, can't beat an album, which oftentimes I think is a bit of a hot take, Atavista. His one he released in 2020, and then everybody forgot about it because he released it on, I think March 15th, 2020, and most of the lockdowns are between March 14th 16th, so just really awful timing, so he re-released it. Also he released it with no title, just like it was labeled 3.15.2020, and all the songs were just like the timestamp they were within the chronology of the album.
Theo Jaffee
bad timing.
Theo Jaffee
Yeah, bro is not Kendrick.
Luke Drago
re-released it in 2024. Bro is not Kendrick. But he re-released it in 2024 with new titles and updated mixing and mastering.
And I think that album is stunning all the way through and very different from his most recent album, which I think is supposed to be his final album, but what does that mean? Maybe the next one's under Donald Glover's name. I think it's the kind of thing where it's almost the opposite, where he can continually make very different hits. And the downside of this is that because they are so different every time, if you like a child's give me an album, you're like, I wish he did more stuff like this. You're just not gonna get that. He's going to just make different the next time he's on the mic.
Theo Jaffee
And then Daft Punk, you said Random Access Memories is also in your top 10, which, excellent choice. Personally, think Discovery is just slightly better, but Random Access Memories, like both of them are easily 10 out of 10 albums for me.
Luke Drago
an unbelievable album.
Luke Drago
reasonable.
Luke Drago
What is your favorite song on RAM? I'm curious.
Theo Jaffee
My god I Think like if I had to pick Favorite song like yeah, I know I'm allowed I'm allowed to pick whatever favorite I want I would pick get lucky, but I think yeah, I think like Fragments of time would also be way up there Within is great motherboard is like really really underrated
Luke Drago
You're right.
Theo Jaffee
Mother word is completely instrumental, isn't it? Yeah.
Luke Drago
I think it is, yeah.
Theo Jaffee
Yeah, like everybody who has sort of listened to little bit of Daft Punk, you know, they've heard Get Lucky, they've heard Starboy by The Weeknd, they've heard Harder But... Yeah, like, you have to go listen to Motherboard. This is like, know, deep cut. Daft Punk just doing incredibly well. Instant Crush is also fantastic.
Luke Drago
They've harder, better, faster, stronger.
Luke Drago
If you want a deep cut, my favorite song in album, by far, Touch, featuring Paul Williams. It is eight minutes long. The first two minutes are like this bizarre intro where like just noise is happening and the robots are like whispering and stuff. And then it becomes a show tune.
Theo Jaffee
Mm-hmm. Reasonable take.
Luke Drago
which ends up becoming like a ballad and canonically the song is about like a robot feeling the sensation of touch for the first time not knowing what to do about it and like being completely floored by this like very human sensation. A buddy of mine and I, two of my buddies and I were in like Italy, we flew into Venice and we're like driving across the countryside and rented a car. We blasted this song at full volume and it was just like, what a surreal experience.
Theo Jaffee
Yeah, that's gotta be amazing.
Luke Drago
Yeah, it was good fun. strongly recommend blasting anything at full volume driving through like the Italian countryside. It's not a bad place to do it.
Theo Jaffee
Yeah, I can't think of any songs on this album that aren't like great. Maybe doing it right wasn't great, but it was still like quite good. Yeah, let's see. Give Life Back to Music, great. Game of Love, great. Giorgio by Marauder. That one's one of my favorites. That one's so good. And it's like, yeah.
Luke Drago
It goes on for a while, yeah, it goes on.
Luke Drago
Great. Georgia by mortar, great.
And also nine minutes long. I think very few artists can do the like post five minutes long and still have it be like relevant and good. And I think Daft Punk can do it consistently, which speaks a lot to like their ability to keep things sonically interesting and just, you know, say I have a through line that survives throughout nine minutes of music.
Theo Jaffee
huh.
Theo Jaffee
Within, fantastic. Instant Crush, fantastic. Lose Yourself to Dance, great. Touch, no, you've already talked about how amazing. Yeah. Get Lucky, just such an amazing good song. Beyond, good but not great. Yeah. Motherboard, great, fantastic. Love that one. Fragments of Time, fantastic. Doin' It Right, good, good. Contact was...
Luke Drago
Fantastic.
Luke Drago
Yeah, I spent enough time on touch. Unreal.
Luke Drago
It's fine. Yeah.
Luke Drago
Yeah.
Theo Jaffee
Very good.
Luke Drago
Yeah, I don't know. I think it's a hard album on the top. And it's definitely the album I played the most by Daft Punk. I think going in, as your final album, deciding we're doing live instrumentation now is a pretty crazy move as a send-off. I wouldn't expect a band to survive this. They survived it with no problems. I think they are better because of it. I think the album is better because it made that bold choice.
Theo Jaffee
Theo Jaffee
Sure. Yeah, I don't know because Discovery was just so good and like it's really hard for me to pick between Discovery and Ram because they're so different and yet both so like excellent. Like Discovery, I think face to face is like maybe one of the best like examples of sampling ever in like music history. It's just sampled so incredibly well. The entire song minus the drums and the vocals is just sampled from other songs.
Luke Drago
Yeah.
Theo Jaffee
And it's so good. Too long is 10 minutes long. yet, like you mentioned, Daft Punk exceeds five minutes consistently. And it's great. Voyager is just like, my god, that song makes me feel like I'm floating through space. Yeah, there's just so much excellence on Discovery.
Luke Drago
Mm-hmm.
Luke Drago
Yep.
Luke Drago
I think the brilliance of artists like Daft Punk is they can do very different things, and you might want 100 other cuts that sound like that, but they've consistently hit 10 at whatever it is they're trying to do. And so means you get this discography of music that is just diverse and interesting, that it always sounds like them, but what they're trying to sound like is different every time. I think like...
Theo Jaffee
And then, yay.
Luke Drago
I'm trying to think of other artists that do this, think, well. I can think of artists that have tried to do this, like, yeah, Kanye's done this pretty well. Miley Cyrus, Sleeper Cut here, has also done this pretty well. Her most recent albums have all been very different. Like, Endless Summer Vacation is quite groovy. But then Plastic Hearts from 2020 is this pretty grimy rock album. And she got really into this, I mean, it's got a cover of Zombie on it.
Theo Jaffee
Kanye.
Theo Jaffee
Interesting.
Luke Drago
It's also excellent. I think she's pretty consistently, if she wants to achieve a musical target, she will smash it out of the park, even if it's quite different. I think the opposite of this is Drake, who think is very good at making a specific kind of hit. And even when he gestures out of that comfort zone, I think he...
Theo Jaffee
Yeah.
Luke Drago
Honestly, nevermind. I'm gonna defend it like a bit of a slept on album. But I think it's very cautious. And I think you can tell his heart's not all the way into it. And it's so not into it that at the end he throws in Jimmy Cook's. like, hey, if you didn't like that album, don't worry, here's Jimmy Cook's. And I think there's no boldness in...
almost doing the thing. I appreciate he went out of his comfort zone on this, but I think had he actually committed all the way through. If Jimmy Cooks is a single somewhere else, another album, had he really said, I'm gonna do this very different thing and I'm gonna excel at it, I think the album would have been better for it. think the energy would have been better as well.
Theo Jaffee
Yeah, I agree. So like what other musicians are in like your top ten?
Luke Drago
Man, a top 10 is always a hard one. What their albums?
Theo Jaffee
Albums, I guess you said yeah college college park uses and ram are all in your top ten Do you have like a written top ten or is it just kind of?
Luke Drago
think I do. Let me find this real quick. I think as a matter of fact, I do. Because I think it's hard to say something in your top 10 and then not talk about your top 10. Other albums here. Let me just pull up my list. What's in yours while I find this? Because I think I go off top of my head, but I really want to make sure I'm being honest here. There.
Theo Jaffee
I have not written a top ten. And if I tried to say one off the dome, wouldn't be very good. I will maybe write one in the future.
Luke Drago
Okay, other albums I know makes this list for me. Daytona by Pusha T. I think it's excellent all the way through. It's from like the soft surgical summer sessions. Standout tracks from that. I mean, basically it's a 30 minute album, so all of it. You've never heard Daytona. The games we play.
Theo Jaffee
Hmm. I've never heard of it.
Nope. I'll have to add it to my list.
Luke Drago
is probably like the peak of this album. The games you play in infrared, I think are like the two stand-ups, but it's a seven-track album. It's from the same session that produced like, Tiana Taylor's breakout album, and also like, same session that produced Kitzy, Ghost, and Ye. Like just like, there was a summer where like seven or eight albums like this just come out. I think what ended up becoming Atavista is on this list for me as well.
I can never decide if it's Flower Boy or Igor that's on this list. It's gotta be one of the, I can't pick both. I think that'd be ridiculous. But it's one or the other that's on this list for me.
Theo Jaffee
Yeah, I love Tyler.
Luke Drago
It's hard not to love Tyler. I bought Chromakopia merch and unfortunately the hoodie is Minecraft green, so I definitely can't wear it. I look like a creeper walker, like the physical, like the Minecraft creeper. It's like very bright green with black text. But it's a tour I probably should get tickets for. I don't think it's come to London yet, I don't think.
Theo Jaffee
Haha, yeah.
Theo Jaffee
Yeah, like, hmm. I think I like Igor slightly more. Like, the opening two tracks on Igor might be like one of the strongest, like, opening, track, like, lineups of any album I've listened to. I think it's, yeah, Igor's theme and then Earthquake. Like, Igor's theme is so, strong and leads well, like, so well into Earthquake, which is, like, the biggest hit off of Igor. let's see. Igor tracklist.
Luke Drago
Why?
Luke Drago
deal.
Theo Jaffee
Yeah.
Gone gone slash. Thank you such a good song six minutes and 15 seconds long
New Magic Wand. There's just so many good songs on here. I think Flower Boy... my god, it's really good, but it's not quite as good. See You Again is maybe Tyler's best song.
Luke Drago
Probably, I think it's such a standout win.
Theo Jaffee
Yeah, and Kali Uchisana was great.
Luke Drago
See you again as what else.
See You Again is excellent. The intro, This Flower Blooms is really, really stellar. Forward and Where This Flower Blooms, both are really stellar. I like, I oftentimes like Jaden Smith, so Pothole's really good for me. I think Jaden can be little too self-indulgent sometimes, but I think his production value's always quite good and his flow is quite good. Let think, other top albums for me that aren't rap. Rogue Cut, Chroma Copies great. Rogue Cut for me, A Fever You Can't Sweat Out by Panic! At The Disco, it's their debut album. It's got...
Theo Jaffee
forward.
Theo Jaffee
I Chromakopia.
Luke Drago
I like Sin... What's the song? Sin's Not Tragedies? Yeah, yeah, It's got Kemosada on it. Build God Then We'll Talk, I think it's fantastic. Which one? I, you know, personally I'm not working on that. What can I say? I think I'm consistently confused as to why a lot of my friends are trying to build God and put him in charge.
Theo Jaffee
Sin's not tragic. Yeah, yeah, yeah.
Theo Jaffee
You like that song, I bet. build God, then we'll talk. Building God? Yeah.
Luke Drago
But especially my atheist friend. I'm a little confused by that sometimes, but that's okay.
Theo Jaffee
How is that confusing?
Luke Drago
I think it's interesting, see I think it's interesting because I think I was on stage at some event a couple weeks ago and I like, it was an AI question and I brought in Augustine and Aquinas and I think it's confused a lot of people on stage. And I think my general take is that like there's a general pervasive notion in the AI community that...
Theo Jaffee
It would be more confusing if your religious friends were trying to build gods.
Luke Drago
you know, if you can build superintelligence, you just have to put it in charge and that you'd be like somewhat stupid to not listen to superintelligence and to let it dictate your life. And I think funnily enough, like, religious had to grapple with this question for a very long time of like, why, if God, then why free will? Why would you enable like your ability to make incorrect decisions? Augustine has a very long defense of this that I think is not exactly relevant, although I think parts of it are. But I think it's like, there are lots of thinkers who grappled a lot with the question of omnibenevolent being that still allows you to make choices in what
inherent value of those choices are. And I think sometimes those conversations are actually more relevant than we think for kind of what we're building right now. I think that it's interesting to me that like, you the late Pope Francis had actually a lot of work he did on AI. His message on like 2024 World Peace Day is I think one of the better pieces of AI ethics work that's been produced in a field that oftentimes is like mired by infighting, given that he's like the Pope, he doesn't have to care about the infighting, he just chooses not to. And so I think sometimes like,
If you're gonna be building omnibenevolent beings, if that is your goal, you should probably look to at least the thinkers who spent a lot of time, who have also grappled with the question of an omnibenevolent being. And I think sometimes that doesn't happen in this space. Robin-Himes looks as like a field that isn't relevant. Yeah, hard pivot, by the way, too. I think that just, but thank you Panic! at the Disco for bringing us here.
Theo Jaffee
This is an underrated take.
Theo Jaffee
Yeah.
True. I think most of my favorite Panic! at the Disco songs are not on this album. I think the only... Okay, favorite Panic! at the Disco songs... I'm trying to remember which of these songs are Panic! the Disco.
Luke Drago
What are your favorite ones?
Luke Drago
I also wonder how you're going to cut up for your listeners, like the AI section, the music section, and then the brief Panic! Disco side quest on the role of theology in AI. I think that'll be good, a good timestamp there. Yeah, it'll be good, it'll be good.
Theo Jaffee
Yeah, just put chapters yeah Viva Los Vengeance is really good Death of a Bachelor Death of a Bachelor was so good Death of a Bachelor is like the song that like I I guess my aspirational song for being able to like sing really well is to do like Death of a Bachelor perfectly
Luke Drago
Death of a Bachelor is quite good.
Luke Drago
Brendan Urie's vocals are just criminally good. think his cover of... god, what did he...
Bohemian Rhapsody, obviously. His cover of Bohemian Rhapsody is really also incredible. And that's a song that I think very few people in the world could cover. For example, Logic's Bohemian Trapsody, I think, does a pretty poor job of emulating its namesake.
Theo Jaffee
Mm-hmm.
Theo Jaffee
I don't think it's supposed to be a cover of Bohemian. It's a different thing. that and Can I Kick It were, I think, the best two songs on Supermarket. They were actually pretty good.
Luke Drago
No, it's not. This is another hit at Logic Supermarket.
Luke Drago
But I think that says a lot about super market.
Theo Jaffee
Yeah, the rest of it was trash.
Luke Drago
Yeah, it's bad. Another underrated album. I know a lot of people have probably not heard of Quadeca, but I think his like 2022, I Didn't Mean to Haunt You. This kid was like, he was a YouTube rapper. He was like most famous for like writing a diss track about KSI that did pretty well. I don't think he was like a particularly excellent YouTube rapper. And then basically out of nowhere, he drops this like really ethereal
extraordinarily well produced album about like death and loss and it has like two features on it which are probably the two most bizarre features to get an album with those themes Danny Brown and the Sunday Service Choir. think Thor Harris is on it as well as like the other feature there as well but like I think he's much less well known and it just all the way through I think is stunning. I can't think of a song on there that was bad, maybe not but I think I still did like that song.
Theo Jaffee
Wow.
Luke Drago
and like its highs. Fantasy World is a seven minute song from a guy that remember was doing YouTube rap before this. Should not be able to do a beautiful like seven minute ballad and just nails it. Both thematically, lyrically, completely slept on album. Couldn't recommend it enough.
Theo Jaffee
Hmm. Okay, so that's how many albums are we at now? We have College Park, Ysys, Ram, one Tyler one, Daytona, Fever, and I Didn't Mean to Haunt You, and so that leaves three more.
Luke Drago
That leaves one, two, three, Cool, do we get out of this though, the Gambino one?
Theo Jaffee
at a vista by Donald Glover. Okay, so two more. Yes.
Luke Drago
So I get two more albums, I get two more.
Luke Drago
It's hard not to pick a Frank album. It's hard not to say like one of these Frank albums needs to be there. Yeah, well, I mean, you have the mix saves too. I am gonna put Channel Orange there.
Theo Jaffee
It's easy to pick because there's only two of them.
Theo Jaffee
Hmm, channel orange over blonde.
Luke Drago
Controversial. I am in fact going to put Channel Orange over Blonde. And I think it is because while Blonde is like an excellent album all the way through, a couple of highs on Channel Orange I think do not get replicated anywhere else. I'm really thinking of Pink Matter, Bad Religion, and Pyramids. I think Blonde is excellent. I don't think anything on Blonde touches my feelings about Pink Matter, Bad Religion, and Pyramids. So I enjoy listening to that album more, but I Blonde is obviously gorgeous.
Theo Jaffee
Hmm. I think,
Theo Jaffee
That's true. I think Pink and White is like one of the greatest songs ever. Really, like I thought it was even like compared to the other songs on Blonde, I thought it was just so far above. I was like, this is really like a like a probably a top 10 top 20 like human song ever. I really like that song.
Luke Drago
Yeah, seems fair.
Luke Drago
Yeah, it's up there. I think I feel the same way about Bad Religion and Pink Matter especially. Like when the Andre 3000 feature comes in on Pink Matter, and I think he delivered, he hasn't wrapped in a while, he delivers like probably, it might be the best version of his career. And it comes out of nowhere, it flows excellently, and all the setup to it has been gorgeous as well. I mean, there hasn't been a mist there. The lyrics are astounding.
I really like Pink and White. think it is sonically pristine. I don't think it hits the same kind of lyrical quality that Bad Religion, or sorry, that Pink Matter does. Although at this point it's kind of hard to compare. You're talking about songs that are just so good it's difficult to make a comparison between them. So I get one more. And it's gonna have to be a Kendrick album. It just kind of has to be. And it's probably not gonna be Mr. Morale or G &X. Although, Mr. Morale has a special place in my heart.
Theo Jaffee
Yeah, that makes sense. And you get one more album.
Theo Jaffee
Yeah, no. GNX was not even... I listened to GNX for the first time. I remember where I was, Depot Park, Gainesville, Florida, on a like perfect beautiful day, you know, like 68 and sunny. And I was just like, oh man, like yeah, there's some good songs on here. I was like, what is this like mustard thing?
Luke Drago
What is your favorite song on GNX? I'm push you on this. What's your favorite song on GNX?
Theo Jaffee
Hmm. right, let me look at the track list. Squabble Up is like, I did not really like that one. TV Off, I did not. this song is on GNX. Yeah, there's like one song on GNX I thought was way better than every other one, which is Heart Part 6. It's Heart Part 6. Yeah. I was like, what is this doing on this album?
Luke Drago
I have like just an immediate hit on this.
Luke Drago
Let's go.
Luke Drago
And it is? It's Heart Part 6. Heart Part 6 is fantastic. Yeah, that is my favorite song on the album. think... It also is not thematically the same as the rest of the songs on the album. It doesn't sound like the rest of the songs.
Theo Jaffee
Yeah.
Luke Drago
I like a lot of this, I like Man of the Garden, I like Luther a lot, I like Reincarnated a lot, but Hart Part 6 for me just stands out, I think that's right. favorite, so we're back, we're down to the big three. Pimp a Butterfly.
Theo Jaffee
You know what my, like, I think most underrated Kendrick song ever is? Duckworth. Which is a song on Damn. The sample that they use is like this Yugoslavian band called September. And they have a song called Ostavi Trag which is like, really, really like hauntingly beautiful vocals that they use. And the story of the song is talking about these two characters named Anthony and Ducky
Luke Drago
Duckworth is a I mean, I'm f-
Luke Drago
Really.
Theo Jaffee
And Anthony is like this gang banger, violent criminal type guy. And Ducky is just like a guy trying to live in the hood and trying to get by. He works at a chicken restaurant. then Ducky sees Anthony coming to his restaurant and he decides to find favor with him. He gives him extra chicken and extra biscuits. And then Anthony robs
the restaurant the ducky works at and decides to not kill him because he had been nice to him. And then like the bars at the end of the song, you know, I'm not going to spoil it for my listeners. You have to like listen to the song and
Luke Drago
No, I think you've got to spoil it for the review. I think you're going to have to.
Theo Jaffee
Okay, he says, mm-hmm.
Luke Drago
And the reveal there is that like Ducky is Kendrick's father and Anthony is his record, the person who runs his label. Top Dog Entertainment. that, it's just unreal, the reveal. I get goosebumps every time I hear that line. I think I'm gonna go with Damn. I'm gonna, yeah, I'm gonna go with Damn. I think that is my favorite Kendrick album. It's a, that is a, I get why. I think the themes on Damn, wrestling with this like.
Theo Jaffee
I got those goosebumps every time.
Theo Jaffee
I like good Kid Med City and TPAT more. But it's up there.
Luke Drago
jaded sense of religiosity combined with like the anger after like after the state of the country and an election. I don't know. I think it is just the most personally resonant as someone who is religious. I think I've thought a lot about these themes. And I don't know. I think its highs are really high. But nothing in the top three is bad. I mean obviously like these albums are just...
Theo Jaffee
You can also...
Luke Drago
exceptional pieces of art that are hard to compare against each other. And I think Untitled on Master is unfairly slept on. Obviously it's like an Untitled album, but I think that like... Is Untitled 03 the one with CeeLo Green? Is that right? No, it's Untitled 06. It's Untitled 06. Pardon me?
Theo Jaffee
It's very good.
Theo Jaffee
I don't know, like that's the problem with making your album untitled. That's the problem with making your album untitled. You have to remember, like what was it, untitled song number three or was it untitled song number six?
Luke Drago
It is.
Untitled 06 is one of my favorite songs of all time. Cee La Green's in it out of nowhere. I really like Untitled 06. think Untitled is, Untitled the Master is my favorite production of any Kendrick Owens. It's mixed between like T-Bops, jazziness, and a Pimp Butterflies, like very leaning into modern rap that I think it does very well.
Theo Jaffee
You can also tell that like, Logic steals a lot from Kendrick. Like, he takes so much. Under Pressure is good Kid Mad City. It is, yeah. And everybody is to pimp a butterfly. But not as good, of course.
Luke Drago
Yes, Under Pressure is in fact good, Kid Mad City. But you know, it's like, it's a good, but it's a good copy, so it's fine.
Luke Drago
See, I think it didn't stick the landing, so I didn't notice it, but even sonically.
like under the peak of the albums under pressure right like albums under pressure song is under pressure under pressure it includes like the same structure as sing about me i'm dying of thirst it just puts like the aggressive part at the beginning instead of the end but it even goes to like multiple letters written to different people and then a letter from logic's perspective and that literally is the exact same form that sing about me i'm dying of thirst takes it's three letters two from other people one from kendrick then a switch and like a diff like a kind of a moral storytelling at the end whereas this one is like the moral
Theo Jaffee
Mm-hmm.
Luke Drago
storytelling like the violence at the beginning, then three letters, the last one from logic's perspective, interspersed with voice messages, which is literally exactly how that same song flows.
Theo Jaffee
Under Pressure is like easy top contender for number one best Logic song. It's so good. And Under Pressure the album is maybe the only Logic album where there are zero skips. Like where every song is actually really good. I kind of like it. I like, you know, it took me like when I was listening to it, it took me about like
Luke Drago
Yeah, it's also his best Kendrick ripoff.
Luke Drago
I think Nikki’s not fantastic.
Theo Jaffee
20, 30 seconds to realize, I get it.
Luke Drago
It was the first time that logic could be a little heavy handed. I think by everybody it just really becomes a problem. But like a little corny. But yeah, I do think that's probably right. think other contenders for top logic's on. I think lightsabers like correctly has a contention here. I think Paul Rodriguez. City of Stars. Which is just like his pyramids contender I think basically, but that's fine.
Theo Jaffee
A corny. Yeah.
Let's see.
Theo Jaffee
Paul Rodriguez, I would put up there.
City of Stars is excellent.
I actually really like, on the Incredible True story, I really like Fade Away, I really like... What's the fourth one called?
Luke Drago
fade away, stainless. I don't think it's, I woe.
Theo Jaffee
Incredible true story tracklist like whoa yeah, that was really good
Luke Drago
Yeah, like what was really nice. I like stainless a lot. Stainless is up there for my like top. I also think...
Theo Jaffee
Young Jesus's Logic's Best Music Video.
Luke Drago
That seems right. Till the End, think is a really good logic song as well. It's like a, I think it's the final song on Under Pressure.
Theo Jaffee
That one's not under pressure. Yeah.
Luke Drago
At least in the main version and I think that's like an excellent outro Heat 6 is such a good producer and you can really tell when like they you know You can tell when he knows he has it because like logic will also get more excited on the beat and they're like Oh, I just know you need to this beat Like I think I confess is like this as well where the beat on confess is just so good that like Logic gets better because of it and killer Mike shows up at the beats way I don't know the whole thing is just really well done
Theo Jaffee
Mm-hmm.
Theo Jaffee
Yeah, I'm impressed by like how many of these songs you can specifically recall off the dome and like the characteristics of each one that you can talk about.
Luke Drago
I have somewhere that like, one of my like weird hidden talents is that like, I've heard a song and I liked it, I have basically the whole thing memorized.
Theo Jaffee
Yeah, so do I, but like, you know, you can take it a step further and say like, you know, there's a beat switch here. I think I need to like read more music theory terms, I guess.
Luke Drago
I, for what it's worth, have never actually formally engaged in music theory. I, for a couple of years, took piano lessons, so maybe that's the peak of it, but I was not particularly good at sight reading. Mostly because I just really preferred to memorize the entirety of the song in one go, as opposed to reading it every time. It was better for me if I could close my eyes and do the whole thing, which is not a good evolutionary pressure if you're trying to get really good at sight reading. So just play the song and then just try to mimic it right there, and then we'll just go from there.
Theo Jaffee
Yeah, very fair.
Luke Drago
My piano teacher was frustrated by this because the problem is it worked. At least there's a certain point I actually could keep competitively, I could play and keep up with level difficulty while never actually learning how to read it because I just listened to it and maybe watched someone play it once or twice and then I would just do it until I had it memorized. Maybe that's indicative of how I listen to music as well.
Theo Jaffee
Well, I should probably get going pretty soon, but this was a great episode. Thank you for coming on the show.
Luke Drago
Yeah, this is great. had a blast. This is, like I said, my first podcast. I hope every podcast covers this much of everything.
Theo Jaffee
Yeah, was great to have you.
Share this post