Does Substack Deserve the Hate?

Speaker 3

A.D.

Speaker 2

hello everybody it is saturday november the 23rd 2024 it's becoming increasingly clear that we're living in a profoundly transformative moment in history especially around ai but it's not clear what kind of transformation uh we're going through in keith tears excellent uh weekly newsletter that was the week he touches on a lot of the issues associated with the future of AI. One of the essays is entitled The Tipping Point, Matthew Harris, which talks about how decentralization won an election. It seems the great theme these days is whether or not all this new technology will spur centralization or decentralization. Another of the The essays that he links to this week is by Matt Mandel of Union Square Ventures. It's called Collaborative Intelligence, and I think it speaks of whether the AI revolution will compound the centralization of big tech or actually enable more decentralization. And another interesting post that he touches on is by John Gruber, the very distinguished Apple journalist who has his own media brand, it's called Daring Fireball, in which Gruber attacks the idea of Substack, suggesting that Substack is really not decentralized media, but re-centralized media. This is the theme that Keith touches on in the editorial. Keith, interesting week. This issue of re-centralization and decentralization It's one we've been through many times before. You and I have been through cycles, the Web 1 revolution, the Web 2 revolution. There was always the promise of decentralization, but always it seems as if tech has resulted in more centralization. Is that the dominant theme of this week's newsletter? That for all the promise of the decentralized network that AI offers, a lot of people fear that it's actually going to result in more centralizations.

Speaker 1

Yeah, you frame it very well, Andrew. That's exactly the competing essays this week. Look at things in different ways. I think it's worth tearing it apart a little bit and understanding a few things. Because I actually don't think decentralization and centralization are opposites. I think they're twins, they're partners.

Speaker 1

The entire internet is built on protocols that are inherently decentralized. So when you put in a router into your home and you publish, you basically are a node, a decentralized node on the internet. And if you had a web server there, people could access it right there in your house. So the internet itself is highly decentralized. But I can't think of a single business built on it that is equally decentralized. Businesses tend to be centralized. They benefit from, in fact, massively benefit from decentralized architectures, but they themselves are almost always, especially the successful ones, are almost always centralized.

Speaker 2

Keith, let me jump in here. We've had lots of shows in the past about... the supposed Web 3 revolution, it seems a little archaic now, people are already talking about Web 4, but these decentralized corporations, which always seem more theory than practice, do you believe in those? Are there still conversations about that these days, particularly in the context of AI?

Speaker 1

There's a lot of people that try to do those, but I would argue that's decentralized with a small d. There's still some kind of a brand raising funding to build those things. So they're not really decentralized. Like working from home, which is clearly a trend, you could argue is decentralizing the companies, but it isn't. It's just a different way of working. I mean,

Speaker 2

if you're working from home for Google or Amazon, it doesn't make those companies any more decentralized.

Speaker 1

Right. So I think what we've got here is some ideological zealotry that says decentralized is better. and no evidence at all that it exists.

Speaker 2

Yeah, but you of all people, Keith, talk about that as ideological zealotry. If that's the case, you are exhibit A on the ideological zealot list. You want decentralization. You don't want a strong state. You don't want people telling us what to do. You're not a corporate type.

Speaker 1

Well, I think if you go into the realm of politics, even there I do believe in centralization. I do like freedom, that's true, but I'm not convinced that freedom and decentralization are the same thing. I think if there wasn't somebody making rules about roads or traffic lights or education curriculums, probably we'd be worse off. So I think there's a balance, there's a clear balance, and you do need some authority, some centralized authority. Democracy probably relies on that. Otherwise, what are you voting for? Now in the long run, and by the long run I mean The very long run. I mean,

Speaker 2

you're tempting me to make the Keynes joke about what happens in the long run, but I'm going to refrain, Keith.

Speaker 1

Yeah. In the long run, if we get to an abundant society where there's literally no work required because of automation and there's abundance... then the need for centralized authorities probably gets pushed out to the edge of life in your local area and global, overarching global rules probably become very thin. But I think that's the very long run.

Speaker 2

I was in New York this week. I was doing some interviews. As you know, I'm working with our friends at DLD. It's their 20th anniversary, their 20th event in January in Munich, and I'm talking to some of the old DLDs. I had a conversation with Douglas Rushkoff, my old friend in New York, this week. And he made, I thought, Rushkoff is a smart guy, a very profound remark about AI. It seems obvious, but I haven't actually heard anyone make this remark. He said that AI is the first native app of the internet. And I think that is very important in terms of what we're talking about today. I'm sure you get what he's trying to say.

Speaker 1

Well, it couldn't exist without it because you need the cloud and you need massive compute power.

Speaker 2

But it comes back to, and I'm trying to make sense of Rushkoff. He's always very five steps ahead of everybody else. So we have to try and figure out what he meant. But in terms of the Mandel piece on collaborative intelligence, AI is quite literally that. AI, the AI revolution, the LLM revolution is... a form of collaborative intelligence. It's downloading our complete intelligence and trying to make it smarter with machines, which is, it's a kind of a very sophisticated version of the original Google page rank. So I think for him, this is the first time that the internet has an opportunity to, so to speak, flex its muscles in terms of an, and Rushkoff is very much of a decentralization champion.

Speaker 1

Yeah. Funnily enough, I know Rushkoff. He's mentioned, by the way, in this week's post of the week. We'll get to it later. But I know him because of that post. And I don't like it when people fetishize technologies. What's really happened is that this is the first time humans have been able to use tools to build something that has the promise of being better than them. And so I wouldn't say it's the Internet's moment. I think it's humanity's moment.

Speaker 2

Yeah, that's where you're showing your ideological colors, Keith, because that's what you believe in. And for you, humanity is whatever you want it to be, like most people.

Speaker 1

Yeah, by the way, we have someone from London on LinkedIn saying, hi from London where it's cold and rainy. And I don't know your name, whoever you are, because it just says LinkedIn user.

Speaker 2

It was cold and rainy in New York. It's less cold and rainy in the Bay Area. Or that was cold and rainy while I was away. It cheered up when I came back, which is horrible.

Speaker 1

It poured down all day yesterday, yeah. Anyway, yeah, but it is a moment. I mean, Rushkoff isn't wrong to understand that it's a tipping point in the same way that the first essay talks about. The last two years collectively are at the beginning of a tipping point. And that tipping point is about to accelerate due to what is in the Union Square article, which is the theme we've talked about the last couple of shows, which is the rise of agents that work together on behalf of us. That's just going to mean things become possible that previously weren't. And they talk about multi-agent systems. They don't really exist yet. I've started using an app called Cursor for coding. And Cursor is built on top of ChatGPT and also Claude Anthropic. And it updates your code from a chat. into your code without you having to do copy and paste that I've previously had to do. So that's beginning to get a glimpse of the software.

Speaker 2

Yeah, it's a glimpse, but it's very cloudy. And of course, the guys at USV, my friend, Albert Vango, really brilliant guy, Fred... Fred, whatever his name is at Union Square. They're all very smart. They're always two or three steps. They're like Roshkov, except they make money out of it. Roshkov sells books. I did want to touch on this first article, the tipping point, how decentralization won an election. You and I, neither of us, I am speaking on your behalf, neither of us are great fans of the general hysteria, I think, about the Trump victory. It's more interesting than just fascism or some other nonsense. What does Matthew Harris suggest about the Trump victory, which speaks about the shifting architecture, the tipping point that you talk about? I mean, the tipping point is, of course, the title of his piece, too.

Speaker 1

Yeah. Well, he doesn't do explicitly contrast with the Harris camp, but he makes a point that Donald Trump...

Speaker 2

And he is Harris. It's Matthew Harris writing about... And Kamala Harris, of course, is not a relative.

Speaker 1

Yeah. He talks about how Donald Trump leveraged podcasts, YouTube, especially podcasts, and how he did not really need to use mainstream media that much.

Speaker 2

Which ignored him, which is astonishing, I thought, mainstream, which makes them increasingly, it sort of digs their grave in deeper for them because it makes them archaic and irrelevant.

Speaker 1

Yeah, they definitely are archaic and irrelevant. It's interesting. I subscribe to the FT and the New York Times, the Wall Street Journal, and they less and less insights are coming from them. They're kind of news fillers, but they're not really insight drivers.

Speaker 2

Yeah, and they're better than the television emics. And one of the things that astonished me about the Trump election is that on the networks, the television networks, they never... They never had Trump people. The only Republicans they ever invite on are people who are critical of Trump. I mean, it was a bizarre experience. And coming back to Harris's essay on the tipping point, how decentralization won an election, I think what captures that more than anything else is that the fact that the Kamala Harris group or people, whoever they are, they refused to go on Rogan because they didn't want to offend their campaign workers. It's one of the most, if that's true, it's not only absurd, but the most symbolic example of how disastrous a campaign she ran. She chose not to go on Rogan. She had the invitation. Whereas, of course, that's... Talking about native apps, it's Trump's native media. He loves going on Rogan. His whole life is Rogan.

Speaker 1

Yeah, well, look, Trump is a pig that likes rolling in the dirt. Harris is an elitist who brushes her teeth and doesn't go out without putting her makeup on.

Speaker 2

And uses toilet paper, Keith.

Speaker 1

And if she would have gone on Rogan, it would have been dirty for her.

Speaker 2

Yeah, it's a pigsty, but that's the nature of media.

Speaker 1

That's the world we're living in. So she basically, like Hillary Clinton before her, abhors real people and doesn't want to mix with them.

Speaker 2

Bores, B-O-R-E-S or B-O-A-R-S? The first, Andrew.

Speaker 1

In the pigsty, of course. Yeah, yeah. And so that... Now, funnily enough, Rogan, I think, is an example of re-centralization. His audience is so big that he's actually become a big... center, even though he's running it himself. I don't think Rogan fits into the decentralization.

Speaker 2

I've been covering this for years. It's the old argument. You have the undermining of mainstream media, and then you have the rise of new media, and then you get people like Rogan, who are bigger than the FT and the New York Times and the Wall Street Journal put together. He has more power than Rupert Murdoch. Why is this different? It's happened before.

Speaker 1

And it will happen again. I don't think it is different. I mean, the main point I make in my editorial is that decentralization and centralization are not... They're both false. They're static concepts. In reality, you've got constant change. Leveraging technology and the ability to gather an audience through it, you've got constant change. And anyone that's sitting on their hands thinking they could just keep doing what they used to do is going to be a big loser.

Speaker 2

So we have two... have two tendencies shall we say on the one hand the power to destroy and undermine the current power holders who tend to be centralized media of one kind or another or centralized powers but then the new powers which come which initially are champions of decentralization are rebellious uh themselves become traditional power holders i mean that's A familiar theme in history. It is.

Speaker 1

Many times before the internet.

Speaker 2

It's got to do with the internet.

Speaker 1

Yeah.

Speaker 2

If only you and me could replace CNBC. Well, I don't know. I think we're better than CNBC, which isn't saying much. Yeah. I mean, they're selling off CNBC. It's one of the ironies of Trump. He himself is trailing behind history. He says he wants to destroy mainstream media, but he's already destroyed it. He doesn't need to shut down CNBC and MSNBC. They're already irrelevant.

Speaker 1

Yeah. Did you read this week that Comcast is selling them off?

Speaker 2

Yeah, it's our moment, Keith. Now.tv, we should buy it and reinvent media. Another subject, another week. Let's deal with very viable because... How does this all connect with Substack? The headline of your editorial this week is, and does Substack deserve the hate? Less people, more AI. You were triggered into this editorial by a piece on Daring Fireball by John Gruber, which oddly enough goes after Substack of all people, the angels. Why did Gruber go after Substack?

Speaker 1

And he was inspired by another article that was attacking Substack from another well-known author, which is Anil Dash. Anil Dash published something called Don't Call It a Substack. So they were triggered by this idea that all of our newsletters are our Substacks, and they don't like Substack becoming... you know, stealing the branding of our newsletters by calling them a substack. So that was the starting point. And, you know, Gruber really goes for Substack. He kind of defends them a little bit early in his article, but then comes down on Neil Dash's side by the end. And it really is just a desire to be fully independent, which is another way of saying don't benefit from anybody's efforts where that person's brand is part of your brand.