Substack and Twitter

Speaker

I've got sunshine on a cloudy day When it's cold outside, I've got the month of

Speaker

April the 14th, 2023, and it's that time again. That was the week with my old friend Keith Teer, reviewing the week in technology, the week in Silicon Valley. And we have some major news to break, Keith. We're not leading today with AI, unlike every other podcast in the world. So is there a bigger story than AI this week? Well, I chose one. It was Substack launching notes and Twitter's response, which I think is, it's like a David and Goliath battle, isn't it? Like between Man United and Seville? Ooh, well, that was a tie. 2-2. So you led with Substack and Twitter, Substack being the David, Twitter the Goliath. What's going on? And of course, it's all around the launch of notes. I, like everybody else, launched on notes this week. What's the big deal? So the big deal really is that Substack is trying to take on the internet business model, which of course is advertising mainly, especially consumer internet. And it's starting with the proposition that advertising is bad for all of us. And in order to pursue that, of course, it has to build a subscription base. They call it the subscription network. And a network needs a network effect. If a network doesn't have a network effect, then you stay very lonely. So they want to build a Twitter-like stream of feed of everybody's thoughts, short form mainly, to allow everyone to find each other, like or dislike, engage and grow audience without advertising. That's basically what they launched this week. The product's called Substack Notes. You don't retweet, you restack. Which I thought was pretty cool. But they're certainly not the first, Substack's not the first company platform to come along with a Twitter wannabe. Why are we leading with this and not Megadon and all these other ones? Megadon? Oh, I know what you mean. What's it called? Megadon or Megadon? Am I confusing my Trump with my Twitter? It's worse. It's Mastodon. I didn't want to make any vulgar jokes. Exactly. Anyway, the reason it's interesting, I think, is because of Substack's core user base is thoughtful, creative publishers. Like you and I, Keith. Or at least like you. Exactly like you and I and the people who read them. So the core user base is kind of highly desirable, I would say. And so it's got a pretty decent starting point. The second is, there's already money flowing on the Substack ecosystem without ads. And I think that this will compound that. For example, I've had a bunch of new subscribers over the last 25 hours because I publish stuff on notes. So there's an internal, if you look, marketplace, I think it would be a good way. So the network effect is kicking in. Yeah. So I don't think it threatens Twitter, by the way. And I said that in my editorial. I think Twitter's in a different league, doing a completely different set of things. The only thing in common is the user interface. And we discussed earlier, before actually we went live, Twitter's announcements this week. It sounded, I said to you this off camera, that Twitter and Substack are both shooting for the same space, somewhere between them. They're both moderating themselves. Yeah, Twitter announced, I think it just was announced today. So it's probably news to many people. If you have the blue checkmark, you can now go to the Twitter page, and you can now post tweets up to 10,000. I think it's 10,000 words, but it might be 10,000 characters. But anyway, very lengthy. You can also sign up to subscription tweets. That means people will pay to see your tweets, optionally. And you'll start publishing things. And who's going to pay? It depends what the content is. I mean, I can imagine using Twitter for signal rank, for example. We predict companies that are going to do well every week. If we were to publish those lists privately to people who are paying, that would be a decent use of... But what's to stop people cutting and pasting those listings and putting them up on their own free Twitter pages? Probably nothing, at the end of the day. Good point. So we won't do that. So is Substack, you suggested that they're not going to become Twitter. What is Substack going to become? In the best case scenario, when Hamish and the others sit down and think, what do we want to be when we grow up in three or five years? Where can they go? Which is credible and serious. Yeah, good question. I mean, I discussed this with Steve Gilmore a lot privately as well. Gilmore Gang is a big user of Substack. I think that, you know, the real answer depends on the Substack leadership team. But I'll tell you what I think they should... What would you do if you were running it? What I would do is, I would recognize that Substack is redefining the relationship between people who produce stuff, the audience, and the payment mechanisms for the content. And so I would encourage them, firstly, to increase the different types of content that they support. I think live video and live audio are obvious content types for Substack. I would encourage... So in other words, they should just buy Restream? They should certainly replicate Restream's functionality, I think. I think the second is they've got to recognize that Twitter and Facebook and LinkedIn are their friend. I actually made a mistake in my editorial. I said they should allow... You never make mistakes, Keith. Well, I actually did this week. I said they should allow notes to be forwarded to Twitter and Facebook and LinkedIn. In the case of Twitter and Facebook, they do. It was buried underneath the menu, and I found it. So they already do that. But being aware that when you're in the content business, distribution is very important, and not wanting to lock yourself in to a closed environment, but using your closed environment as a launch point to distribute into all the other environments, I think super important. So more different types of more distribution, and then more ability to refer to each other to embed and comment on content. My biggest problem with Substack in the way it's evolving, if it wants to become like Twitter, is that I'm not going to pay for other people's subscriptions. But I'm interested in what all these people are saying. My sense is they got to make their bets one way or the other. Either they stick to their original idea of a subscription platform, and just focus on that. Or go for the big play, go for the home run and try to replace Twitter, or at least get ahead of Twitter. But they can't do both. I think they're going to do the first. I don't think they really want to replace Twitter. What they really are closest to, and I've learned this again from discussions with Steve, is because he's coming from a trade press background, they're closest to the trade press, where there's lots of very specialist publications. In that world, sponsorship is a large revenue stream, which is not the same as advertising. Closer to the way podcasts work, the podcasts often have a single sponsor. So I think there is more money that can come into the ecosystem, as long as they're not problematic. I do think the three to five year game is people like you and me, probably can make a living by doing what we love to do anyway, which is publishing stuff that is of interest to an audience. It's interesting that I'm working now with Zencast. Maybe we should do a show because you tend to, I don't quite know why, but you have a very soft spot for Substack, or maybe a very hard spot for Substack. There are other interesting players out there. Zencaster, for example, who are trying to, in their language, democratize podcasting and enable podcasters to earn a decent living without being superstars. So both Zencaster and Substack are in the business of empowering a new creative middle class rather than just superstars. But the problem with Substack is the real winners are Andrew Sullivan's and a handful of other people. They paid large amounts of money to be on their platform. So in a way, Substack is actually compounding the crisis of the creative middle class. Well, I think if you look at their most recent numbers, which they had to publish last week because of raising money on a crowdsourcing platform, they've weaned themselves off those payments. Those payments were time limited to 12 months. And they're now self-sustaining, I believe, or very close to it. But by the way, I don't think podcasts, I think if you look in the rear view mirror, podcasts are going to end up being a stepping stone to what we're doing here. Well, we're always ahead, Keith. There's no rear view mirror for you and I. We're about Musk and Twitter. You don't put many links, but he gave an interview to the BBC in which even he acknowledged he'd shot himself in the foot multiple times this week. How is he enabling companies like Substack and maybe Zencastr and lots of other startup platforms to eat into the Twitter business? I well, I'll answer the direct question, but I'll caveat by saying I don't think anything he's doing is damaging Twitter. I disagree with the commentators who say so. I do think he's probably doing the opposite, which is making Twitter more visible. But what he did this week is he put labels on the NPR and PBA and BBC sites claiming that they were state media, which obviously is just a teenage prank, which is what Musk really is. Well, you say obviously to you, you're part of that world, but for most people it isn't. It's the sort of thing that Trump would do. It's very political. It could be perceived as political. What do you mean it could be? It is political. I don't think Musk has got a political bone in his body. I think he's more like private eye for those British people who understand private eye. It's making fun in a political context. But he has, you know, he makes fun always of the NPRs and the BBCs of the world. He makes fun of progressive. He never goes after the Trumpists. So I don't agree anyway. But it can't be good for Twitter to be essentially alienating major news companies like the BBC and NPR. You know, I think it is because I don't think he respects them. And I think... Well, no, he doesn't respect them. He might enjoy it, but it doesn't actually benefit the business of Twitter. If it wants to be a platform for information and news. I'll bet you that Twitter's usage numbers went up after this knockdown. You think he does these public pranks to get attention, which then convert into attention? I wonder if, ironically enough, the platforms most under threat from this rightward shift in Twitter are platforms like Parler, which were created in response to the NPR style control of Twitter is now being replaced by Musk and his dog. Yeah, why do you need those? You don't need Parler anymore. You got Twitter. Anyway, enough Twitter, enough substat Keith for now. Let's talk about the real news of the week. You may not acknowledge it, but AI is still around. And in fact, you even have a whole section, you have video of the week about AI, Stephen Wolfram, and then you still have a new section in the newsletter, AI of the week. So Stephen Wolfram is one of the smartest guys around. What was his video? It was a very long video. It's over an hour long, and it's wide ranging, but it's the discussion of both the technical basis of AI and stepping over into some of the policy issues that AI gives rise to. So it'd be pretty hard to summarise it here. I think it's definitely a Sunday afternoon, you know, if you've got... Is it state funded? State funded video? Not a state funded video. I've got an hour on Sunday afternoon, everyone go and watch Stephen Wolfram. I've always liked him. I've always a very, very smart man. I thought his written piece on how GPT works is very good. And then what's the AI news of the week? Is it mostly about regulation, debates on regulating it? Very few debates on regulating it, although there was some government... You're a big regulator. You're in favour of regulation. You like a strong state with tech, don't you, Keith? You know exactly the opposite, as you know, Andrew. But funnily enough, we'll talk about it next week. But Tim O'Reilly, they published a piece this morning saying you can't regulate what you don't understand. And I think O'Reilly generally... Did Tim do that or his company? You know, I can't remember who wrote it, but Tim commented on it. So he's definitely involved in it. He probably wrote it as well. I can't remember. He's a very wise man. One of the few really genuinely wise men who you're not quite sure what they're going to say. So what does it mean you can't regulate what you don't understand? It means be super careful not to allow pure fear to drive innovation, you know, being slowed down. The Chinese government this week did try to regulate AI by saying that AI needs to be biased in favour of the state. And it said that no AI will be allowed to be critical of the state. Sounds like the BBC or NPR. Right. So the minute... What do you mean, right? That was a joke. Yeah, but what's clearly obvious is that authorities all over the world will seek to, at some point, regulate AI. China's leading the way this week. And, you know, it's going to be interesting how people calling for regulation respond to the Chinese actions because they should... Not all regulation is Chinese. I mean, it's like arguing when you invented the motor car. Well, we have no idea where it's going. So we're not going to regulate it. There was always a need for regulation. And you...regulation is an evolutionary process. It's not a revolutionary one. I think it's all about whether it's driven by intelligence or fear. I don't think you can have fear. Right now we have fear driven. I mean, if you read some of the stuff that is claimed can happen. Yeah, I sent you the FT essay, which I thought was...it was long. Again, another Sunday afternoon read. Yeah, very long and very well written. One of the better arguments in favour of regulation. Written by a British entrepreneur. Now, that said, I don't think anyone could reasonably argue that there doesn't need to be rules. So the only real question is where do the rules come from? Yeah, I, as you know, I don't usually agree on the regulation side with you. But I do get the sense that there is sort of this simultaneous hysterical reaction, both in terms of investment and a reaction against and fear. And I'm not entirely sure whether either are justified. Since no one exactly can tell me why this is such a big deal. Yeah. Did you have you noticed this thing this week called Auto GPT? No. So Auto GPT is a piece of software that solves the problem that right now you have to write prompts to GPT for it to give you answers. And what Auto GPT is, is a prompt writer. So you say to Auto GPT, I want to go to South Africa at Christmas with my family to Cape Town. Can you please organize a two week itinerary, tell me where to stay, do a budget. And if I if I then approve it, go and book the flights and the hotels. Auto GPT does all of that. It speaks to chat GPT, giving it prompts through a process that it figures out. And so you only have to kickstart it and then it does everything for you. A good example would be this week I wrote, I asked chat GPT to write a Python program to query the SignalRank database, which I don't have the skills to do, by the way, I can code but not in Python. You have I thought you were one of the great coders in the world, Keith. Yes, indeed. Just like I'm the best lover in the world. You are the Sir Alex Ferguson of coding. Yeah. So anyway, gave me a Python program. It didn't quite work. So I said, Can you fix this? It didn't work. He said, Oh, yes, sorry, that was a mistake. Here's a fix. And it probably took me 30 minutes and a few prompts. And by the end, I could run this program. And he gave me back an Excel file with the results in. Well, this is something that should worry us or not. You've just warned us against regulation and not getting scared. You're going to make all the programmers in the world unemployed. Well, well, at least for in my way of thinking, unemployment is a good thing. Well, that's a subject we're not going to, you're not going to drag me down that rabbit hole. Let's talk about some other essays of the week. You've talked about the prohibition of AI and subvert state power in China. A couple of other interesting pieces of news. The end of HBO Max and the rise of Max. Is that is that something that might challenge Netflix's dominance? Can we take this HBO Max thing seriously? It's interesting that the the conversation around HBO Max is really a conversation about the new CEO of Discovery and HBO and all of the assets that sit underneath that franchise, including Cinemax and others. And they they're rebranding and dropping the HBO brand. And most of the discussion is about how stupid that is, that the HBO brand is considered to be one of the better brands. And they're going to replace it with Max. I actually disagree with the critics. I think HBO, only Americans think of that brand the way that we think of it here in America. Globally, it's, it's a weird name. So I think to drop it and to rebrand around, you know, a word everybody can relate to, and then to drive content, probably five years from now, it looked like an okay decision. I thought HBO was state funded, but that's another issue. Anything else? We got Gary Tan. He pissed people off. What's Gary Tan who's running now Y Combinator? How's he managed to upset people this week? I think Gary Tan took a very good decision that's been badly received. The decision he took was to close down Y Combinator's fund called the Continuity Fund. This is a fund that did late stage investments, the opposite of what Y Combinator does. And reading between the lines, the fund didn't make money. So he's closed it down. This is close to my heart, because I believe that there's a phenomenon called opportunity funds. Most venture capitalists have, not most, but many have an opportunity fund, which is extra money to invest in their best companies. And almost none of those opportunity funds do well, because it's kind of obvious that you're going to invest in your bad companies that are kind of okay, as well as your best ones, and you're going to make bad decisions. So they tend to lose money. I think money for the next stage shouldn't be sitting in the pockets of the people who did the early stages. It doesn't work well. You raised money this week, Keith, but nobody else seems to be doing well. You got a couple, one of the essays of the week's Crypto VC fund, close to a crawl. I'm amazed that it's even a crawl. And then another piece in News of the Week about European VC funding dropping 66% as seeds take a hit. Things are pretty bleak on the investment front with the exception of SignalRank. Well, SignalRank's a modest early stage company. So we're not really a weather vane yet. But I think... Aren't you the HBO of Silicon Valley? I hope not, because HBO's going away. Maybe we're the max. You're the max. You're the max of Silicon Valley. You're the max. Yeah, no, look, all of the statistics about VC funding point in the direction of it shrinking, which isn't always a bad thing, by the way. That can be a sign of quality as well, which I do believe is happening. We're probably at the end of that. We talked last week about hitting the bottom. I think we probably have. I can't. And I'm not going to make any more jokes about hitting the bottom. Even on a Friday afternoon. Did you show Janaye, your wife, that? I did. She thought it was very... Is she still talking to me? She is. She thought it was very funny. But she's, of course, a certain generation that finds that kind of humour still funny. Well, Benny Hill is alive. But the European thing, the decline in funding, all of that is true. You know, I published a newsletter yesterday about Series B rounds. And in the first part of this year, there's been 399 Series B rounds compared to 768 in the same time last year. That's about a 50% drop here in Silicon Valley as well. So it's global and it's real. How much of this is just the cull? There was so much dumb money. There's always dumb money in the good times. All it is, is just getting rid of the dumb money. And the smart guys continue to make money. Yeah, I think that's true. I think that's true. And the really quality investors are, you know, they're not fully in a wait and see mode. For example, Level One, which is a crypto play, raised a lot of money in the last company. But when I look at the number of companies our AI predicts will have good outcomes, out of those 399, it was 19. So that is a sense of the quality within the quantity. 19 out of 399 is under 5%. So a lot of money is on the sidelines. It will get deployed, but it will get deployed into good companies. And that probably means this is the best time ever to be investing in venture. Way better than the craziness of the last two years, when prices were high and everything got funded. And one of the smarter guys, who is friends with both of us is John Borthwick, who runs Betaworks. They are your startup of the week, Keith. What's Borthwick up to? Last time I heard from him a couple of weeks ago, he was in Japan. Well, John is a serial entrepreneur over many years. I first met him in my startup, Siberia Cafe, in 1994. He's kind of an East Coast Anglo kind of a guy. And he's got some French in him too. And some French. And he's had a great career. He always focuses in on what's interesting in the moment. And Betaworks has a kind of a large building in New York, which is a startup club, if you will. Yeah, I think he shut that down, but he's still got the building. Right. So what they're doing is an AI startup camp. Basically, they're going to put half a million dollars with some partners into 10 companies, they'll each get half a million. But the companies will be chosen through a kind of a ongoing camp where they talk about application of AI and development of AI. And I think it is a true startup where a few weeks from now, there will be 10 companies with half a million each that come out of that. I'm going to guess knowing John that they'll all be doing interesting things. But how's that different from the startup competitions that companies like Tech Crunch run? Well, this is more hands on mentoring than judging. There is obviously some judgment in putting money to work. But this is I think this is more building rather than judging. You know, there'll be some crossover between the two. But John's a builder not he's not a Yeah, and even the term building is something he's used. Is everyone now jumping into this AI space, Keith, when it comes to early stage stuff? I mean, they're gonna have to because I don't know if you can imagine anything. There's a letter from Samil Shah in this week's newsletter where he tells his portfolio companies that they're going to be irrelevant unless they embrace AI, no matter what they do, no matter what we need to embrace AI on that was the week. Well, I think we we do we do not because we're not building a product or an application. But if we were, we would. I think we could put some AI on top of the newsletter. And, you know, basically allow it to be searchable on topics and such that I'd rather put it on the bottom, Keith. In all seriousness, John, in his announcement suggested that this was the biggest development in his lifetime when it comes to tech. When John and John is not the kind of person just to say stuff. So he must believe it. Do you think he's right? I mean, you went through you know, you've known john since Siberia Cafe, we were all around in the 90s when the internet became big. Is this bigger than the internet itself? That's a hard call. It's certainly huge. The internet was also huge. I even think before that, the PC was huge. And they, you know, the mobile phone goes with us. And the mobile phone era was huge. So but this wheel, this electricity. Yeah, this is the next level. I mean, I just watched the all in podcast, which I think is actually very good. And on that day, Chamath Palihipitaya said,