Born Again: Is Tech Back?

Speaker

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

Speaker

February the 3rd, 2023, the most important date of the year, it's my birthday, I've got sunshine on a cloudy day, but it's also a Friday and it's the end of the week. That was the week, Keith. Before we do anything, happy birthday. Thank you so much. Keith, it's hard to keep an optimist like you down. We've had a few miserable weeks and suddenly the headline is born again is tech back. Is that just you or is that the general feeling in Silicon Valley? I don't think it's the general feeling, but I think it's an emerging subset of what the feeling is that Silicon Valley is very dialectical at the moment. Some people are still suffering from the correction, let's call it. Other people are doom and gloom around interest rates and likely recession. But now there are some fresh spring blooms of people who are saying the worst is over. It's all coming back and that's driven by a bunch of stuff. I didn't really pick the headline. It picked itself because if you look at the contents of the newsletter, there's a lot of good news in the newsletter this week. Yeah, but there's also a lot of bad news. I mean, this has still been a week of announcements of layoffs. The numbers from the big tech companies weren't particularly good. The Google numbers, the Amazon numbers, their revenue is disappointing analysts. So where's the evidence that tech is back? I mean, I take your point that maybe we're not in the last days of tech, but where's the evidence that spring is here? Where's your groundhog day evidence? So it really is summed up very well in the first article in the newsletter after the editorial called The Big Tech Rebound is Underway. The writer is Alex Kantrowicz, who actually appeared with us once on That Was the Week. And it really is all to do with macroeconomics and the public markets more than it is to do with startups. And as we've discussed before on the show, there's a relationship between these things with a time delay built in. And what's happened in the public markets in the past three weeks is it's been an endless series of updates for tech stocks, so much so that Facebook is up 100% since November. So if you invested in Facebook in November, you'd have doubled your money in three months, which is... Yeah, but Facebook is an unusual... I know we're going to talk about Facebook. It's an unusual company. Its stock was down so dramatically that the fact that it's doubled is... It's still... What is it? What are its numbers today? It's still 50% below its highest price. It's not as bad as I would have thought. Yeah. However, the real numbers are that it did the best revenue ever. It's now got 2 billion weekly users despite all of the turmoil. So revenue plus user numbers are through the roof. So that's resilient. If you look at today, just to more prove your point, this is today live. This is my watch list of tech stocks. Most of them are down today. But if you pick any one of them, let's pick Asana just randomly. I actually don't know what this is going to show, but if you go out to a year, it's still pretty bad. What about, Keith, the idea that, and we've talked about this on the show, that tech has normalized, that you're never going to get these... The rockets aren't back because the rocket program has shut down with tech, and it's just another sector like energy or pharma or manufacturing or retail. Well, I think the word tech, its meaning has changed a lot. When you look at AI and you look at pharma, it's all tech, right, these days. So I think that the meaning of tech is that it's so deep in the structures of every business. Look, car manufacturing, Tesla has 25% margins, whereas Ford loses money on every car. Why? Because of tech. And so I think tech as a thing is now so pervasive. So would it be fair to say, then, that the stock market has been, and this isn't a word, I can't think of a better word, techified. So what we once took for granted about the wild swings in the tech sector, in the tech economy, now apply broadly to the entire economy? I think that's true, but I also think there's going to be new bubbles. It's in the nature of technical breakthroughs that lots of money flows very early to very promising things, not all of which works out. But AI is the current flavor of the month. I think Tesla and SpaceX represent another trend where money is still flowing into the idea that you can mine the moon for minerals is a big idea, and lots of people are focused on it, including NASA. So I think we're never going to be the end of the bubble cycle. There's always going to be a new bubble. There'll always be a new correction. In the course of that, what I think is quite rational behavior, it's often depicted as irrational, but it's super rational to put money into the next big thing, and it's super rational to overvalue it early due to the demand to be in it, and it's also super rational that it should correct. So I think this is rational behavior, not irrational behavior. So you talk about the next big thing. I think if you ask most people, the next big thing is generative AI. You have a couple of interesting essays of the week, one a kind of warning about something called generative AI content farms, which sound very ominous and are, in fact, ominous, and then a more positive piece about the growth of GPT. That is the next big thing, AI, for better or worse, Keith, right? Definitely. It's also yesterday's next big thing too because people have been talking about AI for 50 years. Yep, they definitely have. The two articles you're talking about here in Essays of the Week are Dave Karpf, who is basically making a new version of the point that you made in Cult of the Amateur, which is there's going to be a lot of crap out there. Yeah, but at least in Cult of the Amateur, I argued it was human crap. This isn't even human. This is just bot crap coming out of bots or bottoms, the bottoms of bots. Exactly. So he's right about that. The question is, is that a big enough reason to think that the whole thing is rubbish? The answer is really no because there's lots of good things as well. I think what it's going to do, actually, thinking out loud here, and this is certainly what it did to human content in the Web 2.0 age, is it's going to entirely demonetize the low-ending content. It's complete commodification. But it's going to add to the value of good human writing. And then everything in between, and we've talked about this theme many times before on the show, everything in between is going to get lost. Yeah, I think that's right. I think that's entirely right. I mean, I don't know about you. I'm not tempted to use ChatGPT to write anything I write. I am tempted to use it as a research tool because it's very good at retrieving information quickly that's pertinent to any topic, including… Better than, quicker than Google? It is quicker than Google, and it's more prescient. What it does is it takes out the effort of filtering Google for you, so it gets right to the heart of what you ask for, and you can ask it to improve. When it gives an answer, you can say, well, that's good, but what about adding this thought into it? How does that look now? And it'll change it. Yeah, what does Karp say about generative AI content farms? Do they have any humans there? Are they somehow adding some element of personalization, or is it just entirely bot-created? And what's the business model? How are they actually making money out of this? They're offering services. Like, I get these emails offering copywriting services for marketing material, and what you assume is there's some entrepreneur there using ChatGPT to do work and pretending that they're doing the work. And I think over time, things like designing logos, copy editing, are obvious low-hanging fruits, but there's probably more besides. And then content farms are a little bit different. You give a content farm a concept like Hillary Clinton is a criminal, and it can produce versions of that story all day, every day. They already exist. We know they exist, although their influence is widely exaggerated. Yeah, I wonder, the impact of all this on the advertising industry is also going to be dramatic. Online advertising is very vulnerable. I mean, it's always been a very dodgy sector. It's always been – it lends itself to dishonesty in the first place, and confusion with this enormous explosion of content. It doesn't add to the eyeballs, though. That's the issue. I mean, you still need real humans, not bots, to read this stuff. So we still have the challenge of traffic. I mean, the ChatGPT or AI, it doesn't solve that problem, does it? No, and that's where Carp is actually quite good. I don't know if you remember this, but I do. Do you remember Demand Media in 2009? And, basically, it said, we'll write blog posts for you, $15 to $20 per post. And the mistaken idea was that more content is the missing ingredient. But it isn't. It's good content that's the missing ingredient. And eyeballs, basically, there'll be some marginal impact on eyeballs. The more content there is, most likely, the total number of page views goes up. But the total number of valuable page views probably goes down. Yeah, and, I mean, in the long run, it might conceivably even turn people off the internet because people are going to get so sick of garbage. It might actually be good news for the walled garden platforms, the old television networks, the Netflixes and YouTubes of the world, who at least can curate in some way. Yeah, yeah. And, look, it is hard. I mean, I do tweets of my shows and my newsletter. It is super hard to get attention. And I think it's harder for a bot than it is for a human. Yeah, it's even harder now. So in the attention economy, bots are, by definition, not attractive. No one follows. I mean, unless it's a particularly sophisticated, shall we say curated bot where there are real humans behind it and they're shaping it in an interesting way, no one's going to follow bots. People are smarter than that. So that's the bad news on AI. What about the good news? ChatGPT, according to Dan Milmo, is up to 100 million users two months after launch. But those aren't paid users. I know they've also introduced a paid tier, excusing the pun here. Yeah, they've introduced a $20 a month tier, which Bill Gurley tweeted is a steal and you should all sign up for it. And I think if you've got a daily use of ChatGPT as a kind of research assistant stroke intern, you're probably right. There probably is a lot of value in that $20 a month. 100 million users two months after launch, you've got to put that in context. That's a staggering number. I mean, every major success in the history of the internet took a long time to get 200 million users, years. So 20 million users can happen in a year. To get to 100 million in two months just tells you the... It's funny, I shared ChatGPT with my son's girlfriend at dinner the other night. She's a teaching assistant in an East Palo Alto school with all kinds of challenges that any middle school teacher would have with kids. And she asked it, how do I deal with a disruptive kid who wants to play video games in class? And it gave really good answers for strategies for her. And it blew her mind. She's now using it as a kind of an advisor. So I think the universality of this to almost anybody is so big that these numbers are going to get big and the $20 a month people are going to get big. By the way, Microsoft has said they're creating Microsoft Teams Pro version with ChatGPT built in for $10 a month. So this is normally new ideas don't monetize for a couple of years. They build users and they're monetized. This stuff is monetizable right away because there's already value worth paying for. Well, it's certainly not the first or last time we will talk on this show about ChatGPT. And one other subject we always seem to talk about, although I'm not convinced still that it's worth it, is Substack. You managed to slip that one in, Keith. News of the Week, number one story, Substack pledges. I read that and I mean, it's just another way for people to squeeze money out of readers. What's so interesting or original about that? Well, it's very tactical. So it's only really- Is that a euphemism for dumb? Not dumb, no. It's basically targeting free Substacks. Like I have a free Substack. If you go to signalrankupdate.substack.com, you'll see that there's a free Substack which is about SignalRank. And I've already gotten two pledges of people who said, if you start charging, I'll pay $100 a year for it. So it's targeting free Substacks. Oh, so it's a pledge. So it's not a commitment. People say, I will do this if you do this. Yeah, and then when you turn it on, they do it. They can cancel it, but it goes live. If you turn on paid- So it's sort of a research tool really for writers to see, to test the market. Yeah. And so I think it's a fairly good tactical idea. But the reason I put it in really was I wanted some more content other than the video of Hamish McKenzie, one of the founders of Substack, his video interview with Brent Leary, which is super good. I have a self-interest because I helped set it up. I introduced the two of them to each other. Who is Brent Leary? Brent Leary is an analyst who runs a live video platform, kind of like Now TV, but for the CRM space. And he's a very well-regarded analyst. Him and Paul Greenberg, is it? You should have a video of the week, Keith. You've got start of the week, tweet of the week, news of the week, essays of the week. What about video of the week? And then the Leary-McKenzie video could be under it. Not a bad idea, Andrew. There's a product guy in there trying to get out. Yeah. Well, you're helping that product guy get out. What did Leary and McKenzie talk about? What's so interesting about their chat? Well, the most interesting bit to me was McKenzie stated an interest in a concept called Substack Live. And Substack Live allows you to go from writing, to podcast, to video, to live audio and video with an audience that has real-time comments as such. So it's combining what was the Clubhouse and Twitter and all the others. Someone's going to come along with this at some point. All these different pieces get put together and we don't have to go to different bits. It's like Restream as well. And I interviewed the founder of Zencastr on my keynote show this week. And I pitched, I mean, basically that idea to him. I mean, they have to put this stuff together so that you and I as journalists, as video people, as chatters, we don't have to use Zencastr and Restream and Substack and Twitter. Yeah, I think you're right. It's a landing place for creators with all the tools they might need depending on their particular mode of operation at any given moment. I agree with that. And we've talked for years about how The Guardian or The Times or The New York Times or The Wall Street Journal really are not newspapers anymore. They're media houses.