Hey, how are you? All good. How are you? Great. It's nice to see you. We got a white t-shirt thing going on today.
Um, cool. Sorry. I'll go for it, GMA. I'll let you take it from here.
Yeah, for sure. So to give you a little bit of context, Anish, I work at a small venture capital fund in Spain, but I built a small startup before, and now I've built a little bit of a small media company, and I'm angel investing as well, spending most of my money right now investing.
And yeah, so I've built this media company that it's kind of just I started writing on Substack about startups and it just grew and now I'm impacting like I don't know, like one to three million people weekly, and yeah, mostly on LinkedIn, but also on the newsletters I have like half a million reads a month.
Wow. So, and right now like the hot topics obviously like, you know, it's AI. So, Britney sent me one of your writings and we thought it would be really cool to talk about it.
I love it. Yeah. No, that's amazing. And you know, it's funny because I was just in Spain and I loved it. And then two of my most exciting companies, all of the founders are Spanish.
Yeah. Korea and Happy Robot.
Yeah, I know them. Yeah, of course.
You know those guys? Yeah. You'll have to tell me who the third set of Spanish founders will be.
Well, there are a couple really interesting Spanish companies right now, but I'll tell you later if you want which ones.
Yeah, the GRA guys. I remember I was trying to invest in them when they were, I think they did HFO.
And they didn't even answer me. They were already like too important.
Well, if it makes you feel any better, they didn't answer us either. We sent them emails for six months. There's zero response, you know. Yeah. They're very heads down. Hey, by the way, is it GMO or Gal? I see GMO.
Yeah. Sorry. That's like the short thing, but GMO is okay. Whatever you want to pronounce, it's not so easy.
Okay. Oh, yeah. Cool. No, they're wonderful. They're really, you know, they're original thinkers. They got so much energy inside them. It's a pleasure to work with them, you know.
Yeah. And Happy Robot, I've heard a lot about them. I actually spoke with some of their go-to-market team to do some collaboration with the newsletter as well, like I think one month ago.
Oh, cool. Awesome. Yeah. Also lovely people. You know, if I keep investing in Spanish founders, I'm gonna have to have like a paella party or something, you know.
Yeah. Actually, you know, like Spain is a really small ecosystem in terms of tech, but I keep seeing more and more Spanish people going to SF and there's like, it's kind of like there's a little bit of a community kind of like building up of Spanish founders and I think it's really cool. So really excited about seeing, you know, and helping the next generation.
I love it. Great. Well, let's get into it, man. What should we talk about today?
Dude, I don't know. AI. I was reading like, so you built a consumer company before and now you invest in consumer. Obviously like you're looking at a lot of AI consumer tech and KA is consumer. And you were saying in your write-up you were talking about how like many of these apps nowadays are charging $200 per month.
So what do you think? Like I think one of the hot topics that I discuss a lot right now is OpenAI is kind of like building the AI for everything. How do you think new startups should build their products to be able to, you know, not compete with OpenAI?
Yeah. Well, okay, it's a great question. We should talk about everything, especially the price thing, because my theory for this market is that nobody has a marketing problem anymore. You can only have a product problem in this market. You know, if your competitors are charging $200 a month and raising prices, then that just means that if your thing is not working, the product value is insufficient. So, I think there are no marketing problems in this era. There are only product problems. But we'll talk about that.
If you look back to the predictions in '22, you know, the predictions were that there would only be one foundation model to rule them all. And that model would be winner-take-all and all the other foundation models would either be out of business or have to cut prices and be these value players. And instead what you've seen is that all of the different foundation model companies have just ended up pointed in different directions. You know, so Claude is very good for creative writing and code. OpenAI is probably the most horizontal consumer-facing model. Mistral does a lot of things sort of on-prem and security and sovereign AI. And Gemini is just a great overall low-cost provider. And they're all working. They're all growing. They're all raising prices. So it's interesting because it breaks the conventional wisdom that we've had in tech for so many years of winner-take-all outcomes.
So one, we're seeing it at the foundation model layer that these things are actually not competitive. They're maybe even complementary. You know, I bet if we looked at the credit card data, we'd find the sort of highest-tier subscribers of ChatGPT are more likely to be the highest-tier subscribers of Claude as well. So I think that's actually the shape of the market.
So one, I think we're living in a market where there's an opportunity for hyperspecialization. And even when you feel like, oh, I'm somewhat overlapping with the foundation models, even the foundation models themselves have not been zero-sum. So I think for startups, I would worry less about overlapping with the foundation models and more about, look, are we delivering the sort of level of value that we can deliver to the customer to cause them to pay two, three, $400 a month for our product?
Yeah, absolutely. And what spaces are you most excited about right now?
Yeah, I'm spending time in a lot of markets. So a big market where we've spent time in historically has been voice. You know, voice is incredible. It feels like this is the AI insertion point into the enterprise. And you know, companies like Happy Robot that we were discussing have done such a great job of building a really sophisticated set of software around voice. I think voice in and of itself is probably not enough, but it's a great wedge because so much of human business is intermediated this way, right, with voice and with phone calls and eventually with faces. So voice is really exciting.
I think the big thing that's changing in voice is we're going from the first gen, which was last year, where it was sort of this customer support, you know, calls of low value, and it's moving up to being the most important calls that happen in an organization. You know, and Happy Robot, which for your listeners that don't know, it's voice AI for freight brokers. So a trucker will get on the phone with the AI and we'll sort of negotiate on the cost of a load and they'll befriend them. And it's a really important job. It's essentially frontline sales for these freight brokers.
They were the first example of a company I saw that was saying, 'Hey, what's the most high-value phone call that happens in the business? Let's have the AI do that.' And I think that's the trend we're going to see. You know, in a given month or a given year, the most important phone call that happens in a business is going to be handled by the AI and all the way down. So voice has been so exciting and also, you know, we've got emotion coming into voice which matters a lot. So that's been a big area of focus.
Right now I'm spending all my free time in app generation, you know, sort of roughly the vibe coding stuff. And it's so exciting because just like the foundation model thing, I think the misunderstanding is that, oh, it's zero-sum. It's Replit versus Lovable versus Bolt versus... but the truth is all of those products are growing and I suspect they're growing in different directions. And there's many more startups that we're meeting and, you know, this latent demand that consumers have to make software. I think we all underestimated how big it is and how big it's going to be.
Yeah, absolutely. I think it's kind of like the Shopify effect. I was hearing like this chat by Tobias Lütke talking about how when he was starting to fundraise for Shopify, many businesses would tell him that there wasn't enough market because at the time there weren't a lot of small e-commerces because either you had a lot of resources or you couldn't build your technological infrastructure. And basically he opened up that market with Shopify. And I think it's kind of similar. Like a lot of people weren't able to be building applications and what these apps are doing is basically creating a new market now because they're opening up that possibility for these people. And I think that's fantastic for sure.
I love that. I love that analogy. I'd actually love to watch that episode because that's exactly right. You know, I mean, how many people want to open an online store? And it turns out a lot of people do. I also think that there's a type of app that would make no sense to invest in or build two years ago. So I'm working on a very funny app right now that watches TV and day trades based on what it sees on TV.
It's half of an experiment, but it's half of just I think it's going to be funny on Twitter because I'm going to have it watch CNBC and Home and Garden Channel and, you know, all this other silly stuff. But it's really an app that I'm building that's disposable sort of for a joke. It's taken me two days to work on it and, you know, five years ago the idea of building something that complex for a joke, I mean, who would do that, you know? So I think there are also these other types of applications potentially much more disposable that we can now build.
Yeah, I've built a couple of tools as well and I'm building now like another one. Like one of the easy ones that I built was, so I write a lot of newsletters, right? And I thought that on Reddit you find a lot of really good topics. Like there's of course like a lot of startup threads, right? And it's really good because people would ask about really important questions like how can I raise if I'm not growing and I'm like at seed, for example, you know, and a lot of experienced people answer. And so I built an AI that you go on a Reddit thread, you copy the URL and it creates a newsletter. Like basically it scans like the question and every answer and it builds an open...
I love it, like easy to use, you know, and valuable because then it's kind of like I write for startups, so I go copy it. Boom, I have a newsletter.
I love it. How did you build it?
