Robert Ta and a co-speaker argue that staying static in AI kills startups. They identify two market wedges: B2C chat apps for wellness and enterprise HCM platforms needing personalized agents. GrokBot succeeds by solving immediate problems, not chasing frontier models. Building sticky company brains for customers creates a moat.
Do you have fears of like, if we keep doing what we're doing, in a year we'll be dead? Like, do you have those fears that drive this?
It kind of feels like doing anything for the same thing for a year right now is going to die. You know, like the world's changing so fast. So like, in what way should we adapt? And I think our overall motivation and purpose is the same. Like, so it's how do we bring the subjectivity, you know, to a business model that can scale is like, you know, really the open and evolving question I think is really interesting to keep an eye on.
That's essentially our PMF, like where we're trying to figure out PMF is where does subjective modeling of the self and then additional sensemaking to inform probabilistic confidence updates around what does this person believe as they flow through the user journey of acquisition, top of funnel in your marketing stack, all the way through to software application layer. That's been our problem so far is how do we model your user as they go through all the different flows that make up the journey of getting from they know about you, and then all the way to they're loyal to you, and LTV should go up from there. And so for me, One thing to answer this more concretely is I try, especially after listening to the latest Grokbot folks talk about their design philosophy and build philosophy, they just talk about one thing, like how did Cursor become successful even if Cloud Code came after them? And then like Codex came after them and stuff. There's one narrative which is like, are they successful because they got acquired? But it's like, all right, a lot of people would appreciate a $60 billion exit. Not a bad deal, right? Like, I think that would be quite a successful thing. Let's say they are successful. And if you use that framing of like, all right, yeah, they were able to compete against the best. They still are winning deals, as I'm aware of, in enterprise against OpenAI and Claude. And they're trying to differentiate on things now like GrokBot. So you're right that it's hypercompetitive. I also think you're right that maybe doing the same thing over and over for a year, especially with new information and landscape changing all the time, doesn't make sense. But the creators of GrokBot who have launched this wildly successful thing, and now I see Facebook Muse is coming after it too, because they're like, wait a second, personalized agents for the everyday common task makes a lot of sense. It's creating a new market. And my take on this is they had this core design philosophy that was able to disrupt something, even though they don't have the frontier intelligence. Like RockBot doesn't have the frontier intelligence that OpenAI or Anthropic do. Like those 2 teams have the best frontier intelligence, the best frontier models. But what they did have was this obsession with solving a useful problem today. And I think that's what we should focus on too. And currently the problem we are trying to solve today and successful at solving today for 2 categories of customers to be really clear, is B2C chat lifecycle, spiritual wellness, health app. That's like a form factor, smaller enterprise. And then larger enterprise, like $2 billion-ish human capital management, employee experience tech platform, where they're trying to figure out who is this person so that I can inject the right context at the right time to any of my agentic experiences to nudge them along from you were acquired in my funnel, you knew about me, you picked me to solve the problem, and now you're loyal to me. I've retained you and you'll tell other people and refer other people, increase LTV and and our bottom line. So in areas where right now it's employee experience and spiritual wellness and health, I think those are the areas that we could speak to most. And hopefully as we build out these case studies that are baking, we'll attract more prospects and customers along those lines as a market wedge and beachhead to stand on. And hopefully in a year, you know, like I'm sure my thinking will evolve in the next year, but we solve useful problems for the customers we have locked in today, get great testimonials, keep compounding that. One other way that I think about it is that's the product side of moat. Distribution ends up being a moat too, which is can we gain so many successes that reduces the cost drastically of acquisition? And then hopefully, you know, our customers that like our company brain that we provide them, they find that really sticky and they're like, oh, this is great. I don't really— I'll pay you to maintain it for me and evolve it for me. And that helps all my decision making. That makes me money. We'll see from there. But that's the general where I read the tea leaves.
