Robert Ta argues that storing data won't be a competitive advantage once AI matures. Instead, the new moat is building domain-specific AI systems that deeply understand individual goals and contexts, from employee experience to health. Clarity is building infrastructure that creates digital twins of companies and people to deliver personalized intelligence.
Gary Tan says either you die a system of record or you live long enough to become a domain-specific harness. What this means for AI builders is that storing data, even in compliant ways, is no longer going to be the moat if we take the time slider down 10, 20, 30 years or, you know, to just infinity. The real problem we're seeing solved is how do you drive value with this intelligence? And intelligence is coming for your domain. And your domain-specific problems. So Gary Tan's tweet here really speaks towards everybody should be building domain-specific harnesses. And this was a retweet on Jesse Huang. I was just at a YC demo yesterday. Besides hardware, physical things, which obviously has a moat, you got to go figure out your supply chain and distribution for hardware, physical things, very hard, capital intensive. Everyone is just basically just building a domain-specific harness. We see this in legal, in health, and etc. That's a thing. So would love to know what the audience is doing in terms of your own domain-specific harnesses. And whether your company thinks about it that way, in particular, the decision makers behind budget and resources, because that's certainly the way the most disruptive edge of the industry is going. I was wondering what you thought about that in terms of our own business model, the domain-specific harness. I think that in regards to our business model, when we try to meet the goal of, hey, we're AI infrastructure for serving right context, right time, centered around the concept of a digital twin of your company and your company world's model based on its people and its humans and agents that make sense of the world, channeling that signal into— from demand signals to like, what should we build, your asset, and then going to market with that. And that would be like combining the customer world model too. So I kind of see it as if we're building digital twins to help businesses make better decisions at the micro and the macro level to improve their bottom line, essentially. I think we're in the business of helping businesses create their own domain-specific harnesses better. than their competition. That's how I see it. In regards to what is our harness, what is our moat, I think that is something we're still on the quest of discovering. The thinking really to me is, can we help with the alignment problem overall? And the alignment problem, I really like this framed recently by Alex on the Moonshots podcast, which is alignment is just a capability. Like alignment just means better capabilities, meaning the more you can understand an individual's goals, which we spoke to in the last podcast referencing Darquesh's, you know, kind of fearmongering AI civilization essay, The more you could model an individual and understand their goals, our thesis, the more aligned the next agent call is going to be towards that human's goal. And if that human's goal is to further a business direction or company direction, hitting the bottom line or furthering the product strategy and the roadmap to capture market share, that kind of thing for software product teams, then it's how do we, Clarity, build you, customer, company world model and customer world model based on our very opinionated take on subjective self-modeling, or this digital twin concept grounded in self-modeling and principles of neuroscience and computational biology. How do we help you do that faster, better, deeper than the competition? Anyway, so for us, I think the time splitter, taking that to infinity, I think it would be our— now that I'm thinking about the Jev stuff too, we then should have the best belief-specific renderings of real subjective people across as many domain spaces or industries as we can. And that would be our moat. And I think we're starting in areas of high degrees of subjectivity, such as employee experience. When am I next going to get paid? When will I get a promotion? How should my next performance review cycle affect the rest of my career? And how do I think about my career? To things like health and spiritual wellness, like love and finding love. That's another customer of ours. To things like bladder health, like how do I handle my bladder? So those are the spaces we're in initially for the audience. And Jonathan already knows this because he actively works on client work with me. But for the audience, that's how I think about our moat. And again, like, I think we're still questing and figuring it out, but directionally, some dust has sort of settled for me there.
