I’m amazed at how fast everything’s moving. I've always had a deep interest in machine learning, and felt like I have done a decent job of staying up to date over the last ~ 8 or so years. Last year, I would’ve told you AGI might be a few decades away. Now, I wouldn’t be shocked if it showed up in the next year or two.
I’m building a startup right now and finding a sustainable moat is starting to feel like a losing battle. Every time I think I’ve found an angle that’s special, a new model drops a month later, and suddenly my “secret sauce” is basically free. It’s almost comical how quickly AI is turning what used to be advanced features into commodities.
The crazy part is, this might be the most valuable year ever to start something. On one hand, there’s enormous opportunity because these AI models are capable of so, so much. On the other hand, it feels like the window might slam shut soon if AGI makes everything a commodity. I find myself vacillating between excitement—“we can build incredible things!”—and anxiety—“but will it still matter next month?” People talk a lot about how to future-proof your business or create job security. Honestly, I’m not sure there’s any guaranteed moat in the traditional sense. Maybe the moat is just being the first to build something genuinely useful, or maybe the moat, as suggestested by many others, really is just distribution. If you can capitalize on the AI breakthroughs in 2025, maybe you’ll ride that wave for a while. But the key phrase is “for a while,” because who knows how long it lasts.
Sometimes I feel like an explorer setting sail for a continent that could appear on the map next week or vanish in a month. But still—I’m sailing. Because if there’s ever a time to try, it’s when everything is in flux. The way I see it, 2025 might be the most levered year of our lives: if you build something now, it could blow up in the best possible way. And if it doesn’t, at least you’ll have front-row seats to history.
So that’s where I’m at: simultaneously laughing at how fast everything’s changing and heads-down trying to carve out something meaningful before it becomes “just another AI feature.” Maybe that’s what we call a startup story in 2025—part panic, part exhilaration, and a whole lot of all-out sprinting + hoping we’re headed to the right place at the right time.