kizen

ai is the new ui but no one has built something that truly embodies that (yet). we are still in the ibm era, and the agentic world needs a jobs-like visionary to really push for a user experience focused shift in how we use our devices. late 2026, early 2027 is when we start seeing this.

One thing that I’ve never understood about L2s and alt-L1s is the focus on technical improvements. Generally, these can be copied once invented and applied. I don’t see the moat from such innovations (from an investor perspective - obviously it’s good from a user perspective). The problem is blockspace is a commodity business. It’s kind of like in textiles or airplanes. These are brutally competitive businesses where net profits to all shareholders over the lifetime of the industry are negative for airlines (just like blockchains) It’s a problem of second-order effects. Imagine you owned a textile mill, earning some crappy substandard profit margin like 5%. You’re already earning barely more than the risk free rate, but at considerable risk and a lot of work. Now someone invents a loom that can do twice as much work as before. Great! Now you must invest in this new machinery or you will go out of business, since the productivity increases will be competed away and be captured entirely by users, not producers. Or you can invest capital to continue operating a bad business. So I’ve never understood why VCs invest in chains that not only won’t benefit durably from productivity improvements, but also fund the R&D to force everyone into the next leg of the race. It just doesn’t change the dynamics of whether your chain has a competitive advantage. Building deep financial markets seems to be the main advantage for a chain so far. But that comes with its own problems (why am I using a blockchain at all), although I feel like those are easier to navigate