Collection of raw notes while reading a couple of Ben’s recent posts

• https://stratechery.com/2026/apps-agents-and-aggregation/
• https://stratechery.com/2026/frontier-overhangs/

I’m seeing first-hand the transition from apps to agents as the primary interface. Last year I wanted to learn more about gun trafficking as part of my volunteer work to reduce gun violence with Everytown for Gun Safety. I was about to build https://tracer.horgan.net over the course of a weekend while brainstorming with members of that team. Today that would take a couple of hours. Before AI and agents, I would probably passed on the project as the costs were too high. The benefit side of this is still not clear as I ran into the challenge of distributing the app and engaging with community to get more feedback. While AI can reduce the costs of software development, it can’t remove every problem in the value chain. In fact it creates a new set of barriers as there is now an oversupply of “solutions” in the form of apps. I use quotes here over solutions because they are at best partial solutions. Over the last 30+ years many software professionals got used to framing a software product as the solution when it fact it was just yet another tool in the toolbox. It’s like the old saying that people don’t want to 1/4” drill, they want a 1/4” hole, or they want a picture hung or they want to relive their vacation.

One image that Ben shared is from Clayton Christensen’s The Innovator’s Solution https://books.google.com/books?id=r0xxJUzyFHYC:

The space between those lines is one of the hardest places to be in business. The top line is often driven up by insanely focused startups who are cycling through Product-Market-Fit iteration loops as quickly as possible, iterating through ideas, products, and people to find a business that fits their investment thesis. One caveat here is that established founder-led companies can act a lot of startups but are often much more dangerous as they usually have a source of capital to draw from their operating businesses.

The bottom line is often driven by incumbents who have parts of a business figured out but know they are missing the next thing. Their success hinges on trying to find a place to push the value chain apart just enough to squeeze in their modular offerings.

Rough thoughts

• Christensen’s diagram makes the world look too neat. Those lines are super squiggly and run over each other.
• Time is relative. Each company has their own sense and the marketplace is a chaos function.
• Given the chaos, company culture and incentives matter more than technical talent or other commodities.

TODO: pick up from here