The Newly Scalable

A difficult truth that software companies have had to deal with for the past 20 years is that scaling personalization is incredibly difficult if not outright impossible. Those companies that figured out how to do it (or were working on problem spaces that were amenable to scalable personalization) have frankly hit it out of the park. Social media, for example, exploited social graphs to personalize the experience for every one. However, when it comes to things like building houses, helping a local coffee shop set up a mobile ordering experience, or consulting for a company, personalization hasn’t been scalable. At the end of the day, you’ll need the labor and time to support the process.

With AI, however, there is a whole new set of problems that have moved from the unscalable to the scalable. This has become really clear to me as I’ve started to use OpenClaw. As I’ve started reaching out to coffee shops to help build them apps, I can’t help but realize that a lot of this work can be simply automated. Simple agents can send the emails, a voice agent can take calls, another agent can make a Loom video showing what we do. Once the coffee shop signs, agents can spin up the app in a few hours. Now, I’ll still be pretty involved in this process especially early when I’m discovering if this is a strong market to work in. However, once you get past the idea phase and start scaling, the speed at which you can move and the small amount of labor needed is frankly astonishing.

I’ll be thinking about this more and trying to find the strongest ways to implement solutions here.