I am ready to start programming ChatGPT, the same way I have built my own code writing and deploying software on Macintosh. I want to create rules in some kind of macro language that it will never violate. I find it has huge problems with memory, it says it's remembering something, but has forgotten it 24 hours later. This is like the Fail Whale in the early days of Twitter. Cute, because the system is doing something so new, futuristic and useful, but after a while it's not cute because we're using the system for real work. The web is programmable, our operating systems are, of course the AI-o-verse will be programmable too. We are able to create entirely new development environments, these platforms deserve a fresh new look at everything. I'd also like to note that at the same time, the platforms are breaking through in web user interfaces. Remarkable progress. Far beyond what we were doing in the very stagnant Web 2.0 world. They're still stuck on whether or not our writing can have titles. So bizarre to have to exist in a world that is deliberately hobbled, and another with infinite horizons. Anyway this is what I'm thinking about just before hunkering down with my Knicks and popcorn, a Christmas tradition for many many years. Ho ho ho. #