| It was fine. There's a number of related projects that are all incredibly minor who had similar demos and somehow were also managed by similar personality types. I've worked with many people over the years but somehow they get their own special group. Sometimes you run into them at SigGraph, they're these emperors where their kingdom is their mind. They have those messy sprawling websites that read like Doc Brown's scrapbook. It's an archetype I feel certainly adjacent to and fearful of. Regardless, about 6 months ago I left my job trying to catch up on all the AI craze because I think that might be the missing piece in moving this class of projects to the next step. It's going slowly, motivation is hard and this is still kind of a moonshot. The information organization required to make these memex inspired thought navigational systems truly useful was fleeting, subjective, and labor intensive. AI can do that, pretty well actually, and in personal, subjective ways. VR and AR can as well and I explored that enough to conclude it's too complicated. You can certainly express the dimensionality needed and tune things accordingly but it's too complicated to be useful. There's a cognitive limit on the amount of dimensionality and complexity that most people can handle. There are certainly some brilliant people who don't seem bound by these limits but that's not the point here. It's about taking the information that usually only brilliant people have access to and expanding that so that merely average people like myself can gain competence in it as well. Chat bots are fine but that's not going to get you from 0 to say, abstract algebras, modern quantum physics, or field theory, which I strongly believe should be accessible to say, 40-60% of people motivated enough to learn it and I strongly believe it currently is not because of the cognitive limits I expressed above by the traditional linear instruction methods. Getting to the next paradigm is something more people should be working on Turning the Internet into a true learning machine needs more work and hopefully AI can help |