|
I should say more. I think most people are going in the wrong direction. Everyone is focusing on chatbots, and I think that's too limited. Conversation is a slow, low-bandwidth interface that isn't good for most work. I'm hoping to work on tools that will help people be more productive, more self-realized. I think for a small subset of humanity (us, the programmers), that's already occurred, but for people that can't think like a computer, that hasn't happened, and so they only see computers as media consumption devices. I think a major reason that the computer revolution failed is because up until now, every application has only a single user interface. Every individual is different but with today's interfaces, individuals are forced to adapt their minds and behavior to fit the computer, making the human less important than the machine. When you look at most people and how they use software, they're working around limitations in the user interfaces. For example, when you go to a doctor, they're spending most of their time clicking through a shitty UI instead of focusing on healing people. Most non-tech jobs are like this. People waste their lives using shitty user interfaces. After that, scrolling on social media is a way to get some relief. I'm interested in working on generative UI [0], that adapts to individuals. I've worked on a few systems in the past that built some of the underpinnings of this, so I can see how it's doable now with LLMs. I'm hoping that if more people could use computers for things that they care about in a way that fits their way of thinking and working, they will be less likely to fall into manipulative media consumption / AI chatbot manipulation. If they can get into a flow state doing something meaningful, maybe they'll choose that over addiction. I know that it won't be for everyone though. I think the economics of competing with the giants is the biggest problem, though. There's no way that a single bootstrapped company can compete. The only way I can see it working is to foster an ecosystem. I'm inspired by companies like Ghost [1] and even Wordpress. Lots of businesses build off those platforms. In AI maybe it could start from a client with a plugin architecture. I think getting some sort of interop standards between local AI tooling would help as well. I'm just giving a high-level overview of how it could work. But the hope is that a swarm of companies could take on the larger companies, if they all agree to support the ecosystem. Right now, no one is coordinating, and unless that happens, they will all lose. And founding the company as a public good corporation, with a solid Ulysses pact to protect against enshittification might help people to get on board. I've got more ideas about pricing, open source licensing, etc. I think that could be tricky. You want it to be open, but not give away everything. Right now I'm doing a stint at the Recurse Center [2]. If you're looking for people who might be interested in your work, that might be a good place to look. There are a lot of interesting, smart people there. And it would probably be a great environment for you to do your work, too. Lots of people to bounce ideas off off. I highly recommend it. Those are some of my thoughts. Hope this helps. Seriously, email me if you want to talk. Good luck. [0] https://www.nngroup.com/articles/generative-ui/ [1] https://ghost.org/ [2] https://www.recurse.com/ |