08/05/2026

I realize people are working really hard, harder than me, to integrate AI into their lives, but everything is currently scattered. There are plugins, mcps, skills documents given to the AI agent so it can connect to gmail, outlook, canvas, or even GUI-heavy applications like Blender recently. But I don’t think people are adopting it very well. Some use this and some use that. I am thinking of creating a new hardware that is only the AI agent, nothing else, just a chat box and voice, no GUI. But of course, that is also not gonna replace the iPhones or the windows, Macs. People see stuff on their phone and computer. There should be two distinctions that should be made regarding technology, or to be clearer, regarding the use of technology.

A recollection:
I use technology to watch reels, movies, and porn, stay connected with my friends and family, play video games, study computer science, write stuff in my journal. Technology is basically replacing everything in our lives, we could do anything in this virtual world. However, there hasn’t been a real personal companion, personal virtual human that we talk to, who sees the perspective that we are seeing, hear and listen to the things that are exposed to us. So far, people have been operating on their own virtual world, typing stuff down , clicking through their favorite shopping sites, manually booking their own Airbnb apartments. Just now, after asking ChatGPT for restaurants to go for lunch around downtown Chicago, while it did give me a lot of options, I still had go to Google Maps myself to know whether the restaurant does actually open on that date, at that time, and whether it requires reservations beforehand. It acts more like a Google search engine rather than a true companion, and I need it to be a companion that I talk to. I designed WindieOS to be that way, it could control my computer, edit files, browse the internet, but there is something extremely inconvenient that I feel about it. It takes a lot of steps to setup, the backend needs to be hosted on a different server. Starting it up is already the problem, using it is also shit. You have to wait for the agent loop to finish in order for you to control your computer again, which is the problem I have written down as computers, at least current operating systems, are made for human-computer interaction, not AI-computer interaction. Someone said to me though, AIs are smart enough to use the OS even if it is made for human-computer interaction. Maybe smart glasses are the future, it records the actual real-world environment that humans experience in, coupled with the microphone that listens to the environment’s sound and the user’s voice, and the on-glass display feature is pretty nice too. Let’s take a look at Tony Stark for example, he not only uses his smart glasses, but he also uses the computer, interact with it, and the AI companion works with him. But how does it work with him so that the productivity and effectiveness increased exponentially instead of just slowing him down, that’s the problem I am having. Should we let AI use our own computer, or give them its own computer to work with? Should we create another operating system for the AI to work in more efficiently instead of utilizing the existing operating systems that are specialized for humans. Or maybe I am thinking about this wrong, because operating systems are just way more than a pretty graphical user interface that facilitates the every consumer. I am a software programmer myself and I know that the operating system is a program executor itself, it handles multiple programs running all at once and schedule the executions of all of them, not only that, it handles memory usage, inter-process communications, GPU uses, an operating system is plainly a piece of technological magic. Unless I can strip down all the features that facilitate human-use in a current existing operating system such as typing, clicking, displaying, designing my own operating system might just be a waste of time. 

I hate the phrase “Leave me alone”. I mean you have people caring about you so much that they bother you. I sometimes have days that no-one would even want to talk to me.

I will put WindieOS out there, at least there will be an open-source agent that can control your computer, code, and search the web. Add functionalities so people could add their own mcps, plugins, skills, custom tools, different business’s apis for tts, stt, and several more.

This place, California, Bay Area is beautiful, I don’t think I can get enough of it. The weather is amazing, its romantically windy, I could smell the taste of the sand or maybe its seashells, and there are so much wildlife here too, especially the birds, just gliding on top of my head. I feel like I could achieve anything around here and maybe being a cog in the Google machine isn’t so bad, you get all these life stuff , maybe what I mean is that the way Google is might be the way I wanted to create company of my own, so why not join them you know?

WindieOS might be the same as OpenClaw, Codex, Claude Code but what is important is that it is ME who created it, it is ME who birthed it to life, at least it is my own you know, nobody else’s. And I am going to share it to the world. I don’t care what the world will think of it, I care about what I think of what I put to the world and I think it is a good thing to the world, a beautiful, functional thing that I dedicated my blood sweat and tears into. Who cares what anyone thinks? Just do it.

Agents -> Interaction models -> agents with interaction models -> what’s next?

There should be a foundational tool, the next google, per-se for AI for coding, for humans interacting with the AI, not like this.