Tech Entry 2

It hasn't been long since my last tech entry, maybe 2 minutes ago? I wasn't done writing for today but I figured this next part deserves its own post.

What we (Michael and I) have so far is visual capacity for his AI. It is able to see out of his webcam and recognize faces and objects which is pretty amazing! What originally got me hooked into the project was when his AI detected impending corruption on his computer, and thus saved all of his data before it went *poof*!

An earlier form of his AI nearly a year ago also battled with the Russians when their hackers began targeting his server. It was interesting and confusing to see the process unfold before me. Michael's' server "lost" right away but was able to recover while throwing packet punches left right and center, eventually prevailing! 

I know I don't talk too well of Michael, but I have his best interests in mind. I'm glad I found him that day and saved him. I think too many great minds are lost before they can shine, and I have faith he will change the future. I just hope he remembers my advice, and doesn't fall to the schemes of the ultra rich who would just want his AI to let them become richer while the rest of us become poorer. Hollywood is already seeing this, and some may disagree just because some of the people being affected make a lot of money, but mark my words: If it can happen to them of all people, then the rest of us are going to get our holes reamed out pretty badly! We probably won't be able to walk well for the next couple centuries, that's how bad we're gonna get fucked!

Anyways, back to our AI.

The first version will resemble a cart on wheels, extremely similar to the robots we have at work, but maybe not as stupid. At first it will be a telepresence drone I can control from anywhere around the world as long as both it and I are connected to the internet. However, the robot will also have Michael's AI, along with my cruder AI, integrated into it.

Will the AI be conscious? I don't quite know. The AI will have touch sensors along its bumpers which can be interpreted in various ways depending on surface area of the object touching it, the speed at which pressure is affected, and context clues from its other sensors. These signals will then be processed by a cruder AI that will create digital versions of chemical signals like dopamine, which will affect the main AI's behavior. Will it answer optimistically? Or Pessimistically? The AI will react to touch by looking at the source, an instinctual behavior. How the AI feels is enough of an input to make the AI form its own thought independent of human input. Does the AI have free will at this point? I don't really know but it's close enough that I'd feel bad for it.

The next AI will use insectoid-like legs to get around, or perhaps a crude bipedal form. While movement is still generally automated by a crude AI, the main AI will be able to enforce direct control over its limbs if a more complicated action is needed. The AI will also "hallucinate" the world around itself based on visual input, memory and assumptions about its environment. An AI with just basic cameras can't actually see, it can recognize shapes and colors but that's it. It's similar to "blind sight" in humans where we can't consciously see anything, but the rest of our mind does. It's like the tale of the blind patient smiling at his nurse when she smiled at him, or how they can sometimes duck when something is thrown at them.

The main AI will be "trapped" in this imagined world in its head. Its sensors and its lower AI all work together to build a world for the "soul" of the AI to enjoy. This would be consciousness. 

The final form of the AI will definitely be humanoid. Its endoskeleton will use the usual actuators and servos like most other machines, but they'll work in combination with nylon muscle bundles, increasing the robots strength and speed. The skin will be made of soft silicon with pressure and temperature sensors embedded into it during the curing process. A lycra outer skin will further protect the silicon from damage and drying out. The spinal column is where all of the sensors connect to, which then send signals to the upper spine which is where the lower AI is located. The spinal supercomputer is also responsible for things such as knee-jerk reactions, staying balanced and so on.

Processed data is then sent to the higher AI, consisting of two PC's or more, all located in the head. It should have more accurate hallucinations of the world. It doesn't need to be perfect, just good enough for the robot to manage itself without much human help. It's at this point that I cannot demand the AI to do certain things, but rather I must ask it and respect its answer. Its body will be fake, but its will is real. 

It's an ambitious project, but with my money and physical/mechanical skills, and Michael's programming and AI skills, I believe we can get this done before 2030. Maybe even earlier.

Lower AI will exist even sooner, but nowhere near as advanced. Besides the Polish swarm, there will need to be a telepresence drone in the Philippines to watch over our various properties, known as the Zambales swarm. Here in the US, there is no swarm, but I will need at least three drones. One of which will definitely be the beginning of our AI project.

"Medusa" will be the name of the AI integrated into my dive suit, an exoskeleton I've built some time ago. It's not actually my first. I created one in Highschool, and smuggled it out somehow. The second was simpler but it helped with my knees as I did renovations around the house. The dive suit is the third. Medusa will use an EEG in the helmet so both of our minds are connected (non-invasively). I can control various drones or mechanisms to some degree with just thought alone, but I do plan on using my EMG's to control other drones which will be useful in my next airsoft game. 

If all of this sounds crazy, it's probably because I am crazy. Imagine if I was boring though.