Well, now this AI project has a name. I also registered the .com just for kicks, and I think it's punny. More wasted funds. Haha. And a long self-absorbed spiel - check it out on GitHub - not now, maybe in a week after there's actually anything to be found there!


(Seriously, who cares? But anyway...)

I'll admit that this project draws some inspiration from Linux. Once upon a time, there was no grossly dominant open source personal computing platform, no such operating system for multiple hardware architectures. The Linux movement essentially grew into the communal infrastructure which enabled just about anyone to dive into work on operating systems in a scalable fashion. Now anyone can jump into the activity of programming computers with nary a care in the world, thanks to the ecosystems which have emerged since 1991. That problem is resolved.

Today it's 2014, and the field of machine intelligence is still toddling about. Variance in models is high, a few high-profile projects such as Siri (AAPL) and Watson (IBM) dominate the press. We have many robust operating systems, and hardware architectures, both physical and virtual. But there are no simple frameworks which are also abstract enough to be powerful, and elegant enough to be broadly adopted, such that anyone can dive right into the business of making machines which are more human. Artificial intelligence remains the almost-yet-un-consolidated software design domain, of our generation. I think it's only a matter of time before a solution emerges, so I've decided to take a whack at it in my free time. Well actually I've been building my toolset for a few years now, but since I've finally gotten to writing code in C, I figured it's time to try and integrate existing libraries from the open source world at large into this project.

It's only a matter of time before the quality of our machinated minds reaches the biological sophistication required for self-propagation, learning, and self-healing. Soon after that happens, whenever it happens, we will quickly have more people on this planet that are based on electronic virtual machines running off x86 hardware, than people based on organic DNA. It takes years to train a natural human mind, due to our lack of grasp over wetware. Once the algorithms for artificial minds are of quality, it will take fractions of a second to train one to reach the equivalent mental capicity of an educated adult natural human. Entire societies of artificial minds will be born, warring, and extinguished in hours. It may be bright, or bleak, but anyhow, I'm curious to find out what the future will be like when there are more counselours than broken people in need of counsel.

So let's get on with it.

No comments :

Post a Comment