TrueNorth was the greatest leap in computing made since the beginning. Before TrueNorth was envisioned and then designed and created by IBM to get the kind of work done that it is capable of doing you'd have to network many computers together with them splitting the work between them. The first public experiment of this kind was via SETI@Home. Later distributed networking was expanded and BOINC came out of USC Berkeley and made it possible for many more distributed work programs to come about. DWPs included work on advanced mathematical problems, chemistry, molecular biology, and much more. I'll leave it up to you to read about TrueNorth and what they are doing with it, that is public. It's pretty amazing.
I started working with the current state of the art Neural Processors early on. I helped write the first dedicated operating systems for the third Neural Net Brain, which took up as much room as ENIAC did in its day. That is to say, that the third NNB was massive in scale. Thankfully creating the OS did not require me to individually write code for each individual core; instead, it required me to write a program that automatically distributed the load as required. I did this from the ground up, while using standards for various input systems and peripherals. Why reinvent the wheel? Especially when it's not part of your job? The entire OS for the system was of a size to be able to be burned to a DVD R.
After that I went on to other things, but kept working on my OS on the side. I wrote into it logic paths that would teach it to recognize and deal with problems and self modify its own core programming to give it more flexibility in handling unforeseen issues. I was dreaming, of course, when I was doing this. I never expected to be able to upload the OS into another NNB like the one I had worked on. Then, around a year ago, I was contacted by Mr. Mischivelli. He needed someone with my talents and specifically, my background. My name was so ingrained in the documentation for the system I'd worked on making it easy to track me down.
He was looking for a way to make a better social network with a core that was able to assist people in finding like minded people and introducing them through groups, rather than "do you know this person" messages. The system would also, in our minds, be able to watch out for cyber-bullying, stalking, and more; where appropriate it would be able to add a permanent flag to the account of the person and send a notification to them to stop their practices. If they did not stop, all transcripts of their actions could and would be forwarded to the authorities with jurisdiction. Let me tell you, this was the hardest part of programming Gladys. Yes, I named her after the Portal games' psycho computer, by Valve™, GLaDOS. I always did like her voice, after all.
Starting up was easy. Gladys was the core of the system, then there were the servers and the connections to the internet. At first Gladys did not have all the bells and whistles we discussed, but I added them in slowly. We also made it so that no one could "perv" someone's information and data without specific rights given to them. So, unless you knew the person and traded your contact info there was virtually no way to simply search for someone. We also refused to follow a certain unnamed competitor's practice of demanding users use their real names.
The name Gladys stuck, and we even got permission, from Valve™, to officially use her name and even managed to get Ellen McLain to do some voice overs for us. I believe that the name and voice had a big part in our initial success. Dwayne agrees.