You requested a brief description of how our technology works. Well, here it is as best as I can understand it.
So you know that generative AI stuff that has been really popular the last few years? Type in a sentence and suddenly it spits out a pretty picture? Type in a prompt and it’ll write you a novel? What we were building here is the same idea but applied in very different ways. For example, for the human behavior ones, we created a program that converts human behaviors into data points, such as stock market transactions, and constructs data based off of that. It then takes that data, feeds it into a generative AI program, and attempts to use that data to predict future outcome, in this instance, stock market predictions.
The difference between our technology and the generative AI you can play with online is that our AI is simultaneously doing that while optimizing its own code and data processes. Effectively the generative AI is running on its own code to create the best and most efficient algorithms that still accomplish the same goal. So our technology is simultaneously gathering data, generating outputs, and optimizing itself all at the same time.
Now fortunately, the technology itself was deliberately made to be agnostic in its usage. Therefore, whenever you hear people talking about the “shards”, they are talking about a series of servers that are grouped together to run a version of this technology, each group with a different set of data, different parameters, and different objectives. Some of them have been built to talk to each other, for instance, human behavioral models and human historical studies both frequently pass data between themselves. Others are completely isolated, like the genetic one, and the space one.
Hopefully this helps a bit, and if not, fortunately, your job isn’t to manage the shards as much as to manage the computers that run them. Just keep them running and our programmers will handle the rest.
Larry