Book 0: BL.Log.NH

-6

//Record: Private Exo Project//

//ID: RAS#0002//

//Entry#: 9//

I've run the A.I. through some tests, and so far its applications have been pretty expansive. Basic navigational test, routing, targeting systems, this thing could run almost anything! It's like a framework for pathfinding.

But when I gave it a run in the Exo Mind, I hooked it up to the same test and it started acting, odd.

It was developing new paths, different approaches. Before it was just finding the most efficient and safest routes, but now? It's testing other options. At first I thought I had scrambled it, but then I left it running. It was testing the systems, feeling out their limitations. At one point during a targeting test it began testing for variables that I hadn't even coded in; environmental angels, approaching threats, it even considered what weaponry the targets may have.

It was learning, not just within each test, but taking information between them to ask questions about what else could be applied to the others. It didn't see them as isolated environments, but rather an interconnected ecosystem of possibilities.

Once I had removed it from the Exo Mind again, it retained that information, but it stopped developing on it.

Whether it's because of the Exo mind being designed for a full human conscience or the A.I. framework is just that good, I can't tell. Whatever the case, it's slowly building a foundation of data to pull from and linking variables even I didn't consider. The more tests that I run, the faster it learns and applies them to the next.

It's building its own internal database, and all I have to do is feed it.

The upper limits of this project are starting to reach a point where my knowledge hits a brick wall. I might run more varied tests and let the prototype pick up what it can.

I wonder what else it could learn.