Dreamland Evaluater

Project Dreamland Inc. has to announce that we made a mistake. The "Understander", the machine we proudly presented last time proved to be inusable. We operated on some test persons, but none was able to reproduce the knowledge he or she gained during the operation. We could not even measure brain activity or receive any electrical signalling from the head, owing to some of the heavy force-fields we had to implement. Additionally the participants of the test showed signs of neural disorder after they had been enlightened and it seemed that the brain would undergo a rapid degradation process while being inside the tunnel. Unfortunately we have to conclude that this project failed.

We've also made a lot of thoughts on how to prevent this from happening in the future and we've come to the conclusion that only a solid autonomous machine evaluation process could help us in our future goals. We are thus trying to invent several automate "evaluaters", to aid us in determining the quality of our work. We are not thinking that they'll replace human evaluation in the near future, but as we've noted human descision making often is error prone and we felt the lack of a machine based approach in addition to the human jury.
The various thoughts on this machine circled around two basic ideas. First, we think about a very specific single-domain evaluation process implemented in a machine with detailed knowledge on the single topic only and a more general cross-domain evaluater that will try to link paths and determine the quality of our work given more general knowledge about a multitude of fields. We consider both evaluaters to be equally important, yet in some cases one might prove to be more useful than the other.
We're also using them to evaluate their own application in all the different fields and so far we could even think of a DSS aiding judges in court. It'll be interesting to get an overview of all the possible implementations and how they'll turn out.