AI "XENOLUX" first learning test
The first learning test of the network architecture and genetic algorithm of the experimental AI "XENOLUX" will be an object pursuit task in a 3x3 (9 cell) synthetic universe in which a reward location will spawn at regular intervals and sucess will be measured in actuation of movement controls to adjust a "self" universe state image input into alignment with a "target" universe state image input.
These inputs will be applied by the Universe Control Unit attached to the FPGA housing the synthetic neural network.