zlacker

Obviously feeding more data won't do anything besides increase the knowledge available.

Next steps would be in totally different fields, like implementing actual reasoning, global outline planning and the capacity to evolve after training is done.