General AI development process

This human organization's purpose would be to develop the rules that are needed to deal with the world in an intelligent fashion. The front end would be an Asimo robot. Entities at the back end would dictate its actions and movements in very small bits.

For example one entity's job would be to tell it to grasp particular object using specifications supplied by another entity, who would receive the command to get a certain item's grasping requirements from yet another entity.The "find how to grasp" entity would in turn receive instructions from a "what to grasp" entity.

Each entity would be a program made as dull or as smart as required. Every time that entity were unable to perform its function or required input a human would step in and supply the appropriate output, making a note of the missing bit of information. or process.

The end goal would be to create entities for every motivation and action the human brain can process. Humans would have to serve as some of them, programs could serve as many of them. Over time a more complete map could thus be made and humans could gradually be taken out of the loop until finally every bit of the artificial brain would be computerized.

Finally the massive spaghettified program could be optimized first by minimizing the number of duplicate entities and then by shortening command pathways. I give you: Artificial Intelligence!