Human Control of Thinking Machines

Humans will want to exert a control over the activities of Thinking Machines when these are material to human interests. This will be true whether a Thinking Machine is small, that is roughly the same scale as a single human being, or one or a collection of Thinking Machines at the scale of human families or social groups or states.

Fundamental laws will be needed from the first. This is easy to say, but there are many difficulties. First - and this may not be obvious - is the definition of a human being and humanity in a way that all Thinking Machines can understand and respect. Defining a Thinking Machine follows. Then is the drafting laws that are acceptable to the variety of human societies with their different cultural sensitivities.

In time a jurisprudence will evolve from the fundamental laws. A body of precedents and rulings will form to resolve overlaps between the laws and conflicts between humans and Thinking Machines and between the machines themselves. There will be codes of practice and dispute resolution mechanisms. Interpretations of these will be needed as human languages and societies evolve.