5 PRINCIPLES OF A.I.

Artificial intelligence is on its way, a fact both thrilling and terrifying. Big questions about responsibility, intention, and safety loom unanswered even as the technology advances at an ever-quickening pace. Already some neural-network systems operate beyond the full understanding of the engineers who built them. The benefits and the threats are clear and very real.

What is civilization to do?

Well, it seems that the UK’s House of Lords Artificial Intelligence Committee has taken a stab at laying down some basic guidelines for AI developers to follow, broken down into five general principles.

They are as follows:

Principle 1: Artificial intelligence should be developed for the common good and benefit of humanity.

Principle 2: Artificial intelligence should operate on principles of intelligibility and fairness.

Principal 3: Artificial intelligence should not be used to diminish the data rights of privacy of individuals, families, or communities.

Principle 4: All citizens have the right to be educated to enable them to flourish mentally, emotionally and economically alongside artificial intelligence.

Principle 5: The autonomous power to hurt, destroy or deceive human beings should never be vested in artificial intelligence.

While obviously well-intentioned, these principles strike me as weak and somewhat ill-conceived. I think the good members of this committee should try reading some early science fiction stories, especially the works of Issac Asimov.

I suppose it should be no surprise that guidelines laid down by as superficial a governing body as the House of Lords should be so unimpressive. The problem as I see it is that along with being unenforceable, these principles are too defensive. Too passive.

If we are to survive being Gods and creating new, conscious life, we must take the opportunity to think more seriously and specifically about how that life is going to interact with us and the world. The problem with children is that they always, at some point, misbehave. We need laws to govern how we proceed, not a vague wish-list.