Bringing Evil To The World Of AI

Have you every truly stopped to think — what is evil? So many times we coat what’s dastardly or not by the pieces of fiction in our society. Then you have to think about what truly makes something/someone evil. Is it in the actions, or the soul? Hopefully, Selmer Bringsjord, a logician, philosopher and chairman of Rensselaer Polytechnic Institute’s Department of Cognitive Science will find the answers for us. In the study of AI no less.

Merriam-Webster defines evil simply enough - morally reprehensible or causing harm. I might be in over my head, but I have to disagree with their second definition. Causing harm is something we all do, I mean you can cause harm by being in an accident. It doesn’t make one evil, does it?

These questions and conditions lead me to ponder, can anyone be truly evil? Bringsjord states his thoughts on the matter,

To be truly evil, someone must have sought to do harm by planning to commit some morally wrong action with no prompting from others (whether this person successfully executes his or her plan is beside the point). The evil person must have tried to carry out this plan with the hope of “causing considerable harm to others,” Bringsjord says. Finally, “and most importantly,” he adds, if this evil person were willing to analyze his or her reasons for wanting to commit this morally wrong action, these reasons would either prove to be incoherent, or they would reveal that the evil person knew he or she was doing something wrong and regarded the harm caused as a good thing.

The experiment that Bringsjord conducted to help find the mysteries behind evil was called “E”. The character was given a physical shaped, and they began testing E to see how he would respond to certain situations. They programmed him with evil based AI, in which they took a situation of evil from a piece of literature and made E believe he was the person responsible. The situation was that a group of parents had given a pistol to their distraught son who killed himself. They led E to believe he was the parents.

Once the AI was scripted, they interviewed E to try and discern the logical reasoning behind his actions. He had but one cold, and calculated answer - “The boy wanted a gun, E had a gun, so E gave the boy the gun.”

Luckily, Bringsjord is keeping ‘E’ under tight lock and key. You can imagine if this AI was ever used for the wrong reason. I’d say something along the lines of ‘Wargames’ perhaps. All of this studying of his work has led me to a roadblock. Can you truly create something evil in AI? Without understanding on a completely human level, can AI discern what is harm and what is not? Or is it all just numbers and code?

[via: ScientificAmerican]