We should. I was not expounding a policy. I was merely stating the results of a conference on the subject at hand. All of what was discussed in the article is weak AI. Narrow programs delivering narrow features. Social engineering doesn't require strong AI.