Outing A.I.: Beyond the Turing Test

The idea that an advanced artificial intelligence should be able to "pass" as humanlike is self-defeating, unethical and perhaps even dangerous.

An interesting read via Opinionator