The set of actions, behaviors, and outcomes that a human or other Modosophont would view as benevolent, rather than malevolent; nice, rather than malicious; friendly, rather than unfriendly; good, rather than evil.
Friendly AI A device that does what you ask em to, as long as it doesn't hurt anyone else, or as long as it's a request to alter your own matter/space/property; a device which doesn't cause involuntary pain, death, alteration, or violation of personal environment.
- Best Friends - Text by Michael Beck
A psychological simulation is made of a targeted individual, allowing the user to gain near perfect insight into the target's preferences. A cloned "companion" is then grown, and threats to the companions welfare are used to pressure the individual. The companion does not even have to be physical (expensive), but could be a virtual copy.
- Empathy and Deductive Telepathy
- Perfectly Friendly AI - Text by M. Alan Kazlev, based on Creating Friendly AI
 Friendly AI which never makes a mistake and always takes the maximally friendly or baseline friendly action.
 A benign turing, superturing, or hyperturing.
Text by M. Alan Kazlev, based on Creating Friendly AI
Initially published on 29 October 2001.