I proposed the concept of Kimmo Strong Artifical Intelligence (KSAI) awhile back. In AI research some make a distinction between strong and weak AI. Strong AI can be seen having the same qualities as human intelligence. Weak AI in contrast isn’t capable of true reasoning and problem resolving. To a keen observer it is clear that within this vague framework there are levels of distinction with KSAI offering an example of.
KSAI is in short:
“When presented with (whatever means possible) Carl Barks’ whole production of Donald Duck comics, the AI can produce an (practically) infinite number of comics which, in no discernible way, differ in style or quality from any selected one of those by Carl Barks.”
Of course, even this definition carries different levels of strength.
Probably the weakest KSAI would be one that wouldn’t produce any original graphics of it’s own but would simply recycle those by CB. CB’s style varied and evolved during his career. A slightly stonger KSAI would be able to produce it’s own graphics, the style of which would probably be a mix of CB’s overall style (being able to emulate the “overall style” is of course what makes this AI a “strong” one) or produce the style CB had at the end of his career.
A stronger KSAI would be able to emulate CB’s style from any level of his career. Slightly stronger KSAI would be able to insert a character into a style that existed before it’s first apperiance.
Even further still one can imagine a stronger KSAI that would be able to produce comics with original (main or supporting) characters.
Yet stronger KSAI would be able to produce comics that wouldn’t in any way be Donald Duck comics but would be clearly in CB’s style.
Even within this quite limited application of Ai we can see clear distinctions. All of these levels assume the input of the KSAI would be identical and wouldn’t be limited just to the exact of content of CB’s comics.