feed the computer it's own actions (SPOILER ALERT)

Thoughtful maybe but far from original. This was lifted directly from Star Trek episode "The Changeling", among others. But since the whole premise was lifted from Eagle Eye I'm not surprised.

As for the political thought, is that really necessary here on IMDB? Can't any part of the net be free of political bickering? Especially when you're so wrong :}

There was no implicit mechanism for feeding it's own actions as input into an analysis... and as you see, when explicitly directed to, it went into an infinite loop... (or at least the reasoning and deduction part of the software).

Infinite Loop aside, it definitely broke itself from the loop when it reached a certain reliability of result.

The power of the programmer is that we can design ANYTHING into a system. We literally can make a system work in anyway... So anything is possible.

Can we say that the AI achieved consciousness of self? First we need to define consciousness and that goes hand in hand with defining the self... those are two definitions that we have historically been unable to understand.

It seems like we run into similar self-referential problem... can the hammer hammer a nail into itself?... Can we see our own eyes directly?

We have to look at something else to tell see ourselves and then it's not reliable because we can't certify what we get is real.

This line of thought always leads down to a matrix-esque concept of "What is real". We can't trust our senses, we can't even trust our brains, reasoning and deduction.

So ultimately, we have no real concept of self, and consciousness appears to be simply moments when we focus on something fully in the moment without mental judgments, categorizations, or interpretations... at all other times we are unconscious in a very real way... on auto-pilot.

This AI was simply following programming which never mandated that it analyze it's own actions... until it was told to do so.