Fallible perfectible: Schizophrenic boring intelligent machines

Feedback is an essential part of a intelligent system. Indeed it constitutes the architecture of what an intelligent system is: we can say the intelligence is produced by the every fact of being re-fed. A feedback is an information about the environment but most of all a reaction of the system itself as a entity in the environment. So the most important thing feedback is telling, is about the system itself. An eternal intelligence, a perfect system not requiring feedback will be difficult to imagine. We can call this hypothetical machine, a god's mind. We can say, cybernetically, that feedback is the speculation of the machine about itself. A god's mind actually wouldn't require this kind of information. A god's mind would prefigure every possible configurations of the system. So in other terms, a god's mind doesn't think. In a certain way, given our tradition to describe intelligent systems, a god's mind is not intelligent at all. What an irreverent conclusion! And all for the power of feedback...

In a very crude, engineering perspective, feedback is what make our algorithmic reasoning machines the fallible, perfectible humans we are. So we like feedback? Well, yes and no (of course, damn human condition: you always want to swing in between two opposite poles: why???). The feedback, insofar constitutes the very horizon of our possibility to know and to reflect and to be the fallible perfectible systems we are, also feeds our suicidal epistemic appetite.

A feedback is designed to come back to the system and to inform it about the new configuration of the system itself. Per definition a feedback is the most valuable element for the system to know. Knowledge in mathematical terms can happen instantaneously. A theorem is what it is regardless of time. On the contrary real machines (and for this argument, living beings are real machines), require to be place in time. The process of knowing is placed in time. The intelligence of the intelligent systems is placed in steps orchestrated by feedback. A feedback is the intrinsic reminder of the system of itself. An intelligent system is constantly put together by the memory the feedback is feeding. We can say that the “space-time” of the intelligence of an intelligent system, is made possible by feedback. Or that a mind is made possible by its reminder

Alleluia feedback? Unfortunately what is the happiness of engineering is also the damnation of the human condition (in this case only...). First, our mind is not just an intelligent system: it's also a theory generator about the environment is which is placed (which is not entailed by the mere fact of being an intelligent system). A human mind will tend also to give description of its place in the environment and consequentially of itself in the environment. But, accordingly to the distinguished philosopher Willard Van Orman Quine, we can give an infinite number of different descriptions (theories) of the same configurations of world. These infinite descriptions can be reciprocally contradictory, though with equal descriptive power. For our argument, our feedback can confirm and remind an infinite number of different variations of the same intelligent system that originated the impulse of the feedback: the feedback is reminding the identity of the source intelligent system, but it is reminding an infinite series of possible identities. This is the Pirandello fallacy: an intelligent system is ONE, but it will forget his own intelligent process unless reminded by feedback. Feedback can remind the intelligent system of 100 000 alternative, concurrent, reciprocally contradictory but descriptively equal intelligent systems. Consequence, for Pirandello, we don't have ANY true intelligent system.

The second paradox lies in the contrary effect of the feedback reminding: the genuine exploration of different possibilities of description from the intelligent system will be constantly aborted by the feedback, that will prevent the intelligent system from changing: if the real process of knowing is the evolution of alternative description of configurations and feedback is reminding the original descriptive system, then every feedback will reinforce the source request and will prevent every genuine mutation of description.

I hope not to have raised a generation of feedback-haters. This was just a recreational philosophical Con-Und-Drum. Though some real perplexities can remain, if we consider that since the appearance of Homo Sapiens Sapiens, we are actually nothing more nothing less than humans: always with the same annoying philosophical problems. On the other hand, in comparison with our ape mates, we present a much higher mind instability. Good night.