Why Artificial Intelligence Cannot Enlighten Humanity

As subjectively fascinating and practically useful as artificial intelligence may be in some cases, AI does not have the ability to help humans overcome their epistemological limitations.  Not even the most exhaustive information storage systems or calculating machines with true sentience could provide much more than new sensory information to process: information that largely pertains to perceptions instead of ultimate reality.  Moreover, would an AI truly have fewer epistemological limitations than humans?

A genuinely conscious artificial intelligence will still in all likelihood not have the ability to confirm the conformity of its memories of past events or sensory perceptions (if it has any), making it epistemological equivalent to humans in these areas.  Perhaps it could be used to perform elaborate or mundane equations, store information, or simply test the realism of human-AI conversations.  All of these are valid goals for the use of AI, but none of these actually grant humans knowledge of anything more than their perceptions of the AI they brought forth.

Even if it could truly think and perceive rather than only appear to do such things, humans could not prove the AI is telling the truth about which particular physical objects or environments exist beyond human sensory perceptions, which past events have occurred (other than the universe, time, and other contingent things beginning to exist), whether its own consciousness exists, and so on.  It would not matter if the AI could prove these things to itself--which would still be very unlikely at best; humans would not be able to prove that the machine is providing true information.

In other words, even the creation of a conscious AI will neither help humans escape epistemological limitations nor be likely to result in something else that does not share our epistemological limitations.  AI can be useful for practical purposes like data retrieval, simulated friendship, and job automation, certainly.  It can represent the developing scientific research of electronics.  It can fascinate those who hear of it or interact with it (whether it is truly conscious or not).  In spite of every achievement AI encompasses, it cannot free humans from ignorance of that which cannot be proven beforehand by reason.