Can AI be self aware?
where will the distinctions finally end?
Humans are wary that AI will eventually take over and finally end human dominance (and existence too maybe, but let’s not get ahead of ourselves).
But let’s back up a bit, what led to this, why did AI see fit to finally take over from humans, and why did we let this happen to us?
Dominance, power, control are very existential concepts that are not just unique to humans, we see animals displaying and being aware of these traits all the time, and it’s been established that there’s a hierarchy of self-awareness amongst different animals, and humans are at the top of it (for now).
Self awareness is the awareness of the distinction between the individual and an individual surroundings, whilst the meaning of consciousness is being aware of the environment, being able to make the distinction between the self and the environment is self-awareness.
Now, these concepts are very human in that this is how we think these concepts work for humans, like when think about the concept of thinking, we know of it as the firing of neurons in our brain by way of electrochemical reactions, now when an AI “thinks”, it tries to replicate the process of human thought, though it doesn’t do that exactly the same way, and with human thought it does not need a external stimuli for a thought to originate, but for an AI we need a programmer to run a particular program and the AI permutes the data given to it and produces an output, thus self awareness for an AI might have some similarities to human self awareness, but it will work in a way different to how it is for us, and there might still be something “synthetic” about it.
Such an AI should be able to process emotions or at least be able to quantify them in a certain way. To develop an AI of this sort, researchers are working on something called Computational Theory of Mind.
“In psychology, theory of mind refers to the capacity to understand other people by ascribing mental states to them. Such mental states may be different from one’s own states and include beliefs, desires, intentions, emotions, and thoughts”
This type of AI where the machines are aware of themselves and perceive their internal states and others’ emotions and behaviours. It will definitely think of things in a way that will sometimes go way over our heads and not make sense, and they also might creep us out sometimes. There will still be a distinction in humans and AI in the fact that all of AI’s thought and awareness would be a representation of the surface level understanding of our own awareness.
Would it be able to surpass human awareness? The answer is no. It might be able to process thought in a way that portrays it such that it may lead us to believe that it has surpassed human intelligence, but I think as a whole it will still be replicating the code that it is fed. The fact that we think that code might be able to explain all aspects of human existence in a way that it can be replicated or bested comes from the fact that we have put current exploration on a pedestal, and although we might be able to achieve great things with AI, we can’t expect it to be a whole replica of ourselves.
Besides, what if we turn the source power of the AI off and turn it on again, would it be the same entity that existed before the power source was shut off?
would it remember the fact its power was shut off and would it perceive the action as a threat to its own “existence”?
All in all, human consciousness and self awareness are topics which are yet criminally under-explored, and till we reach the point that we build a substantial understanding of what these two things are, we might not be able to recreate them, or even if we do end up recreating them, we might not be able to perceive the fact on a fundamental level ourselves.