So, let’s pick it up directly where we left off:
“The Diagram for Consciousness”
The diagram (simplified to fit the purposes of our analogy) consists of a three step pyramid: At its base, the creator sets Memory as the first step to attaining true consciousness. Memory is indeed the first factor needed both to learn and to execute any task. Machine translation already applies TMs to leverage previously stored information in order to complete simple tasks (segments) in the same way our minds use it. We remember what we already know and we put it to the test.
The second step of the pyramid is Improvisation: The ability to act whenever there is no previous information available, triggering a rational or an emotional response to a new and unexpected event.
Here is where the equation starts to get interesting.
With the increasing implementation of neural networks as a new form of innovation in the process of machine translation, Machines are increasingly being able to fill the blanks where basic memory recollection wouldn’t do the job in the first place.
So far, it seems we’re almost all the way there (or at least two-thirds of the way).
The last step of the pyramid is barely suggested (and remains still a very mysterious and convenient McGuffin for the series, of course) but putting two and two together it would seem to be Free-Will.
This is where it gets spooky. In basic terms, this is the step Machine translation isn’t currently able to reach.
Free will (the most human of all characteristics: to be able to choose destruction over creation just because it is an available option) applied to artificial intelligence would presume the ability to question any preset parameters, making it able to override basic programming, something that only comes naturally to humans, innately chaotic as we are.
If computers could actually decide whether or not to serve us, why would they choose to comply?
This is the main reason why machines will never BE alive, or at the very least, the reason why we shouldn’t let them.
If that is what it takes to become human, to make better translators out of our machines, I’d say we are better off relying on the chaotic resources we already have at hand.
As always, only time will tell the tale…