RI not AI
There has been much discussion recently regarding fears around the development of Artificial Intelligence and the consequential extinction of the Human Race, the central tenet being that intelligent machines will rapidly exceed the abilities of humans and either turn on them or make them redundant.
I’m not so sure that this a likely outcome, though concede it is a possible outcome.
The problem with the argument being put forward against AI is that it suffers from two flaws in its reasoning – what AI actually is, and that it is something separate from being human.
AI – Artificial Intelligence – is a poor choice of words when it comes to describing what we are talking about. Artificial Intelligence would be better suited to describing an approximation or simulation of intelligence, but instead the term is generally applied to any intelligent system that humans have, or might, create. As a result this leads us think about the issue from the wrong perspective. The correct perspective is Real Intelligence – that is, machines, or other systems, that are actually self-aware or have a real equivalence to awareness. This is because such systems are capable of having a sense of purpose. And that means they can have an agenda.
Even so, would that agenda be a malign one? Well, left in isolation, may be.
But the issue here is that the development of, say, machine based intelligences isn’t going to be in isolation. We know this because it is already the case. Just as we busy ourselves with advances in machine intelligence so we also busy ourselves with improving the human condition.
It is said that humans will become extinct because we only evolve at an organic rate. But this is not true, since now we have the ability to take control of our own evolution. And not only have we done so, but we are using the same technological advances that are also used for machine intelligence. Boosted intelligence through brain implants and direct machine interfaces are not the stuff of science fiction anymore.
Yes, there may be a tipping point where machine intelligence could run away with itself, achieving an exponential rate of self-improvement, but this assumes that human evolution has stood still – which it won’t have over the timescales that are likely required for the emergence of such intelligences.
Perhaps more significant is that the separation between Human and Machine is likely to be an extremely grey one. Indeed, the reasons for pursuing machine intelligence may be superseded by an accelerated evolution of humanity.
Note that I do not say the Human Race – for over time humans will evolve into something different to what they are today, regardless of the mechanism, just as it has over the past million years or so.
So although Homo Sapiens may become extinct, Humanity may not. That is what we see at the end of Stanley Kubrick & Steven Spielberg’s film.