And, Apparently, We All Are
I have as of late been listening to a number of very bright mathematicians state that mankind is now doomed due to super-intelligence and its near immediate arrival. Supposedly we are too far down the evolutionary road in relation to AI, that it can be stopped, curtailed or controlled.
Several of these individuals including the ‘father of AI’ suggest that mankind’s future is now limited and that we, in fact, will no longer exist in short order. One has stated that this will occur with such speed that it could happen by the time Donald Trump is no longer President. And while it would give me great personal satisfaction to see some robotic being excommunicate Don the Con from this planet, it somehow doesn’t seem all that satisfying if I’m going to join him at the same time.
And yet, Dr. Roman Yampolski states that mankind will suffer 99% unemployment within five years. AGI, the term used to indicate that artificial intelligence exceeds human intelligence, he suggests, is already with us, although in narrow domains. But, within five years it will have surpassed AGI to super intelligence where AI can perform any human task better than any human.
https://youtu.be/wP24u80lOGY?si=g-tljl9Y8oIOkQET
So, question becomes little more than a rhetorical one. If no matter what we ‘think we know’, a machine, a robot, can perform that function quicker, more accurately at either no cost or very limited cost, what does the future hold for mankind? Even if somehow this super intelligence does not immediately determine that ‘man’ is nothing more than a parasite on the resources, the flora, fauna and species of animals on this planet, what rationale might it have to ‘maintain’ us?
At that point, man has no particular value, in which case our continued existence becomes impracticable. Consider that for just a moment. We would then be expendable by pure rationality.
The key characteristics of AGI are as follows.
He suggests that our hubris reassures us. In fact, “AI safety is not only hard, it’s mathematically impossible, with exponential improvement in the former and linear/constant in the latter. And the difference gets bigger every day.
In other words, whatever we devise, super intelligence will have already built a back through which it can bypass any and all safety systems as it sees all possibilities and probabilities, and by ‘knowing’ makes any and all actions by man, moot.
If he is right, and several other theorists and experts in the field say he is, like the father of AI, Geoffrey Hinton, then there is very little that you and I can do to either prepare or survive what is inevitable in the future.
Mankind, will become irrelevant. Just as Donald J. Trump and wealth will be. In a perverse sense I actually find that comforting.
Good luck to all of us. We are apparently going to need a ton of it.
Leave a Reply