While discussing on the subject of Singularity according to Jay Kurzweil within the Xavier Group Ltd on LinkedIn, I was asked to comment on a phenomenon they called the "Brick Wall" where, as I understood it, humans will try their best NOT to hit the Singularity point where AI becomes more intelligent than humans. Mostly because of religious groups, people in power, and the fear that these AIs will eventually treat us as an inferios subspecies.
Truly, I think this is very far from what may happen. Here is my response.
Taking from the research of Paul Chappell who did a great job in his speaking engagements to pull out all sorts of evidence that fundamentally, humans are peaceful beings that only wish to have the basic requirements of living, and beyond that, be free, love and be loved, and live in harmony with everyone else. He shows in his lectures that violence is in fact something that is conditioned by the people in "power", who themselves are at arm's length of the violence perpetrated, and direct good people down the violent road for their own selfish ambition.
This is actually very important when we consider AI.
It means that for the most part, us humans will create AI that are.. well, like us. computers that aren't violent at all. Right now, there are still people in "power" that try to control us. In the most communications-dense countries like Europe, Japan and North-America, its getting harder and harder for the "powers" to control people. The real information is literally at our fingertips and instant, and penetration of the truth through social networks just reveals the bullshit of the "powers". So in these countries, where most of the highest tech is developed anyway, we're not really moving towards a system where we're going to build super AI's in the pre-Singularity timeline that will be agressive and out to destroy.... At least that is my hope.
The acceleration of technology and propagation of information into other less-developed countries will bring them to par with us and the population as well informed as us in just a few years. Peter Diamandis of the Singularity University estimates 66% of people on the planet will be Internet connected in the year 2020. It is estimated that almost all people will be connected by 2029.
So it seems to be going in a direction where super AIs on the point of singularity will be programmed to support humanity as they surpass our intellectual capabilities.
Psychologists today say that humans are wired to feel pleasure, to love and to be loved. That's basic to our innards beyond food, sleep and security for us, once more a programmed survival instinct. Therefore intelligent humans that have all those things feel content.
Therefore it can be assumed that AIs that are wired or programmed for work, dangerous missions, or, get this, deriving satisfaction from say creating tasty vegetables at a highest yield, will use their intelligence to achieve just that.
If we treat AIs just like people, their basic needs and programming cannot be changed so if we program them to get satisfaction from something we feel is boring, repetitive or mundane, that's what they will enjoy doing, just like we'll enjoy whatever we wish to do.
IMHO then and in conclusion, we are directing ourselves towards a coexistence with AIs. There is also nothing preventing humans from improving on ourselves either mind you. Heck, the scientific programmed AIs can probably help us out of pleasure too!
Cheers!
No comments:
Post a Comment