RE: Artificial Intelligence
July 14, 2015 at 10:49 pm
(This post was last modified: July 14, 2015 at 11:04 pm by JuliaL.)
What the heck is 'general intelligence' and how is it measured?
Hard for the youngsters to grasp, but at one time being able to play chess well was seen as indicative of high intelligence.
Saying that a generally intelligent individual should be able to find creative and novel solutions to difficult problems begs the question of what is a problem. Ability to solve a problem implies that there is a framework of valuation under which the problem can be said to be solved or not solved. There have to be goals.
If the goal is the accumulation and coordination of data, then mechanical means have long since surpassed human ability. Scientists can't work without libraries, publications and other databases.
The question of whether conscious machines can be made is more interesting. Certainly they can be made self aware in a limited fashion. Any PC that keeps a register updated with CPU temperature or fan speed does that. But one that has enough power to keep tabs on itself and its surroundings...One with effectors to give it the ability to affect its environment and a payoff matrix where its own homeostasis is weighted as the highest utility...that we haven't seen. It would take a lot of selection to give a trojan or virus or worm the ability to take over the power plant on which it depends for continued existence. That's the bad boy that Elon Musk fears. Humans could get wiped out in an ill conceived optimization.
Though I'm not conversant on the current state of affairs of computer hardware and the hard limits to Moore's law. I'll stay agnostic on how far it is likely to go.
I wouldn't want to be remembered for a quotation like that attributed to Lord Kelvin, "There is nothing new to be discovered in physics now, All that remains is more and more precise measurement." Supposedly said in 1900, five years before Einstein's annus mirabilis.
I voted 'good' because only our robot overlords will be disinterested enough in the kind of petty human power grabbing that stratifies society, destroys all other life for its own gain and causes endless human suffering. Maybe they'll keep us as pets.
And it's the only way that consciousness will ever life extend long enough to get off of this rock and go out to explore the rest of the galaxy.
Edit---General intelligence. OK, I read the wikipedia. It's what humans do. Kinda anthropocentric isn't it?
I mean really, 'the coffee test???' Someone isn't intelligent if they can't go into a house and make coffee? Nobody outside of Yemen was intelligent before the 16th century?
Hard for the youngsters to grasp, but at one time being able to play chess well was seen as indicative of high intelligence.
Saying that a generally intelligent individual should be able to find creative and novel solutions to difficult problems begs the question of what is a problem. Ability to solve a problem implies that there is a framework of valuation under which the problem can be said to be solved or not solved. There have to be goals.
If the goal is the accumulation and coordination of data, then mechanical means have long since surpassed human ability. Scientists can't work without libraries, publications and other databases.
The question of whether conscious machines can be made is more interesting. Certainly they can be made self aware in a limited fashion. Any PC that keeps a register updated with CPU temperature or fan speed does that. But one that has enough power to keep tabs on itself and its surroundings...One with effectors to give it the ability to affect its environment and a payoff matrix where its own homeostasis is weighted as the highest utility...that we haven't seen. It would take a lot of selection to give a trojan or virus or worm the ability to take over the power plant on which it depends for continued existence. That's the bad boy that Elon Musk fears. Humans could get wiped out in an ill conceived optimization.
Though I'm not conversant on the current state of affairs of computer hardware and the hard limits to Moore's law. I'll stay agnostic on how far it is likely to go.
I wouldn't want to be remembered for a quotation like that attributed to Lord Kelvin, "There is nothing new to be discovered in physics now, All that remains is more and more precise measurement." Supposedly said in 1900, five years before Einstein's annus mirabilis.
I voted 'good' because only our robot overlords will be disinterested enough in the kind of petty human power grabbing that stratifies society, destroys all other life for its own gain and causes endless human suffering. Maybe they'll keep us as pets.
And it's the only way that consciousness will ever life extend long enough to get off of this rock and go out to explore the rest of the galaxy.
Edit---General intelligence. OK, I read the wikipedia. It's what humans do. Kinda anthropocentric isn't it?
I mean really, 'the coffee test???' Someone isn't intelligent if they can't go into a house and make coffee? Nobody outside of Yemen was intelligent before the 16th century?
Quote:Coffee cultivation first took place in Southern Arabia.[5] The earliest credible evidence of coffee-drinking appears in the middle of the 15th century in the Sufi shrines of Yemen.
So how, exactly, does God know that She's NOT a brain in a vat?