Maybe Antonio Damasio might have something. He believes that emotions cannot be ignored in a full understanding of consciousness and mental processes.
He's absolutely correct. The research that's been done with people literally unable to feel emotion (usually due to brain damage) shows that our decision-making process is actually based far more in emotion than reason. Those without the capacity for emotion tend to become completely indecisive, unable to make any decision without huge amounts of prompting which suggests that we are far more emotional beings than rational ones and further, that much of our "rational" reasoning is actually self-justification after the fact.
How you create a computer capable of feeling emotion (short of rigging up a rudimentry limbic system) is beyond me.
He who oppresses the poor shows contempt for their Maker, but whoever is kind to the needy honors God. ~ Proverbs 14:31
And therein, lies the question. No matter how sophisticated we can make programs these days, we cannot yet program a computer to exhibit will. How we do so is rather beyond my understanding but the people doing the serious research into it at places like MIT might have something for you.
Maybe Antonio Damasio might have something. He believes that emotions cannot be ignored in a full understanding of consciousness and mental processes. It seems to me this might have a close connection to the idea of will. And I do not know how you can program in true emotions. We experience emotions not because of the higher processes of our cerebral cortices; they arise in the deeper (and evolutionarily older) parts of the brain. And they seem to have as much to do with chemistry as wiring.
A computer can reason and store and process prodigious amounts of information. But I do not know how it can be made to want to do something with the passion people bring to their lives.
The thing I question in AI is, how do you, at some point, program will into a machine? This is a key step for an intelligent machine, and a truly independent machine, the sort of machine people worry about "taking over." Humans, including our brains and our consciousness, have will programmed in at the most profound level, simply by evolution. The simplest creature on the planet has more will to do anything than the most powerful machine we have today. It is one thing to program a machine to play Jeopardy; machines will have truly arrived when, one day, some machine decides on its own that it would like to be on Jeopardy.
And therein, lies the question. No matter how sophisticated we can make programs these days, we cannot yet program a computer to exhibit will. How we do so is rather beyond my understanding but the people doing the serious research into it at places like MIT might have something for you.
" It is not capable of going beyond it's original program which is traditionally the boundry of true AI and at which point, we need to start worrying."
" It is not capable of going beyond it's original program which is traditionally the boundry of true AI and at which point, we need to start worrying."
The IBM supercomputer Watson won its second "Jeopardy!" game in Wednesday's edition of the TV show, completing a sweep of its two human opponents
In a bit, I may share a piece of personal info which bears on this, but in the mean time-
Anybody see the shows?
Any comments?
Are we indeed doomed to be overtaken by the machines?
It's not a triumph of machine over man, it's "man's created machine against a man itself".
It's really man vs. man.
It's the programmers and the designers of Watson against the minds and reflexes of their opponents. It's like racing a car vs. a runner, it's like putting a computer against someone doing long division. It's like someone with a rifle vs. someone with a rock. It's like having a competition as to who can plow a field fastest - a person by hand with a hoe (or even without), or a John Deere tractor?
I can go on and on . . .
I predict that there will be further competitions like this, but there will be two categories of contestants: people vs. people ("organic"), and computer vs. computer. Who is to say that other engineers cannot come up with a better "Watson" ?
A few points here- from a first hand perspective (Ok, I was myself on Jeopardy many, many years ago...)
In spite of its overwhelming performance, Watson is not yet close to passing the Turing test. Some of the wrong answers (as well as the consistently overcalculated wagering) could never have been mistaken for human responses.
First hand, there are three keys to Jeopardy:
- An exceedingly large knowledge base. This Watson demonstrated. They tell you, before you go on the show, that outside of a few stock categories (e.g., state capitals, Academy Awards, US presidents) it makes no sense to study. Most of the stuff, you either know or you do not. Watson's data base was obviously very deep and broad. But- a machine that size can cram (or be crammed)- you just keep keeding it information. Nothing profound there.
- Quickly making connections between perhaps very disparate things or concepts. This was the heart of Watson's programming and I have to say it was impressive. This is a world different from the computer which beat the world's chess champions. That was nothing but number crunching in a very limited context. Here, the machine needed to parse the question, figure out context and the categories relevant to the subject at hand, search its massive data base, make the connection, and decide on an answer. Really well done.
- Ring in quickly. From observing the game, I would say things were slanted towards the machine here. Humans have a given reaction time, which cannot be bettered. In fact, when you play the game, Jeopardy is much more than knowledge, it is very similar to a sport where your reaction time is as important as what you know or how fast you can figure it out. A machine is not subject to any such constraints. Mostly, if Watson knew the answer, it beat the other contestants to the buzzer.
All in all, though, it was an impressive step. But, there is still much farther to go until Watson or its successor is ready for that Turing test. The thing I question in AI is, how do you, at some point, program will into a machine? This is a key step for an intelligent machine, and a truly independent machine, the sort of machine people worry about "taking over." Humans, including our brains and our consciousness, have will programmed in at the most profound level, simply by evolution. The simplest creature on the planet has more will to do anything than the most powerful machine we have today. It is one thing to program a machine to play Jeopardy; machines will have truly arrived when, one day, some machine decides on its own that it would like to be on Jeopardy.
Seriously though, Watson is a very sophisticated data search-and-retrieve system. In that regard, it's not all that different from the NCCC or PROFILER (computers used to store and organise the vast amounts of data generated by major criminal investigations). It is not capable of going beyond it's original program which is traditionally the boundry of true AI and at which point, we need to start worrying.
He who oppresses the poor shows contempt for their Maker, but whoever is kind to the needy honors God. ~ Proverbs 14:31
The wonder is not in the winning of the game. The wonder, if there is any, is in the CREATION of the game to begin with. And no "super computer" has yet shown the power to create from scratch.
Democrats think the glass is half full. Republicans think the glass is theirs. Libertarians want to break the glass, because they think a conspiracy created it.
That a supercomputer would win over two people seems--ahem!--a no-brainer to me. I wonder why anyone was dumb enough to think that people would have half a chance against a device that not only can store far more trivia and factual knowledge than any human but can also access it super-fast without such human lapses as stating nearly the correct answer or forgetting form-of-a-question.
This smacks to me of putting on American Idol people who clearly can't stay on pitch or sing on key "for the entertainment factor."
Guess we've gone well beyond thinking a fair and honest competition between well-matched opponents is what makes for topnotch entertainment.
