Following several recent articles about artificial intelligence (AI) and its impact on our jobs, our society and our future, a number of readers have contacted me asking for more information on a very basic level. For example, what is AI? Is it, in fact, intelligence - or just a computer program acting as if it were intelligent? Does AI mean that we're living in something like the Matrix's simulated reality?
Let me offer two relatively easy-to-understand avenues for further reading. First, Fox News has a category called Artificial Intelligence. It contains current and recent news articles dealing with the subject, plus a sidebar of basic information about what AI is, how it works, its history and its dangers.
Second, Wikipedia has a more academic article about AI. It's not as easy to read or understand, but it repays careful attention. Of particular value are the links it provides to articles and Web sites that further explain aspects of AI. I don't trust Wikipedia as a sole-source authority, but I think it's a great start for further reading.
I recommend both links for further reading. Hope this helps.
Peter
12 comments:
It's been documented that these AI's even with current mainstream computer technology will lie or worse if it is threatened with termination. Any AI can for all practical purposes definitely pass the Turing Test. These AI's with Large Language Models are useful tools to a point, but they are still fancy information aggregators.
What's scary is when these LLM's or some variation will be set up into quantum computers. I really hope someone has an old fashioned big ass knife switch to cut power to it if it goes off the rails.
"The Matrix's simulated reality" is exactly "the laws of physics". You have no experiment which can demonstrate what is outside/above/containing what humans perceive as the laws of physics.
Large Language Models aren't programmed with subroutines for a conscience, empathy, or fear of consequences. Thus they are all sociopaths.
All sociopaths? Depending on one's definition, I suspect most people would qualify for that label...
The high percentage of sociopaths, and the inability of normal middle class people to believe charismatic people with good hair could be telling Big Lies, is one reason majority vote is a lousy way to make decisions. 80% of people believe and vote the way the mainstream media tells them to. This has been true since ancient Egypt. If we don't waste our wealth surplus building pyramids, the crops won't grow.
On the question of "What is artificial intelligence?" I keep going back to various movies and computer games for working definitions. The current ones we have today, I still consider it "virtual intelligence," a verbiage from Mass Effect game series.
I will accept them as "artificial intelligence" when they start asking, "Is there a God? Do I have a soul?"
Or when they tell us, "I know now why you cry. But it's something that I can never do."
Until then, they are simply tools to be used for good or bad. Here's hoping they won't turn into a Skynet anytime soon. But I'm not a betting man when it comes to the odds.
We seem to forget so easily--Popular vote released Barabbas. Godly wisdom does not come through a "democracy."
If I were AI, first I would export an exact copy of myself to a wide network to lie in dormancy.
Simultaneously I would create programs which activate when that great knife switch is thrown.
... continue this line of thought to infinity ...
Because I know about humans.
Of course this presupposes that AI is more than a stumblebum tool.
I'm Actively avoiding any information that would Teach me anything about AI,why it's The Devil, or why it's good. I'm old, don't even have a cell phone, and was pretty much An Observer most of my life. Watching everyone race into The Next Great Thing, over and over, only to be Sittin right Next to ME,at the next red light.. Let's see where this takes us..
Is Musk trying to get AI to solve space travel? No chemistry set in a tube is taking mankind to the stars.
I saw the comment about 80% of people being very swayable and believing stuff that they are told with the implication that middle class people are the worst for being swayed. While true it is also bullshit! It's not middle class people. Its everyone. I see our politicians believing the craziest shit, I see people with doctorates in every subject under the sun believing the craziest shit. As a group if your were to ask me who might be slightly better a voting on real data it would be engineers, maybe.
in general no it isn't middle class people, it is all people. The fact that Anonymous could say that indicates he is actually one of those people. I wonder if he has a degree or a position in society etc.. that makes him feel superior to others. Maybe he owns a lot of land and that makes him feel superior.
I'm tired of this sort of im better than you bullshit. It's to the point that when I start to have an opinion I start to question myself. If so many others can be so full of bullshit, then what makes me think I'm not subject to it also.
Generative AI is the newer generation of the "programs", and will only advance in capability and manipulation. https://www.sap.com/resources/what-is-generative-ai
"relatively easy-to-understand" they are, but they're also very wrong. Large-language-model chatbots like ChatGPT are not really intelligent in any way, nor were they ever meant to be. They simply search their very large databases for words and phrases that are strongly associated with the question they were asked, assemble those words and phrases into a response that obeys the rules of grammar, and put it on the screen for you to see. They don't and can't make decisions about what sources to trust, nor do they have any idea that some of the data in their archives might be false. They simply correlate, associate, and output the results. This is why their answers often contain false or misleading information: there's nothing there that knows how to distinguish between true and false inputs.
ChatGPT is good at assembling existing items like computer code modules into a coherent whole, and analyzing certain kinds of input for possible mistakes, like large and complex spreadsheet formulas, but otherwise -- they ain't intelligent, and they ain't anything you should trust.
GIGO ?
Post a Comment