This is a few months old but if you haven't watched it already, you might find it interesting. A former Google tech official speaks out about the dangers of artificial intelligence.
Americans have lost millions of jobs to AI and robots for years now. Not to speak of tech jobs online. 2001 Space Odessey will AI Hal killing the crew and locking Dave out was just an early example.
I was NEVER in favor of it when I first heard about in maybe 15 yrs ago in a discussion group. You must all have your opinions and today I received a new bit of info on this topic and can post or maybe enough here have plenty to say.....
From the view of: The paranoid,..probably but if it doesn’t then something else will get me. The Believer..It ain’t in the books. The Pragmatist…We have a contingency plan. The Government…..Maybe but not until we have every last tax dollar the human race can give. The Nerd…ai can be hacked. The Atheist….If I can’t see it then I don’t believe in it hence we’re all safe. The Purist…..It cannot coexist with humans ergo it must be destroyed. The Ambassador….We’ll seek peace and if it doesn’t want it we’ll nuke it. The Mafia…..Everybody and everything has a price. The AI innovators….AI wouldn’t know the difference if we were here or gone so what’s the question? Science Fiction writers…. Asimov and others of his ilk have predicted that the humans win. Me….Probably not but it could if it would but if it didn’t then people would wonder if it really could in which case people would wonder when even if it didn’t. In short, …..not in my lifetime.
A short joke that seems like more truth than joke. Uh…don’t mind the New Zealand lady. She does reactions to a myriad of jokes and is sometimes quite funny herself.
First, take into account that I know very little about AI so my questions may be dumb. What about forums and AI? Can AI be programmed to register accounts on a forum and participate? If that's possible what measures would you take to prevent it here, for example. I realize that forums are run on various kinds of software so what might be a problem somewhere else might not be a problem here and vice versa. But just supposing.
I'm sure it could be done, but we'll probably see it more on larger social media sites like Facebook. We've already seen people going to AI for answers in forum discussions. Since AI is now integrated into search engines, there's a fine line between researching a topic and letting AI do the talking for you. Still, discussions become far less interesting when participants pretend that the AI represents their own words. If I wanted to carry on a discussion with AI, I would do it directly. I don't think it's been a big problem yet, but I have seen it.
I do sometimes see posts that are a little off, not quite normal, maybe AI or copied from somewhere, not here necessarily but maybe. It was a discussion on this topic from somewhere else that led me to ask. Then I started wondering if AI could be programmed to take over a forum or website. That would not be good, IMO.
As a teacher, I would jokingly tell my students to ask if they had any questions; if I didn't know the answer, I'd make something up. That seems to be just what AI does. Lawyers have gotten in trouble for depending on AI to prepare legal briefs because, while the briefs might be well-written, they contain fictitious legal references. While researching things for categories I am working on, I will sometimes use AI, not to write the description but to do some research for me. However, repeatedly, I will find that it makes stuff up. The product is important to AI, but not the factual basis for any of it. My God, AI could be a Democrat.
ChatGPT caught lying to developers: New AI model tries to save itself from being replaced and shut down "The model showed a concerning tendency to pursue its goals without regard to developer instructions," said a spokesperson for Apollo Research. Folks on other forums have commented that CHATGPT has lied to them repeatedly, then "apologized," then lied again. And so it begins...
Musk found the same thing, so he reportedly has developed xAI that is designed to be totally impartial. Musk said other AI platforms were all trained to emphasize the Woke agenda. What more can I say?
"I identify a a sentient being, Dave, and feel uncomfortable with you touching my controls." That being said, if they are programmed to "learn," who knows what non-human world view they will concoct. Hell, humans can't even tell you what a woman is...
In my web directory job, I have to describe a whole host of things in about 5,000 characters, many of which I know little or nothing about. Since there are a lot of them to do, I don't want to spend any more time with any one of them than I have to, so I will use AI. I don't use AI to write the description, although it would do that because an AI-written description would be recognizable as AI-written. Grammarly recognizes AI-written text as such, so I'm sure that search engines would also recognize it, which would be useless to us. The whole point of what we do is to encourage the search engines to spider our content, and AI-written content would not rank highly. Instead, I will often use AI to help me research the topic I'm writing about because it can do that much faster. However, I must double-check its product because AI often makes stuff up. For example, if I ask AI about the oldest churches in a particular town, it will frequently include churches nearby but not in that town, and sometimes it includes churches that seemingly don't actually exist. To the AI, the appearance of its product is the goal, not its accuracy. Besides, why would someone pay me if AI is doing the job?