So I see where the Trump administration is attempting to overturn California's state protections over the use of AI. They're framing it as opposition to anything that threatens the U.S. dominance in AI.
What poppycock! I guess you can make comparisons between AI and the Trump administration - they both spread misinformation, and are rarely right.
My recent forays into getting Windows automation software to run under Android highlighted the issues with AI.
Virtually all the approaches that Google AI summary suggested, along with the software it recommended, were incorrect. I will admit, it gave me the starting point for using Winlator, but nothing else was useful or even applicable. It was painfully obvious that the AI was just regurgitating information found on the Internet (and we all know how accurate that can be), with absolutely and positively no understanding of the request, or that information itself. The only 'intelligence', if you can call it that, was in the summarizing of the bad information. The old adage - garbage in, garbage out - applies.
I read the other day that AI and other related technology stocks make up up to 40% of the Dow Jones Index (and in only 7 companies). At some point, there's going to be reckoning.
What scares me even more than that, however, is that a lot of people in power actually believe this stuff is real right now. That says a lot more about their intelligence than the technology's.
What ai presents is like gossip, it either has grains of truth or it's a twisted fabrication that sounds true. Like you said, it can provide an easy starting point, but it's very evident it has a tendency of completely making things up - there's been several times I wondered "where did it get that information?" - so go to the links it provides as citations only to discover that the link sources contains absolutely no reference whatsoever to confirm it, it really is quite common for ai just make up stuff out of the blue.
But it very apologetic and admits it's mistakes when you put it on the spot concerning it's errors. I guess that's something.
From what I see especially when you look something up on Google like what we were talking about with the running of Windows programs on Android AI has no intelligence of it's own. It gets it from other sources and puts it in it's own words.
It seems to be generally accurate as I found. Unless you try to fool it. It's a quick way to get some info at a glance.
But of course you can find the sites that give the info you want to know.
I agree that if info is wrong it can't distinguish that on it's own. It's not "real" intelligence.
... It seems to be generally accurate as I found. Unless you try to fool it. ..
Nahhh... Well, maybe generally accurate in many cases.. but still totally unreliable. I'm not sure what you mean by "unless you try to fool it". It's the AI trying to do the fooling.