AI, a blessing or a curse?

If a company like BT with a million customers tried to use humans to deal with customer complaints that would have to employ tens of thousands of people. Get real.

kelper said:

If a company like BT with a million customers tried to use humans to deal with customer complaints that would have to employ tens of thousands of people. Get real.

Click to expand...

Or they would suddenly find the motive to build a more reliable product.

Maybe the billionaire club will reproduce themselves in AI and beam themselves off planet to live out their digital lives among the stars or Mars maybe

I'm starting to find widespread AI annoying now. Just about every blooming website has a "ask AI" thingy prominently - even clothing websites! I hope and think that eventually these things will just become ignored - like cookie banners. An annoyance to be ignored or clicked out of the way. Took me ages to work out how to turn off the intrusive AI assistant in Adobe Reader/Acrobat.

I like Chat GPT sometimes but I OPT to open that website when I choose - not have it in my face.

I've learned to be skeptical of AI. I've used it to ask technical questions relating to specific IBM mainframe software. It's usually accurate but sometimes it's not, and the answers always look like they've been pasted right out of an IBM manual which is misleading when the information is wrong. To me, for now, it's more like a front-end to a Google search than anything else which probably reflects on the unsophisticated way I use it.

OldMainframeGuy said:

I've learned to be skeptical of AI. I've used it to ask technical questions relating to specific IBM mainframe software. It's usually accurate but sometimes it's not, and the answers always look like they've been pasted right out of an IBM manual which is misleading when the information is wrong. To me, for now, it's more like a front-end to a Google search than anything else which probably reflects on the unsophisticated way I use it.

Click to expand...

To me it's like Wikipedia: an overview that must be recognized as fatally biased and often incorrect in the details, sometimes even in the large. But it can be a marvelous resource to harvest for references to do further research. Just ask ChatGPT to also provide references and sources for its responses since it doesn't always do so.

XxXxX said:

its called deniable plausibility or in the words of the layperson 'not me guv'

Click to expand...

That's a nice way of putting it. I call it juvenile irresponsibility, myself.

And my point exactly...



But the machine did it!

The machine did it!

Not my fault!

Not my fault!

mommy....mommy....mommy....

whaaaa....whaaa....whaaa....

A Guy

ultrajv said:

We don't have AI yet. It's currently just large language models. Great at processing questions, poor at answering. There aren't many alternatives at the moment. Until then, AI is a way off yet. That won't stop people from forcing it's adoption without regard to the consequences. Currently referred to as AI slop.

The bubble will burst and people will get stung, companies will collapse in it's wake.

Click to expand...

Yup.. I've read several books on this subject.. the short story is this:

Until we have quantum computers with a few ten thousand stable q-bits, there is no such thing as a real intelligence. What we have now is nothing more complicated than a fancy parrot.. a talking toaster.. they call them language models as you say.

It will take another 50 years to advance quantum computing to the absolute minimum level we need, then they say another 50-100 years until the software required for such intelligence is created.

And when it is created, it will basically become a god, limited only by the laws of physics.. and even those might turn into suggestions rather than laws.

On a different note, I read a book a few months ago called "Fly Bot by_Dennis E. Taylor". pretty entertaining, very realistic... Probably pretty accurate. He's a good scifi writer.

AI we have access to are "scrapers", information aggregators. But then the garbage in, garbage out applies. It would take a deep dive to filter out the false information, and only provide the correct possibilities. But these days truth, like beauty, is in the eye of the beholder. I see politicians straight up lying these days as if their "alternate facts" are acceptable, so perhaps todays AI learns all too well...

A Guy