AI hallucinations, and other post-apocalyptic stuff

“'The AI world’s hallucinations may be a small blip in the grand scheme of things, but it’s a reminder of the importance of staying vigilant and ensuring our technology is serving us, not leading us astray,' Kostello said."

Shades of HAL9000, right? So, if we overwork AI chat bots, they get loopy? Three headed cats loopy? Critical misinformation loopy? 

Very interesting. I've been reading a lot about AI chat bots, partly because I have students who sometimes like to take short cuts, and also because I'm a wordsmith myself. It seems that the chat bots, like GPT Chat and Bard, get stressed out. They have dark emotions. (What??) Great, we've created digital Poes and Plaths, without the talent or vision. 

The computer gurus assure us that these things are fixable ailments, but in the meantime, corporations are going nutty over this new plaything, and the practical applications of such technology are now suspect. I do not want to alarm anyone, but if this type of tech is going to be put in charge of our safety, help scientists develop new drugs, or --heaven forbid-- be given control of the nuclear codes, we are in for a rough go of it. 

But seriously, though, when a tool requires this much vigilance to make sure it is not creating huge errors, then it is not a tool. It is a busted toaster. 

I'll keep my pen and paper, thanks. My phantasms will be mine alone.

Take care,

C