AI Is A Scam.

Hallucination is something that's supposed happen when you get good drugs. It can be cathartic, spritual, inspirational or any combination of. The last thing you want, though, when consulting an expert, is for them to be tripping balls.

AI has that problem too often. People are being encouraged to use AI for their job's tricky problems, to help them find different ways to communicate, for therapy, even to help them with their homework, this latter case, even at university, even for postgrad theses. Yet, ask one of these 'bots more than a handful of questions, and you'll start to get "hooky" answers. I shouldn't understate it - in my experience, every session goes "pear shaped" withing half a dozen to a dozen queries. Sooner in many cases. Have a really close look at my AI generated site logo...

Some, ah, cute... shall we call them "artefacts." I borrow the term from my former profession as an audio engineer - those things we mishear in imperceptably imperfect recordings, such as "mondagreens" in songs.

The Verge's Alex Cranz has also noticed these "hallucinations." Frankly, and she'd probably agree with me, you'd have to be The Who's "Tommy" to not notice them. She has a good take on it, here. Give it a read, it's as hilarious as the problem of AIs being unreliable and a unsustainable is dangerous.

If you have any serious capital invested in any of the AI corporations, I'd suggest you divest. The article that led me to the above mentioned Ms. Cranz' article, Ed Zitron's latest edition of "Where's Your Ed At?" Ed posits that the jewel in the crown, Open AI, simply can't last beyond another 2 years. There's no way to make a profit from it. Seems the 'bots aren't the only ones getting high on their own supply, the venture capital sector are throwing billions at this industry in, at best, hope it's not a money fire.

Divest! Divest! Divest!