Something I didn't share during the weeklong frozen pipe saga (mostly followers only) is that a few people recommended ChatGPT to get ideas.
"It's great! You can talk to it..."
I cringed inwardly. I think it was inward. I wanted to scream, but did not.
"Yeah, I work in tech. I know how it works and I won't go near it. Search worked fine. The Bob Vila website had good ideas."
This conversation took place at my local building supply store's project desk.
One of the staff piped up: "Bob is great!"
The acolytes weren't interested.
I don't currently technically work in tech, but close enough. I have IT training. I work in the online business world. I used ChatGPT a few years ago b/c a client required it, before I knew what was really going on.
Anyway. I didn't have the spoons to explain the multiple harms, or the hilarious and horrifying inaccuracies which were particularly relevant in this situation.
Inaccuracies like the viral story about using glue to stick cheese to pizza.
Or the lesser-known story about cleaning the wound and applying a bandage before putting a turkey in the oven to roast – the "intelligence" totally missing the mark on roasting a turkey with dressing. (I suspect this is cultural, "dressing" vs "stuffing.")
And then there's the story, published on Futurism, where ChatGPT recommended mixing bleach and vinegar for cleaning – which creates potentially lethal chlorine gas. How many people would know that?
The lesson: Never, ever trust a large language model aka "AI" with anything that could harm you, your family, furkids, vehicle, or home. Never.
There is zero intelligence here. It's scraping, blending, and predicting. It's a bloody damn algorithm.
Also, "AI" is a vague term that applies to many things, coined in the 1950s. It's devolved into a marketing term that benefits the billionaire bros to the exclusion of everything else that could be beneficial.
And now I will use my soapbox for kindling. So tired of all this shit.
#AI #AIslop