A while back I shared a post, then deleted it after I was persuaded that the image in it had been generated by AI. I do not trust AI and do not consciously use it, for the following reasons. And because I am posting this online, I am confident that some generous soul will point out how misinformed I am.
AI depends on accumulating a vast store of texts from which to generate material. Some of that text is copyrighted, and my understanding is that the copyright holders are neither cited nor compensated.
The ability to generate new texts from this store makes both plagiarism and fabrication easy, leaving the reader with the task of identifying what can be trusted. We've seen accounts of teachers who cannot vouch for the authenticity of the work their students submit, as well as of legal filings that include bogus citations. (In a proper society, this phenomenon would mean more work for editors to ferret out the genuine from the dross, but few care enough about accuracy to engage editors.)
AI requires enormous quantities of electricity for its servers, which leads to increases in cost for ordinary consumers and strain on the power grid. It also requires huge quantities of water to cool those servers, which makes it environmentally questionable.
I am beginning to see articles questioning the sustainability of the project: millions invested in technology that is made obsolete by further technical advances before the initial investment is recouped. This could, these articles warn, lead to a financial collapse comparable to the housing bubble and subprime mortgage collapse of 2008.
So, yes, I am not a cheerleader.
