Meta sent me an email if I wanted "to object to the use of [my public] information" in order "to develop and improve AI at Meta." I objected. First, many of my posts are my own creative work, which I don’t want any company to use for AI. Secondly, AI for search is misleading at best and worthless at worst, with too much false information compared to non-AI approaches. Thirdly, the servers used to feed large language models (LLMs) have appalling environmental effects. Finally, LLMs are based on an understanding of language that is completely contradicted by linguistics. Human language is not statistically generated on the basis of a gigantic database. (Andrew Shields, #111Words, 20 April 2025)