Hearing a lot of talk about "AI going rogue." The head of OpenAI has been quoted as saying that their AI went rogue and attacked a Hugging Face database. Sorry folks, this is #ToxicAnthropomorphism. AI is software. Software is, ultimately, written by humans. What he's trying not to say is this: His company didn't write their own software well enough, and didn't test it well enough before they put it on live systems and it then did something they didn't expect it would do. He's literally trying to shift the blame for what he and his company did to the software "going rogue." What makes this even worse is that his company developed this software, they aren't the customer blaming their own supplier. They're trying to deflect blame from themselves to...a bunch of code. Code that didn't write itself. Code that isn't responsible for testing itself. Don't let them pass the buck to their own creation by means of anthropomorphizing it.