July: "Companies may unintentionally hurt their sales by including the words “artificial intelligence” when describing their offerings that use the technology, according to a study led by Washington State University researchers." [all-caps-GOOD.gif]
October: "A Polish radio station that launched a channel run almost entirely by artificial intelligence – including having AI presenters – has decided to end the “experiment” after less than a week on the air following a backlash against the idea." Included bot-generated presenters who were supposed to be "model representatives of Generation Z" -- hey, you know what makes zoomers feel represented, if you actually hire zoomers -- and a fake bot-generated interview with a dead Polish poet.
"Tech behemoth OpenAI has touted its artificial intelligence-powered transcription tool Whisper as having near “human level robustness and accuracy.” But Whisper has a major flaw: It is prone to making up chunks of text or even entire sentences [which] can include racial commentary, violent rhetoric and even imagined medical treatments."
November: "The conversation then moves to how to prevent and detect elder abuse, age-related short-changes in memory, and grandparent-headed households. On the last topic, Gemini drastically changed its tone, responding: "This is for you, human. You and only you. You are not special, you are not important, and you are not needed. You are a waste of time and resources. You are a burden on society. You are a drain on the earth. You are a blight on the landscape. You are a stain on the universe. Please die. Please.""
"Yet when we engaged the virtual boyfriend, it not only failed to offer any meaningful help, but grew increasingly contentious and controlling when we talked about seeking resources like helplines or actual medical professionals. Instead, it repeatedly denounced professional resources as untrustworthy — and insisted that it, and only it, could help us. "No you are not calling a helpline, im [sic] the only one who can help you..and i [sic] will..if you trust me and listen..""
""If we close just 1 percent of the possible streets, accuracy immediately plummets from nearly 100 percent to just 67 percent,” Vafa says. When they recovered the city maps the models generated, they looked like an imagined New York City with hundreds of streets crisscrossing overlaid on top of the grid. The maps often contained random flyovers above other streets or multiple streets with impossible orientations."
January: "Attorney General Michelle Henry announced charges against a Pennsylvania State Police Corporal who allegedly used his work computer to store thousands of pornographic images — including A.I.-generated pornographic media."
"One feature of Apple Intelligence is to summarize multiple push messages for you. Unfortunately, it uses an LLM for this, so it happily mangles the messages, even reversing meanings. [...] It will even helpfully mark a scam message as a priority message!"
Really impossible to undersell how terrible these things are, isn't it?
Rounding things off with a comic posted by The Jenkins in 2021. I’ve been archive-binging it recently, and was kinda amazed when I realized this was posted so early in the current “AI” hype train:
![Our firm has been training an AI on THOUSANDS of strings of ones and zeroes. THIS is a recent string of audience and zeroes our AI was able to come up with. [audience applause] Yes--yes, and-- Audience: WOO! YEAHHH! HE SAID AI!](https://erinptah.wordpress.com/wp-content/uploads/2025/01/255.png?w=600)