garbage in, – Speaking at the Babell Literary and Cultural Festival in Porto, Portugal, Margaret Atwood said she used an AI chatbot once—Anthropic’s Claude—and got the wrong answer, describing how a model can confidently mislead when it’s fed the wrong signals. She also cal
Margaret Atwood didn’t treat the AI question like a tech beat. At the Babell Literary and Cultural Festival in Porto, Portugal, the conversation turned to artificial intelligence—and she answered with a bluntness that landed like a verdict.
Atwood said she had used an AI chatbot exactly once, Anthropic’s Claude. She went looking for information about the British detective series Father Brown. Instead, Claude gave her the wrong answer.
“Claude gave me the wrong answer, or it lied. Of course. it didn’t know it was lying because it’s not a human being; it’s a large language model… It had skimmed and sampled a lot of television reviews. but they never give away the ending in online criticism. so it was misled by the things it had read about the show.”.
For Atwood, the mistake wasn’t a fluke—it was how these systems work when they’re asked to perform like a person. She pushed back on the trust people place in them, arguing that a model can’t “understand” the way humans do, especially when it’s relying on whatever it has been trained on.
Her criticism didn’t stop with the technology. Atwood also had sharp words for the people she sees leaning on AI for shortcuts. She called them “opportunists” looking for the easy way out.
“Human beings are not robots, but they are opportunists, so if there’s an easy way to cheat and it’s hard to detect, people will do it… But the thing about AI is that it’s garbage in, garbage out. Even people who use it for business reasons have to check it because it makes mistakes.”
The thread running through her remarks was simple: AI language models can only be as reliable as the information they absorb. In Atwood’s telling. that means scraped. previously published. and potentially out-of-date material—and then the uncomfortable reality that if the data doesn’t contain what you need. the output won’t magically become truthful. The risk, she suggested, is not just error. It’s misplaced confidence.
Margaret Atwood AI Claude Anthropic Father Brown large language models garbage in garbage out generative AI artificial intelligence
So basically AI is wrong when it’s fed wrong stuff… shocking.
I don’t even trust those bots. Like if it “sounds confident” that’s how people fall for it. Also the Father Brown thing—did it actually spoil the ending or what?
Wait she used Claude once and it gave the wrong answer and that proves the whole technology is garbage? I mean, humans do the same thing, they confidently wrong too. Opportunists cheat anyway, doesn’t matter if it’s AI or not.
I feel like she’s kind of missing the point… people aren’t asking AI to be “human” they’re asking it to summarize and save time. But yeah “garbage in garbage out” is real. Still, I’ve seen people online argue the AI is accurate because it cites stuff, then it’s like… just random scraped pages from 2017. So if you’re using it for anything important you gotta verify, but nobody does because it writes fast and pretty.