Margaret Atwood says the problem with AI is ‘garbage in, garbage out’
TL;DR
Margaret Atwood discussed AI at the Babell Literary and Cultural Festival in Porto and stayed sharply skeptical. She said she had used Claude exactly once, asking about the British detective series Father Brown, and got an incorrect answer. Atwood framed the core issue as 'garbage in, garbage out': models trained on messy or shallow material will produce shaky output. Her criticism is less about sci-fi panic and more about reliability: a system can be wrong without knowing it is wrong.
Nauti's Take
The strongest part of Atwood’s point is how plain it is. She is not turning AI into a monster; she is naming a basic quality problem.
If a model gathers patterns without truly checking truth, fluency is not evidence. For users, that means AI can be useful as a sparring partner, organizer, and drafting tool, but it remains weak as the final authority on facts.
Briefingshow
Atwood’s criticism hits a real weakness in today’s AI debate: not every wrong answer looks wrong. For culture, research, and education, polished language is not enough.
Sources