Can AI detect a “glaring error” in a 1976 image?
A user says they use image-spotting challenges to evaluate AI models and claims Anthropic products fail nearly all the time.
TLDR
A user asks what looks out of place in an image they date to 1976, calling it a “glaring error.” They say they use such tests to evaluate AI models, claim Anthropic products fail nearly all the time, and predict the company will use their images to train future models.
Combined views
35.8K
1 Source, first seen 3h ago
likes
