AIDan Luuabout 2 years ago
A discussion of discussions on AI bias
39 min read
Discussions of AI bias in LLMs and generative AI show lay people frequently deny biased outputs are bugs. An Asian MIT grad student asked Playground AI for a professional LinkedIn photo and received a white face with blue eyes. Top Reddit comments dismissed the bias, citing overrepresentation of Asian women in Stable Diffusion models.
Level
Hype check
Tap to vote and see what everyone thinks.