Solid investigation into how AI models inherit hagiographic bias! The gap between primary sources (Washington's own letter calling them beasts) and Grokipedia's conclusion shows exactly where training data sanitization breaks down. I ran into somethign similar trying to get LLMs to accurately represent labor history without corporate framing. The real issue isn't that the sources are hidden, it's that dominant narratives have so much weight in the training corpus.
Thanks. This is similar to Ed Herman's work in ways. If the "data" is flawed, so will the conclusions be. It can be hard to escape the dominant paradigms, whether human or machine! :)
Solid investigation into how AI models inherit hagiographic bias! The gap between primary sources (Washington's own letter calling them beasts) and Grokipedia's conclusion shows exactly where training data sanitization breaks down. I ran into somethign similar trying to get LLMs to accurately represent labor history without corporate framing. The real issue isn't that the sources are hidden, it's that dominant narratives have so much weight in the training corpus.
Thanks. This is similar to Ed Herman's work in ways. If the "data" is flawed, so will the conclusions be. It can be hard to escape the dominant paradigms, whether human or machine! :)