Kialo requires JavaScript to work correctly.
Should there be strict rules regulating the data used to train LLMs?
Loading Discussion
Selected thesis
There should be strict rules regulating the data used to train large language models (LLMs).
Pros
Cons
Pro claim 1
LLMs produce biased and discriminatory output based on their input data.
Pro claim 2
Many LLMs are trained on data taken without permission.
Con claim 1
Limiting the use of data for LLMs could make them less functional for users.
Con claim 2
Limiting the training data for LLMs would be a barrier to developing this technology.