this result is obvious to anyone who knows anything about language models. studies like these are conducted in bad faith to further legitimize the technology for these use cases under the (wrong) assumption that fundamental limitations can be satisfactorily corrected with further investment
this result is obvious to anyone who knows anything about language models. studies like these are conducted in bad faith to further legitimize the technology for these use cases under the (wrong) assumption that fundamental limitations can be satisfactorily corrected with further investment