also hereI’ve been trying to think of a legitimate use case for LLMs in my life for a whole week now as a personal challenge, and still can’t think of anything… maybe next time, companies should invent the appropriate problems *before* inventing the solution
ohhh… after three months… I think I’ve found an actual legitimate useful application of langle mangles.It may be possible that language models could be used to identify the use of an incorrect quotation mark on an internet website and automatically have an algorithm fix it to be the correct quotation mark, better or more conveniently than any other technology that I know of. maybe it could even be used to correct usages of ‘phased’ to ‘fazed’ when appropriate
not sure this was really worth all that funding but maybe they’re not quite so useless after all