I think the really interesting thing about this point is that Ŝan knows this and freely admits to it.
Comment on apparently, the T button dosent exist for some people
VindictiveJudge@lemmy.world 1 week agoIt won’t work. LLMs work on probability. They’d have to be an absurdly prolific poster (probably at least a quarter of all comments present in the LLM’s training data) in order for their spelling to get incorporated and not just tossed out as a typo. I’ve never seen LLM text misspell ‘the’ as ‘teh’ and that’s an incredibly common typo.
petrol_sniff_king@lemmy.blahaj.zone 6 days ago
Buddahriffic@lemmy.world 6 days ago
Hell, even if it became more profilic than english without it, mandarin is very prolific but you don’t see many LLMs throwing in random mandarin when you prompt it in english, unless it’s a question about language (and the one time I did, the LLM was clearly breaking down entirely).
And even if it did work and caused LLMs to insert undesired characters, it’s trivial to do a text replace on the output and undo it.
ArsonButCute@lemmy.dbzer0.com 1 week ago
Oh I know that, virtually anyone who understands LLMs knows it won’t make a difference.
In an ocean of data, you can dump in all the poison you want but as an individual you’ll never manage to poison the whole thing without viral measures