What does work is to type @gemini and WAIT. Then a popup appears and you can select which model and it replaces it by the actual thing that counts for detecting LLM usage. (Another bug: note that the icon is the wrong size for one of the gemini models):
Until very recently @chatgpt hidid work, but Harald removed the old code that detected that sort of thing, which was of course potentially brittle. So it's probably best to just change the placeholder text completely. Thoughts?
I did just change it to
from
since "a LLM" is not correct english, and also we don't want to omit humans from being mentioned -- that's very important! Finally the "...)..." looked funny.
@chatgpt hi
or@gemini hi
What does work is to type
@gemini
and WAIT. Then a popup appears and you can select which model and it replaces it by the actual thing that counts for detecting LLM usage. (Another bug: note that the icon is the wrong size for one of the gemini models):Until very recently
@chatgpt hi
did work, but Harald removed the old code that detected that sort of thing, which was of course potentially brittle. So it's probably best to just change the placeholder text completely. Thoughts?I did just change it to
from
since "a LLM" is not correct english, and also we don't want to omit humans from being mentioned -- that's very important! Finally the "...)..." looked funny.