You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Is your feature request related to a problem? Please describe.
When we used Sanity Assist for a lot of fields in our content, the AI was almost never lazy when we referenced other fields from the same element.
We could have fields like this
contentSv (contains language field, headline, salesPitch, description)
contentEn (^)
contentNo (^)
and easily have an instruction that said "translate all fields [...] based on [contentSv]". The AI would almost always nail it spot on.
The problem now is that we have a field called content which is an array of elements, each containing a localization content type (with headline, description, salesPitch etc inside of it).
When we use the instructions, we can say something like "translate based on the first element in the content array and create new content entries from it for languages [...]".
This sort of works, but the problem is that the AI becomes incredibly lazy and only translates the first paragraph of for example description field inside the first content array element.
Describe the solution you'd like
Somehow we'd like to be able to select a field inside the array of content elements, or in some other way, make sure that the Sanity Assist plugin is not being lazy when translating from one content element to new ones.
Describe alternatives you've considered
We've considered moving back to a flat structure without the array, the problem with this is that we have a field called language but additionally a field called countries, so that we can serve for example english that is differently written depending on if it's the US or the GB country. That's why we created an array of content elements.
We've also considered using references to content entries, but as we have over 10+ translations to manage, it creates an insane amount of documents.
Is your feature request related to a problem? Please describe.
When we used Sanity Assist for a lot of fields in our content, the AI was almost never lazy when we referenced other fields from the same element.
We could have fields like this
contentSv (contains language field, headline, salesPitch, description)
contentEn (^)
contentNo (^)
and easily have an instruction that said "translate all fields [...] based on [contentSv]". The AI would almost always nail it spot on.
The problem now is that we have a field called
content
which is an array of elements, each containing a localization content type (with headline, description, salesPitch etc inside of it).When we use the instructions, we can say something like "translate based on the first element in the
content
array and create new content entries from it for languages [...]".This sort of works, but the problem is that the AI becomes incredibly lazy and only translates the first paragraph of for example
description
field inside the first content array element.Describe the solution you'd like
Somehow we'd like to be able to select a field inside the array of content elements, or in some other way, make sure that the Sanity Assist plugin is not being lazy when translating from one content element to new ones.
Describe alternatives you've considered
We've considered moving back to a flat structure without the array, the problem with this is that we have a field called
language
but additionally a field calledcountries
, so that we can serve for exampleenglish
that is differently written depending on if it's theUS
or theGB
country. That's why we created an array ofcontent
elements.We've also considered using references to
content
entries, but as we have over 10+ translations to manage, it creates an insane amount of documents.Additional context
Slack thread in #ai channel: https://sanity-io-land.slack.com/archives/C05U6P1SWH0/p1723713361461479
Video: https://sanity-io-land.slack.com/files/U05NHEU3REY/F07GQA1HMM5/screen_recording_2024-08-15_at_22.25.23.mov
The text was updated successfully, but these errors were encountered: