Skip to main content

Hi there

I just have some questions about the new version of the answers feature using the question mark.

We are in the process of testing it out and have noticed that AI doesn’t appear to be able to decipher content held in images, for example cost tables - would that be correct and if so will future developments allow for this? (My general research outside of Guru seems to suggest it can be done).

I am not sure if there are any specific training guides to help us - the only ones I found in the Help Centre related to the previous iteration of answers, so I don’t know if it would still be the same? We would like to learn more about the background of educating the AI when flagging/updating the answers.

I hope some of this makes sense - the AI side is a new world to us that we ae trying to get on board with - any further pointers would be greatly appreciated, thank you:)

Things like tables, etc. should be built out as actual text tables in the card itself. I just did this today! That way Answers can generate great answers from that content.

 

Thank you so much, I just tried it and it worked!

We are moving onto a new CRM soon and will need to upload training materials - I am assuming in this case that the AI element won’t be able to pick up detailed information from screenshots or videos?

Any ideas on a best practice for doing this to get the best out of the AI would be very much appreciated, thank you:)

Reply


Terms & ConditionsCookie settings