Chatbot deployment

[pauljsasmith]

I have access to GPT-3 and looking to create a chatbot to demo. However in order for the chatbot to work i would need to be able to input text and output text to the same area which is constantly calling the code from my local machine. I was thinking of using a text area to create it but not sure if I can populate the AI response text following what the user has said in the same area?

Is this possible or can you suggest another way? maybe text input … label … text input … label … etc

[hugetim]

Maybe using a TextArea could work. You could just add a new line to its text attribute to populate the AI response. But the potentially awkward thing would be that the user could edit the AI response (and their previous entries).

I’d probably use a RepeatingPanel, something like this:

Hangouts-like Chat Design

I created a chat design that looks similar to Google Hangouts chat, with messages in speech bubbles, inside a card that scrolls to the bottom automatically. https://anvil.works/build#clone:X3NLEWNXN7EKTRVA=Z25OY6KF7Q26MPLO4BOUCJJS I guess the innovations are the CSS that creates the speech bubbles (probably considered pretty basic but it was a stretch for me), and the repeating panel template that displays a different format depending on whether the message is from me or the other person. Oh, …

[pauljsasmith]

Thanks that’s very helpful, will check that out