![]() What's pretty neat about that is the ChatGPT integration can remember conversations, so in certain instances you can build workflows with a memory. Say, for example, you're building a Zap that has ChatGPT answer questions that are posted in a specific Slack Channel. You can set up the rules or basic knowledge (whatever company info or guidelines you want ChatGPT to abide by) within the Zap's instructions field. ![]() Then, you just add a word to the memory key field (like RULES-THREAD_ID). Now, every time the user asks a follow-up question in the same Slack Thread, ChatGPT will remember the user's previous conversation history. If you build other Zaps with ChatGPT and want them to have the same context, all you need to do is add RULES-THREAD_ID into your memory key field and ChatGPT will remember those specific instructions for your new Zap.įor any workflow you set up with the ChatGPT integration, you can provide it with more tailored training instructions and incorporate important company information like FAQs, writing guidelines, or other rules you want ChatGPT to incorporate into its answers. It's a pretty handy way to instruct ChatGPT to have better conversations versus just one-time prompts. On the other hand, the OpenAI integration (which uses GPT-3) enables you to generate text but does not have built-in conversation memory. You might use the OpenAI integration to summarize meeting notes, create email and Slack responses, or draft up articles. It also has multiple actions (in contrast to ChatGPT), which are broader and incorporate more of OpenAI's features like DALL.E or Whisper.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |