Crew AI build AI agents team, anyone interested?
Added 2024-02-25 19:40:45 +0000 UTC
Testing some features here.
I will do some practical script that run for some real example.
:)
Comments
By using prompt, the response of the LLM are just 1 time processing that an AI model generate content. Verse in AI agent, we are able to set role for different processing in a workflow. For example, agent A write an article, agent B review and point out some problem and giving feedback to agent A to improve the article. Then agent A are able to edit the draft and make changes. This way, a lot of researchers have found the quality of generated content are higher quality. And this is not just for article, it can be making a mobile app, create reports with data.
Benjamin Law
2024-02-26 04:41:01 +0000 UTCI have seen examples of things like having a researcher, a writer, and an editor. But I think the question is what makes sense to accomplish in a prompt vs a handoff to a separate agent? What if you had a researcher, a first draft, second draft, a fact checker and final draft with different editors at each stage. Is it better or just more expensive?
Andrew Luetgers
2024-02-26 04:21:52 +0000 UTCGreat guys :) give me some scenario if you have any. Like for example, we can use AI agents of 2 LLM and talk to each other, create a blog post of a topic. Or go all detail in a research , and summerize stuffs for us. etc.
Benjamin Law
2024-02-25 19:44:48 +0000 UTCDefinitely, saw some interesting things about it, I’d be curious to hear your thoughts comparing it to the other systems you have tested
Dungeon Master
2024-02-25 19:42:32 +0000 UTCWould love to learn to build one
Good Dude
2024-02-25 19:42:13 +0000 UTC
