← Back to context Comment by predkambrij 14 days ago tell more please 4 comments predkambrij Reply loopasam 14 days ago by using conversation traces and observed tool calls to capture the human reasoning and logic, and generating synthetic data from that for instance loopasam 14 days ago or simply using chat sessions for analytics and product prioritization (what are the topics users care about? What's trendy? Etc...) jpardoe 14 days ago That sounds a lot like training on your chats. vdfs 14 days ago They can just train on your data, throw it in the pile
loopasam 14 days ago by using conversation traces and observed tool calls to capture the human reasoning and logic, and generating synthetic data from that for instance loopasam 14 days ago or simply using chat sessions for analytics and product prioritization (what are the topics users care about? What's trendy? Etc...) jpardoe 14 days ago That sounds a lot like training on your chats.
loopasam 14 days ago or simply using chat sessions for analytics and product prioritization (what are the topics users care about? What's trendy? Etc...)
by using conversation traces and observed tool calls to capture the human reasoning and logic, and generating synthetic data from that for instance
or simply using chat sessions for analytics and product prioritization (what are the topics users care about? What's trendy? Etc...)
That sounds a lot like training on your chats.
They can just train on your data, throw it in the pile