3 ms·
The training data could include internal docs that describe how it ignores or not the robots.txt file.
by gtirloni 4y ago
The training data could include internal docs that describe how it ignores or not the robots.txt file.
- amluto 4y agoIf I were involved at OpenAI, I would not include the internal wiki, Slack archives, Dropbox folders, etc in the training data. While it would be highly entertaining, it would not be a good idea.
- iKevinShah 4y agoI agree on that - that private data (in a best case scenario) should not and would not be included in the training but there would be some parts of internal documents which would be public (lets say public website) - It is expected that chatGPT would know at least those ..