In the current state of AI? I guess it would depend heavily on the job at hand. My concerns with current iterations are privacy ( how do I know sensitive data i

In the current state of AI? I guess it would depend heavily on the job at hand. My concerns with current iterations are privacy ( how do I know sensitive data ingested to feed the model won't be re-used ala co-pilot or gpt for other entities and some information should stay where it is ) and errors which necessitate human babysitting anyway. There is a clear benefit to automation ( in fact, its usually my spiel ), but it is going to happen anyway so I am not sure what the argument is in this instance. Would you be willing to elaborate?