It would be helpful to compile a thread. Eventually I could turn it into a github repo.
This was inspired by seeing the Zoom news (disappointing).
Doesn't use content for training by default:
* GPT-4 API and GPT Playground
* Apple Notes, Apple Reminders
Does use content for training:
* ChatGPT (I think)
* Gmail (I think)
* Google Docs (I think)
* Google search
Ironically, AI could probably parse the LLMs for this. If anyone's looking for a side project, LLM-parsed terms of service that summarize whether a service will use the content for AI training would be helpful.
See also: https://news.ycombinator.com/item?id=34807687