I Do a Deep Dive on DeepSeek, Part 2

I really dig into DeepSeek's terms and privacy policy, and provide a link to the spreadsheet where I'm keeping track of differences between the six LLM providers I identified in Part 1.

Just in case you didn’t see it, here’s a link to Part 1, where I discuss DeepSeek more generally and touch on three aspects of the ToS that have caught some folks off-guard.

I Do a Deep Dive on DeepSeek

And here’s a link to my still in-progress spreadsheet that tracks differences between policies across the six named companies I mentioned. As I mentioned in the footnotes, I relied on Google’s NotebookLM to scan and analyze relevant sections of the ToSes and Privacy Policies/Notices, which I then review. NotebookLM is a phenomenal tool for heavy research like this, not because it gets everything right, but because it allows me to easily compare details without having to flip between 12 different tabs while I go silently insane.

Instead, I can ask a discrete question (e.g., ‘what is the data retention period’), get an initial answer that not only links to the relevant document, but also highlights exactly where in the text the answer is. Because NotebookLM uses natural language, it also finds similar or comparable terms. That leaves me to do the hard stuff (actual analysis), and saves me from CTRL-Fing myself to death trying to think of every variation of ‘retention’ that might exist. This is, IMHO, an excellent use-case demonstrating how LLMs can augment the research process.