Ask HN: Your company puts its secrets into Google Docs, MS365. Why not AI?
Companies put all of their most secret material into cloud services every day, into Outlook, Gmail, Google docs, Microsoft 365, any number of Amazon systems, cloud based accounting systems, workflow systems, development systems and source code systems.
Why do they trust these yet freak out about putting information into AI systems like ChatGPT and Claude?
one quick reason: if amazon has my data, then they'll work with it. Depending on my AOB (area of business), amazon won't be a competitor for my. Also, they won't use the data in a big & disrupting competition based on customer data, because then no company will use their services anymore for sure. So, there is a slight barrier of data-usage. Also, contracts are in play and for sure amazon does have a clause like "your data is safe. Only you can see the data" in the contracts.
With AI like chatgpt, your data is their business and their business is also the competitors of yours. You can't control who is using it (potentially the whole world), you can't control what & how it is used (potentially everything) and you don't have any control about YOUR data, which is also produced by YOU with YOUR money. AI is trained on that and use it without paying, f.e.
so, having providers saying "we dont use your data" and having AI providers saying "we use your data" - its a trust issue and an issue of disclosing crucial business intelligence so your competitor can save on money building that knowledge..
something like this.
Outlook, Gmail, etc. are (1) very familiar, (2) have always presented themselves as "just like having your own server in your own server closet...but without all the hassles", and (3) offer all sorts of access controls, to (in theory) regulate exactly who at my company has read & write access to which bits of data.
In the current perception, AI offers none of those.