Generative AI has developed the habit of publishing my municipality and the City I live in whenever i google a query (often even though I haven’t asked anything to suggest I was looking for something in that particular area.)
I wonder how it is allowed to make such an invasion of privacy?
I asked AI, this is what it replied:
I’m not aware of having ‘frequently mentioned location based queries’.
My location setting is turned off.
Suggests being ‘aware of how my data is being used’. AFAIK it’s not made clear that it includes publishing my location on the World Wide Web.
Although I goggle my searches I’m not signed in the Google account
(No matter how many times I’m invited to do so) but I notice that in order to minimise AI replying one needs to get into the Google account.
It will most likely come from IP, connection point and device data as well as through history of data collected. Welcome to metadata collection and its use.
And it is allowed as terms and conditions as well as use agreements spell out how data is collected and used.
When you’re not signed in to a Google Account, we store the information we collect with unique identifiers tied to the browser, application, or device you’re using. This allows us to do things like maintain your preferences across browsing sessions, such as your preferred language or whether to show you more relevant search results or ads based on your activity.
I do understand that there’s information being collected, even though I’m not signed into a Google account. What I object to is AI making a public reference to the Municipality and the City I’m in. One example, a while ago I asked about a population estimate of an Australian town, and the reply included the population of the municipality I’m in and naming it, a totally irrelevant piece of private information.
Generative AI is now admitting that the responses may include mistakes, and not only that but ‘Made-up facts, Statistics, or even entire Articles’!
So, AI is up there, the first reply to any googled query, and I have to wade through mistakes and made-up facts? Easier to ignore it and move down to replies by well known reliable sources.
It can be full of mistakes and not relied upon itself. As a good friend said recently about AI (writes code for on of the big US tech companies)…
rubbish in = rubbish out. One rotten apple can spoil the whole box.
Everything should be checked against reliable sources. AI can compile information, but can’t (yet) determine the reliability of such information.
Hence ‘mistakes’.
It possibly is a bit like ‘Chinese face’ (miànzi (面子), which anyone who has lived in China understands. Ask a question, and one will get a response even if the answerer might not know the right answer. Response is given to save ‘face’.
