No, reading personal email to deliver advertisements is creepy. It's better than it used to be but we all remember seeing things like casket advertisements when an email mentioned dead code, etc etc.
It's creepy, and I would gladly pay google monthly to prevent them from abusing my privacy.
But Google doesn't want money from me, it wants to abuse my privacy. That's far more profitable.
It's unfortunate but too many services point to my gmail for me to be able to switch.
Let's define what "read" means in this context. A computer program, most likely a C++ binary, takes the string of your emails, splits on U+0020 and sends the resulting array through a series of algorithms like Levenshtein distance to match to its database of ads.
Text mining has come a long way since 1985, so I don't think that's how it's done (in fact I know it's not done that way), but I guess what you mean is that no human ever looks at the data. That's possible, but not reassuring.
Many of the shady purposes this data could be misused for don't require a human to ever read it. Say, one day, that C++ binary computes a score that indicates how reliable, credit worthy, dangerous, healthy, honest or skilled you are.
Say another C++ binary makes a decision that influences your life based on such a score. A decision on health insurance, banking, job applications, security matters, etc. Is that better, just because no human ever looked at the data and giggled because you searched for some embarrassing disease?
But what business sense does it make for Google to sell the data directly? You can only sell data once; targeted advertising can be sold over and over again.
It could make business sense for someone like Google or LinkedIn or Facebook to provide a service to employers to score job applicants in terms of loyalty, skill, reliability, security risk or whatever. It could make business sense for Google to provide credit risk scoring to banks or lifestyle/health related scoring to health insurance companies. They would not have sell our personal data to do that. They would just compute scores.
Granted it would not make business sense for Google to do that right now. But the future might look very differently. What about a desperate future Google that has been deprived of ad income by regulation happy governments or threatened by anti-trust legislation? What about a Google disrupted and superseded by some other new and cool company, lead by someone put in place by activist investors to monetize its assets?
Things change, but our ability to react to such changes is very limited when it comes to taking back information that has been put out there.
No, but that's a separate problem from algorithmic ads. When Google shares personal data and not merely statistics about personal data I'll being doing the call-to-arms along side you. What Microsoft wants you to think, however, is that there is an office full of people some where reading your emails and picking ads that they think you might respond to.
Some would say that sharing a computed score that claims to say something about me is not sharing my personal data. But I don't even think that Google would compute and sell such a score at this point in time. I trust them to some degree, maybe even a bit more than I trust others.
What I'm concerned about is that data that allows such a score to be computed is out there and I have no way to control it or take it back. Google could change and start to interpret their privacy policy slightly differently. The data could be stolen or accessed by governments. Many things could happen that are outside of my control.
But I agree that Microsoft may try to insinuate something different from what I'm concerend about.
You seem to comprehend how Google runs its business and pays for its services. Isn't it incredibly ludicrous to call it creepy?