Wait, so you trust them to store the email, and parse it for something like spam detection, but not for some smarter feature? What specifically are you concerned about that is a problem with this and not with them having the email in the first place? Sounds like you shouldn't be using Gmail at all tbh.
You trust snail mail carriers not to read your mail.
The digital equivalent of the envelope is encryption - the carrier is allowed to see the outside but not the inside. The function of encryption is lack of comprehension - the carrier still knows the exact encrypted bytes of your mail, but cannot comprehend it. We generally accept that this is already a useful proxy for privacy.
Imagine a mail protocol that somehow lets you encrypt everything but the part necessary for spam identification. The carrier is able to apply some algorithm to determine whether it's spam, but no other information. I'll assume everyone thinks this sounds good (though impossible). Because encryption is just lack of comprehension, an algorithm that is only capable of parsing an "is spam" feature from email has the same level of privacy as this hypothetical. But an algorithm that is capable of parsing "your credit card payment is overdue" is able to comprehend something else, which is a different level of privacy.
(Of course, encryption is better than "They're not running anything that can parse this," because it says "Nobody can parse this.")
My mail carrier can't read my mail because I'm in the US and there are laws against it. There are certainly mail carriers in the world that can and do read mail.
That is a philosophically incoherent model of privacy. The singularity notwithstanding, computers cannot “comprehend” anything. Spam filtering and credit statement parsing are no different. It’s all just ones and zeros being shuffled around.
Everything is just bits of energy being shuffled around. There is no value in seeing computers as being unable to comprehend things, when comprehension is little more than maintaining an internal model of some processes. And computers do that.
Plaintext is also an internal model. The bare fact that they have your emails means they already have an “internal model” of your financial statements before any parsing is done.
> What specifically are you concerned about that is a problem with this and not with them having the email in the first place?
That they use millions of computing-dollars to extract this information and store it.
Supposedly spam detection does not need to understand the numbers in your bank emails. Supposedly it needs mostly information about you previous contacts and spammy patterns from the rest of the internet.
The point being that you cannot misuse the information you do not have. Google is very good at thinking of a lot of this kind of quality of life functionality, they do not seem to realize how creepy and disconcerting they can be.
Actually I want to make a slightly different point. The problem is that "intelligent" systems are unpredictable. In the future google could wrongly decide that your business is inconvenient blacklist you from searches. Maybe it got that idea from your name being mentioned in relation to some extreme stuff and unified you with ISIS.
Computers cannot “understand” anything. You are inappropriately anthropomorphizing an algorithm. There is no epistemological difference between spam filtering and credit card statement parsing. Either way it’s just a cpu pushing around ones and zeroes.
Yes there is, whatever information google extract form an email during spam detection need not be associated with your profile. That is not the case for financial data.
> Either way it’s just a cpu pushing around ones and zeroes.
What is even this supposed to mean? also the information "kill on sight" on your profile of a terrorist association is just one bit, do not worry about that. The fact that they are digital does not mean that the data won't affect your life nor that human or machine will take complex decision based on them.
I think I'm not understanding if you really answered the question you quoted - what specifically is the concern with having that information extracted? Is it just "it feels creepy"?
My point is that they already have the information, so any potential misuse or abuse is possible regardless of what types of parsing they're doing. It might just be inconvenient.
One example is holiday photos: I am ok with stranger taking photos with me in the background. I am not ok with someone spending a lot of effort going around collecting all the photo with me in the background trying to build a personal profile on me.
There is also a security issue, I trust google to keep my email secure, I trust less google to keep my profile metadata secure as selling it is its main business model.
>What specifically are you concerned about that is a problem with this and not with them having the email in the first place
My issue is that federal law regulates what information my bank can share with 3rd parties and allows me to opt out. The ethical thing to do is at least hold oneself to a similar standard. If I had worked on those projects I would have made it policy to ignore emails from healthcare providers and financial institutions outright, since there is far more harm that can be done than good.
I have no problem with data collection when it's done responsibly. But I get the feeling that most developers/engineers working on these features have warped senses of ethics and don't consider the large scale implication of what they're doing, or if they do they just ignore it. Either way, it's the blatant disregard for privacy that bothers me.
But I get the feeling that most developers/engineers working on these features have warped senses of ethics
Nonsense. It's a totally reasonable ethical viewpoint to assume that a user who opts in to their financial institution to send their data to Gmail has opted into Gmail reading that email and trying to provide awesome features on top of it. It's not like Google keeps it a secret when it does this kind of stuff. This is the bread and butter what Google is, and anyone not interested in that kind of thing should choose another email alternative.
Google taking privacy seriously means they treat the data you share with them with respect. In this situation it is you, as a person who does not trust Google, who is apparently not taking privacy seriously when you opt into sending them your financial records.
Do you think customers have absolutely no responsibility over what applications and services they choose to use?
I realize the downvotes will come... But also take the time to let me know what I'm getting wrong.
