Pulled from the other thread, a way to delete all of your activity.
When I hit the "remove purchase" button, it even says "To remove this purchase, delete the email."
If you don't trust them to handle the derived object according to your desires, then deleting it should not give you any assurances, since they could just rebuild it from the source material and the code they have on hand.
I think what you're really looking for is a mechanism like, "don't use this derived object in ways I find objectionable." I'm not sure exactly what that would look like for you; it would probably be different for every individual. For what it's worth, at least if you believe the news and Google's public statements, it doesn't use the contents of your email for ad targeting.
The problem is that they are constantly thinking along these lines, and building products to enable this. Also, they are an advertising company, so this is something that is of value to them. Fox guarding the hen-house stuff.
In my mind, its kind of like hiring an IP lawyer at a company. Even if your company is not litigious, there is a chance that at some point the lawyer initiates legal action to "prove their worth" or even advance their career inside the company. Same thing with this, why is Google paying developers to build this product if they're never going to use it? What is the business justification? Its certainly not for PR or marketing reasons.
Having an apparatus ready to go is very different than having no apparatus. This is part of the reason why weapon buildups are bad across the board even if people claim "nobody" is going to use them.
Useful features attract users. Users are a good thing. Like I can apply the exact same argument to email labels:
Why does google let you label emails? They're an advertising company, so labels must be valuable to them. What is the business justification? It's certainly not PR or marketing.
And now suddenly labels sound like an ominous feature, anti-user or anti-privacy or something, maybe. Which they aren't. But the implication is there.
I think its a bit of a stretch calling this feature useful but I suppose you can re-define anything to be useful. Ads are useful too, because you can save money by buying the exact thing that you need. Its too bad that people are downloading ad-blockers by the millions.
Google snooping on emails and extracting purchase related information to stick inside a dossier on you, is in no way the same as you voluntarily, creating a label for an email which they have already scanned anyway. An advertising company wants to know what you're buying, how frequently, and how much you spend on things. That is like a goldmine for them. The labels, are almost noise because its data that is an abstraction away from telling advertisers what will get you to open your wallet - which is the end goal.
>Like I can apply the exact same argument to email labels:
I don't think you can, and I could not follow the argument you made.
> I think its a bit of a stretch calling this feature useful but I suppose you can re-define anything to be useful.
It's used to alert you on upcoming packages you may receive or bills you need to pay (if you have an android phone/google assistant). I think that's useful.
> The labels, are almost noise because its data that is an abstraction away from telling advertisers what will get you to open your wallet - which is the end goal.
Statements like this amaze me! You really think that manually curated labels would be less useful than automated inference about an email? Supervised learning is basically always easier than unsupervised, and email labels are, quite literally, labels for the data. It's a model-maker's dream!
> I don't think you can, and I could not follow the argument you made.
That's because you personally find labels useful and non-invasive, but feel that this other feature is not useful and is invasive. That's the only difference. If you felt you got value from the purchase history tools, which again, some people do, or if you didn't have an unfounded belief that everyone was lying to you and this data was being used in ways it isn't, you wouldn't find what is essentially automatic labelling of some emails invasive.
The argument is that labels are useful to some people. Google could misuse the data derived from labels. Similarly, purchase history info is useful to some people. Purchase history data could be misused by Google. If you trust Google to not misused labels, then why don't you trust them to not misuse purchase history? If you do trust them in both cases, this doesn't matter, and if you don't trust them at all, why are you using an email client that you think is completely lying about how they're using your data?
Its not about a line. Advertising functionality is simply not a core function of an email service. It is something that is tolerated in exchange for a free service. That is not the same as ceding control and letting a vendor do anything they wish simply because a vendor claims its useful.
>It's used to alert you on upcoming packages you may receive or bills you need to pay (if you have an android phone/google assistant). I think that's useful.
Well, how does storing the item purchased and the dollar amount and other purchase details help with tracking anything? AFAIK Amazon for e.g. has stopped putting in tracking information in their emails. There could be a useful function as far as bills are concerned, but its weird to take an email about a bill and then generate another wasteful notification about the same bill. Or to take a package has shipped email and generate another notification. But that is a separate discussion.
Also, if Android and google assistant users have chosen to enable such features with informed consent, I don't see why every single gmail user has to have this this enabled. I mean that in a rhetorical way, I do see why an advertising company would do such a thing. Anyway, I'm simply registering a protest and I encourage other users to do the same. Several invasive Google policies were reversed due to privacy push-backs, and so all hope is not lost.
>The argument is that labels are useful to some people. Google could misuse the data derived from labels. Similarly, purchase history info is useful to some people. Purchase history data could be misused by Google. If you trust Google to not misused labels, then why don't you trust them to not misuse purchase history? If you do trust them in both cases, this doesn't matter, and if you don't trust them at all, why are you using an email client that you think is completely lying about how they're using your data?
That to me, is needlessly polarizing a nuanced issue. "Either you trust us completely or you don't" is not really an argument worth having is it?
They want to give users the ability to inspect the source data for various further derived features they provide, like shipping notifications.
> at least if you believe the news and Google's public statements, it doesn't use the contents of your email for ad targeting.
True, but its important to note that it used to do that for many years. They didn't make it clear if they removed information mined from previous scanning of emails.
That is definitely not the problem. Email you send is owned by its recipient. Generally, the recipient can do what they want with it, including delegating to Google and allowing Google to scan it for purchase information.
> They didn't make it clear if they removed information mined from previous scanning of emails.
That information was ephemeral, IIRC. It was scanned when you loaded an email to allow advertisers to make bids for ads relevant to the given email. In any case, bits rot at Google like anywhere else. If a feature or system is no longer being maintained, it quickly stops working. There is approximately zero chance that they retained any data that they theoretically would have generated from emails scanned half a decade ago.
I don't agree that everyone thinks this is the case.
This is precisely why I avoid sending emails to gmail accounts. Fortunately, almost nobody that I personally know uses gmail.