Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Proton spam and the AI consent problem

dbushell.com

Post reply on HN
Print view View original

391–400 of 446 posts

Re: Proton spam and the AI consent problem

#391

Post by supriyo-biswas »

microtonal wrote: post #260
hojiron wrote: post #131

Earlier quoted context omitted.

Still happy that Tuta Mail is anti AI, and does not push ads on you via email. I wonder who told Proton that it’s a good idea to copy big tech tactics.

* I wonder who told Proton that it’s a good idea to copy big tech tactics.* But people subscribe to Proton because they want to move away from big tech. What’s the point of paying them if they get as bad. Though for now I’ll assume that it’s a genuine mistake with things not properly escalated by customer support.

With customer support positions, escalating to engineering is also seen as a negative metric. They might blame customer support for this but it’s likely that they’d have been turned away with “why are you escalating this stupid thing to us?”

Re: Proton spam and the AI consent problem

#392

Post by jedahan »

davidee wrote: post #231

I've been using proton for a year after migrating from Rackspace and I'm done. Not because of this article, but I might as well pile on: 1. I use a custom domain. Turns out that there are two competing features, not-at-all documented. If you use a catch-all, like I do, AND use specific addresses for sending, the two are incompatible to some degree. Which is bonkers. Example: with a catchall I can create any address I…

I do the same for years and had no idea that disabled emails are blocked, is that documented anywhere? Will have to switch if so.

Re: Proton spam and the AI consent problem

#393

Post by chrisjj »

tmarice wrote: post #388
chrisjj wrote: post #165

Earlier quoted context omitted.

The FM Masked Email is insecure in that there is a circumstance under which it can leak your real email.

Do elaborate.

"WARNING: Fastmail Masked Email insecurity" https://www.emaildiscussions.com/showthread.php?t=81287

Re: Proton spam and the AI consent problem

#394

Post by willis936 »

davidee wrote: post #231

I've been using proton for a year after migrating from Rackspace and I'm done. Not because of this article, but I might as well pile on: 1. I use a custom domain. Turns out that there are two competing features, not-at-all documented. If you use a catch-all, like I do, AND use specific addresses for sending, the two are incompatible to some degree. Which is bonkers. Example: with a catchall I can create any address I…

I want to emphasize just how bad bridge is. It's so bad that they nuked the issues section on the repo. They didn't lock it, they removed all record of issues. I found out when trying to click on a useful/familiar issue that was cached by a search engine. Proton says to talk to their support directly, but that is a totally different solution for a totally different problem. The fact that they deleted public history of comments really flares my unsubscribe feelings.

https://github.com/ProtonMail/proton-bridge

As for the "why is bridge bad / why were you searching for issues": keeping it logged in on a headless server is an exercise in pain. It will latch onto whatever keyring it feels like then fail to integrate. Okay, capitulate and do it through the GUI. That works until the token expires. So you're expected to log in every few days for email backups? I only have so many weekend hours I am willing to waste troubleshooting with an llm before I say "fuck it, I'm going somewhere else".

Re: Proton spam and the AI consent problem

#395

Post by chrisjj »

vee-kay wrote: post #389
chrisjj wrote: post #375

Earlier quoted context omitted.

But the suggestion "get marked and reported as spam" can lead to future mails getting junked before even reaching the spam folder.

Agreed. That "Mark as Spam" facility not only moves the offending message into Jink/Spam folder, it also allows the Email Service provider to identify that type of email as spam, so future incoming messages that match that may criteria can be categorized as spam, so they'll go into spam folder automatically, rather than into the Inbox. You can find them in the Jink/Spam folder. However, if thousands of users report s…

> (Emails from known Contacts are auto-dumped into Spam folder)

Oh?

Re: Proton spam and the AI consent problem

#396

Post by chrisjj »

MiddleEndian wrote: post #390
vee-kay wrote: post #389

Earlier quoted context omitted.

Agreed. That "Mark as Spam" facility not only moves the offending message into Jink/Spam folder, it also allows the Email Service provider to identify that type of email as spam, so future incoming messages that match that may criteria can be categorized as spam, so they'll go into spam folder automatically, rather than into the Inbox. You can find them in the Jink/Spam folder. However, if thousands of users report s…

>However, if thousands of users report same domain or sender as spam, then the email service provider may take stern action, including blocking the sender email id or domain at the server level, so their messages will never reach your mailbox. This is a good thing. If you spam thousands of users, you are a spammer, even if you also happen to send legitimate emails. If anything, it should be applied more broadly. When…

> This is a good thing. If you spam thousands of users, you are a spammer

Or you got hacked by a spammer.

> even if you also happen to send legitimate emails.

And also a bad thing. E.g. for the user losing a critical legit email.

> if Gmail marked all their emails as spam until they stopped, that would be a major quality of life improvement for everyone.

Sorry absolutely not for everyone. To me, receiving legit PayPal email is far more important than being protected from PayPal spam, prevented from employing my own protection.

One size does not fit all.

Re: Proton spam and the AI consent problem

#397

Post by chrisjj »

MiddleEndian wrote: post #322
chrisjj wrote: post #313

Earlier quoted context omitted.

The disconnection was the fault only of the spam filter hiding the service mail.

I mean if said company first spammed you and you marked them as spam, then it is on them. No different than if someone sent you a bunch of unwanted letters and you threw them out, but one of them happened to be relevant. It's on the organization sending you junk.

This is not you marking them as spam. This is "all email providers (especially Gmail, the largest provider) should mark ALL of their emails as spam".

Re: Proton spam and the AI consent problem

#398

Post by chrisjj »

solid_fuel wrote: post #333
chrisjj wrote: post #315

Earlier quoted context omitted.

Agreed, but the false positive I am referring to is the cathching of the non-spam message from the source of the previous spam message.

They shouldn't send marketing mail from an address they want to be read. I think that's been the standard for a while, in practice - most actual transactions come from orders@ or something similar while marketing mail comes from a dozen other addresses.

Agreed, but many do. They want all mail to be read. Worst offender here is a bank.

Re: Proton spam and the AI consent problem

#399

Post by Citizen8396 »

chanux wrote: post #361
bl4ckneon wrote: post #234

Earlier quoted context omitted.

You also get simplelogin for free, give that a try. Will probably fix your first issue

Does SimpleLogin only support ProtonMail addresses? This was my impression the last I tried and hence moved on instead of fighting it.

No.

Re: Proton spam and the AI consent problem

#400

Post by Bender »

brauhaus wrote: post #3

This is not an AI problem, it's an "data privacy + lack of consequences problem". It happens everywhere. I mean, have you ever tried making an airline company to stop sending their shitty miles newsletters? Only way to stop is to start fining these companies.

Only way to stop is to start fining these companies.

There is a way to fine them regardless of where they are operating from. Get them on the DNSBL/RBL sites such as uceprotect, spamcop, spamhaus, etc... There are many others. They are still used to this day though indirectly behind the scenes instead of outright rejecting email from those listed. They affect spam scores and are also used by some commercial server products. In some cases this is still a fine regardless of regional laws because one has to pay to get removed immediately rather than waiting for the penalty period without more reports to pass. Uceprotect is well known for this. Some see them as extortion sites and I love it. Spammers should absolutely be extorted to send more UCE.

Post reply on HN