This is an inevitable consequence from the push to make cloud companies responsible for the content they host. Analyzing all of it can only be automated; automation will deliver weird false positives.
I made a Nextcloud to host my references and storywriting notes. Nextcloud is horribly buggy, difficult to maintain, and has very poor user experience compared to Google Drive - but at least I don't need to worry about my entire digital life being auto-terminated by an overzealous robot, with no reasonable appeal process beyond "knowing enough people to make a stink".
Why do people always forget the 'cloud' is just someone else's computer.
Nothing important should ever exist solely on the 'cloud'.
I wouldn't even trust a backup copy sent to email, yet I see that constantly as well.
https://news.ycombinator.com/item?id=31652650 ("I've locked myself out of my digital life")
The article you referenced was getting locked out of accounts. We are speaking of having documents backed up, not google accounts.
The best way period would be to periodically backup to an external HD or even SD cards.
People need to get over this cloud bullshit years ago.
All of these problems have been solved since the advent of the internet, these are all self-imposed problems.
and where do you store those?
you really want multiple backups, and an encrypted cloud storage is one of them.
Honestly, that seems like a feature of many administrative processes, even those administered by government. So far, in the process of trying to become licensed as adoptive parents, my wife and I have been dropped twice now by licensing agencies, without any explanation of why or recourse to appeal the decision. As far as I can tell, the only way to gain the right to appeal a decision is to actually be convicted of a crime.
Probably my most memorable lesson in unexpected failure modes for complex, automated software.
Back when Gmail was invite-only, most people used Hotmail or Yahoo Mail, which would offer a base level of (I think) about 50-100MB of storage at the time. If you wanted more storage; you'd have to pay for your email service - an idea I don't think anyone born past 2000 has even heard of.
Gmail came along, and suddenly offers us 1 GB of free storage, for free. This was around a quarter the size of some people's hard drives at this point.
Of course we didn't care if there was no support!
Are you kidding? I don't think Hotmail had support, either - (maybe I'm wrong?) - even if you paid for it; but here Google was offering a substantial value for anyone using any other existing mainstream email service, immediately.
They even had STMP and IMAP support, meaning you could use whatever client you preferred. Back in the day most people used computer-side email clients and had local backups of emails anyway - so if one went missing from Google's side it wouldn't be so bad.
There's a reason Gmail is still one of Google's strongest and unusually longest-lasting product of Google's. There's less reason now - but my God, when it was introduced it was an honest-to-God mindfuck as to how they were offering such a large amount of space and features for nothing.
The selling feature of Gmail was "Now you can keep all your mail and have a permanent searchable history of all your electronic communication".
"In-Q-Tel scouts the global market for commercially focused technologies with the potential to contribute to national security." - Source: iqt.org
Of course, now we all understand that they were harvesting all the personal data this gave them access to, in order to monetize it, and sell access to it to anyone who would write them a check.
And then we found out that this information was so valuable, and in such a concentrated place, that the CIA placed fiber taps on Google’s data center drops.
Also, can't they distinguish a spam account from a legit account based on the account history? I can't see why this is a difficult problem to solve.
Have you ever tried to fight spam at scale? It’s trivial to block the 50% of obvious spam accounts, but we are at the stage of humanity where the activity of very dumb humans and the activity of very clever bots, EASILY overlap. A lot.
Bots will use proxies that use residential IPs, they will maintain session info (Cookies, User-Agents, etc) they will move the mouse and introduce jitter, their access patterns aren’t random but are engineered to work in the day night hours that match the geo-data of the IP they are using. They solve captchas, at the same rate that humans do (time, error rate)
Some human accounts look like bots. They sign up and upload a couple of files which immediately get high traffic. They use NordVPN so their public IP is shared by thousands of “known bots”, their access patterns are weird and unpredictable.
Yes, you can use machine learning to try and identify theses but then you end up with false positives and real people having problems like the poster above.
And on top of all that, bots are constantly subverting detection so whatever solution you have now won’t work next week.
To avoid detection, you can't just make API calls for six months. You have to run the official client on the machine for six months, and then the official client can collect more data on your usage pattern. Imagine the cost to run hundreds / thousands those accounts.