1. publishers want to be able to put content on the Web without undergoing background checks
2. everyone wants to be able to discover content with as little friction as possible
3. consumers don’t want to drown in unwanted crap
The incomprehensible Algorithm is the result of trying to square that circle. Give up any of those requirements, and the arms race would end:
Give up #1, and it’ll be possible to do all of the rules enforcement reactively, with no algorithms and no inhumane call centers, because when someone is banned, they’ll stay banned. The ban will be tied to a legal name and anyone caught ban-dodging can be sued.
Give up #2, and it won’t matter how much spam you make available on the web because nobody will fall victim to it. The web becomes less like a publishing platform and more P2P, because you basically only find content on there through your in-person social contacts.
Give up #3, and you don’t need Safe Browsing any more. Good luck selling that to everyone, though.
In order to sue them, you need to come up with something that they should’ve done but didn’t. Having a human review every web page that’s ever published is obviously dumb, so they’re going to have to go with the algorithmic approach.
> Give up #1, and it’ll be possible to do all of the rules enforcement reactively, with no algorithms and no inhumane call centers, because when someone is banned, they’ll stay banned.
This doesn't actually work because the people doing bad stuff are criminals with no qualms about committing crimes, like identity theft. Some large fraction of spam is sent from compromised but otherwise legitimate mail servers.
> Give up #2, and it won’t matter how much spam you make available on the web because nobody will fall victim to it. The web becomes less like a publishing platform and more P2P, because you basically only find content on there through your in-person social contacts.
This is the one you can actually fix because it's a spectrum rather than binary. It's also something that doesn't need to be a monopoly, and not being a monopoly would significantly reduce the consequences of mistakes.
Discovery is also fundamentally a search issue. Not putting something you suspect of being spam in the first page of your search results is a world away from shutting down some guilty until proven innocent third party's DNS or hosting.
> In order to sue them, you need to come up with something that they should’ve done but didn’t.
How so? If you sue for damages, you only have to prove you were harmed by Google's actions, no? And actively misrepresenting your website as dangerous and deceptive to your customers is sort of libelous and clearly damaging.
You’ll at least have to prove negligence if you want to sue for libel (assuming you count as a private figure).
Now, I’m not going to actually say that you’re wrong to claim that Google runs Safe Browsing in a negligent manner. But I will say that, if you’re going to go with that, then you’re going to have to say what they neglected to do. Have a human review all their entries? Apple does that, and they get just as many complaints. Get rid of Safe Browsing entirely? It was created to solve real problems, and those problems aren’t just going to go away.
Google has no obligation to list you on their search results or allow access to your site through their browser.
> ctively misrepresenting your website as dangerous and deceptive to your customers is sort of libelous and clearly damaging.
Except the OP even said someone uploaded a malicious file that was put in a place publicly accessible. Google was not being libelous. There was a malicious file.
1. publishers want to be able to put content on the Web without undergoing background checks
2. everyone wants to be able to discover content with as little friction as possible
3. consumers don’t want to drown in unwanted crap
The incomprehensible Algorithm is the result of trying to square that circle. Give up any of those requirements, and the arms race would end:
Give up #1, and it’ll be possible to do all of the rules enforcement reactively, with no algorithms and no inhumane call centers, because when someone is banned, they’ll stay banned. The ban will be tied to a legal name and anyone caught ban-dodging can be sued.
Give up #2, and it won’t matter how much spam you make available on the web because nobody will fall victim to it. The web becomes less like a publishing platform and more P2P, because you basically only find content on there through your in-person social contacts.
Give up #3, and you don’t need Safe Browsing any more. Good luck selling that to everyone, though.
In order to sue them, you need to come up with something that they should’ve done but didn’t. Having a human review every web page that’s ever published is obviously dumb, so they’re going to have to go with the algorithmic approach.