There are so many ways to do human verification that have worked for years. The biggest reason bots are plaguing the internet is because these corpos don’t really try anymore. There’s literally no reason to do face scans or IDs other than to unanonymize people and take their data.
Name one way
How about at a random interval once every couple months it will ask you to draw a picture of a cat in the browser and if it finds your drawing process too similar or the image too similar to one that’s already in the database it will flag it without telling you. Three strikes and your out kinda rule. It’s like drug tests but for the internet.
Even if you did, like, a line across the screen to save time there’s no way in hell it’d be the same as anything else in that database unless you are extremely unlucky.
Sorry dude, but this comment makes it clear you have no idea what you’re talking about. If I’m given an assignment to write a program that generates a huge variety of novel cat drawings, which won’t register the same in this database, I bet I could personally single handedly have that done in a week of work.
Some other problems :
-
Many real humans will draw roughly similar cats. Unless you compare the drawings pixel by pixel, or very precisely, then there will be lots of false positives.
-
If you then decide to compare the drawings pixel by pixel or very precisely, then I can trivially generate new drawings by altering them. I have plenty of time to reverse engineer how your database is comparing things even though you don’t tell me when I get a strike because I can make lots of fake accounts via hiding my IP address and run experiments by feeding different variations of drawings and seeing which ones eventually get banned.
-
If you ask once every couple of months, then my bot gets to do three times whatever that duration of months is worth of damage and misinformation before it gets banned. That’s a pretty long time, especially for a bot. I can deploy new bots much faster than your system gets rid of them.
-
As Anivia pointed out, I can pay real humans to draw the different cats for my bots when needed. This is especially trivial because I can just ask them to draw a whole bunch of cats in advance. I don’t even need to wait for the cue from the system. I can just have a big library of pre-drawn cat drawings that I upload as needed.
Cool, you don’t have to be a dick about it. :P I already know that I don’t know much about but I wanted to make a suggestion because the guy asked.
Sorry, I wasn’t trying to be a dick, just being honest. I felt it was important to correct your unjustified confidence in the original comment where you claimed many such solutions exist
But why do you need to talk down to me like I’m an idiot instead of just correct me politely? I’m genuinely curious because it seems to be a very common thing nowadays where percieved confident incorrectness and ignorance is met with distain and anger. I didn’t mean to offend anyone when I made this comment and was simply brainstorming while bored at work.
Because you made a dumb claim with confidence. You were talked down to because you were acting like a fool. Which you’re still doing btw.
-
Spez (Reddit CEO) just put out an announcemen talking about how they’ll verify humans via a new rectal probe in collaboration with Meta AI that 3D scans your log factory.
That sounds tempting, but still won’t make me rejoin reddit.
The only way I’ll ever rejoin Reddit is if they fire that piece of shit, Spez, and every other piece of shit who had a hand in the monetization of their API access which destroyed third party apps like Apollo; and if they change that monetization, either making it free or making it so you have to be a paid Reddit subscriber to have expanded API access.
They did the whole thing with their API completely backwards, on purpose, to shut out the third party apps – when they could have still been able to make money by doing it properly and not alienating a lot of their userbase.
And now that Reddit is effectively a right-wing cesspool of lies and bullshit, just like Twitter has become, even if they fix what they broke, it may not be worth rejoining.
or making it so you have to be a paid Reddit subscriber to have expanded API access.
Isn’t that what they’re doing? I’ve used Relay Pro via API until recently by paying for it.
The only reason I don’t use that anymore is cause I went de-googled, so no play store subscription and thus no API access anymore, and since I can’t stand the original app I came here.
They made the prices so insane that most 3rd party apps couldn’t justify the higher subscription price
No, they made the third party app developers pay the license fees, not the users. If they changed that since I left, then that I’m ignorant of. It made zero sense to force payment from the developers and every sense to simply create a two-tiered API where maybe only a read-only front page works (no commenting) unless you’re a paid subscriber to Reddit, which could be available via an authorization scheme on the logged-in user. It’s simple to do. But they decided (at the time) to screw over the third party developers directly by forcing them to pay.
Thing is, even Christian (the guy who made Apollo) said a nominal fee makes sense and he would have been fine paying that, but he would have had to pay literal millions of dollars within like a month’s notice in order to keep going, and have to pass that exorbitant cost onto his users who ALREADY paid him for yearly subscriptions/etc.
Simply put: Reddit should only have charged users directly, via subscription to Reddit, in order to use a fully-featured API irrespective of which client they use it through.
Spez is a greedy piece of garbage.
