this post was submitted on 10 Jun 2026
231 points (97.1% liked)

me_irl

7856 readers
3047 users here now

All posts need to have the same title: me_irl it is allowed to use an emoji instead of the underscore _

founded 3 years ago
MODERATORS
 
you are viewing a single comment's thread
view the rest of the comments
[–] Fermion@mander.xyz 41 points 2 days ago* (last edited 2 days ago) (3 children)

Captchas are for free image recognition training, not verifying humanity. My ability to identify fire hydrants is not what makes me human.

[–] lurch@sh.itjust.works 1 points 23 hours ago

There are additional hidden JS checks on those captchas tho. The fire hydrants is just their way of extracting value, so they can offer them for free to the site owner.

Cloudflare, for example, gets their money another way, so they can skip the fire hydrants.

[–] StillAlive@piefed.world 7 points 2 days ago (2 children)

Captchas are to annoy people using VPN.

[–] TankovayaDiviziya@lemmy.world 2 points 2 days ago

Two of my apps probably blocked me temporarily for using VPN. It's sorted now but for months I was not able to log in to my shopping app and I have missed out on some deals. My gym membership was also inaccessible for two months.

[–] atomicbocks@sh.itjust.works 14 points 2 days ago (2 children)

When captchas were first a thing computers couldn’t do them yet so for a very brief period of time they were effective at filtering robots. I can believe that some poor Google engineer from 2009 or whatever didn’t realize the implication of what they were creating.

[–] Dave@lemmy.nz 4 points 2 days ago* (last edited 2 days ago)

Around 2023 when a bunch of Lemmy instances got stood up after the reddit debarcle. Most instances were still open sign up at that point, then the Lemmy devs removed the (basic) captcha from the signup page.

That day we learnt that even the most basic captcha still stops a LOT of bots.

[–] zikzak025@lemmy.world 6 points 2 days ago* (last edited 2 days ago)

The earliest ones that asked you to type two words were also used to help digitize the Google Books library, for cases when the OCR wasn't quite able to match a word confidently. That's why you'd get a pair, one that was clearer but "jumbled" to be unreadable to computers, the other looked more like a fuzzy scan (or occasionally just a complete mess of misplaced ink), though both had a wavy effect applied to them. Only the clearer one was needed to pass the test, the other was just free training for their OCR process.