The point of captcha is for Google to track users, not to prevent bots.
this post was submitted on 20 Aug 2023
353 points (98.1% liked)
Mildly Interesting
17133 readers
2 users here now
This is for strictly mildly interesting material. If it's too interesting, it doesn't belong. If it's not interesting, it doesn't belong.
This is obviously an objective criteria, so the mods are always right. Or maybe mildly right? Ahh.. what do we know?
Just post some stuff and don't spam.
founded 2 years ago
MODERATORS
No, the point is to train the AI.
That's a side project, and doesn't explain all the services that use it without images.
Regardless, you always have to connect to 3 different domains - typically the website itself, google.com and gstatic.com. These 3 domains allow for very accurate triangulation across the internet, and each one will fingerprint your browser. They might not know your bank account number or social media account name, but they know that someone using your browser banks with this bank and has an account on that social media, along with thousands of other data points.
Not anymore it ain't. If they were to keep using it that way, bots' scores would go down.
It’s so obvious it’s a scam too because of how vague some of the “challenges” are. “Click all the squares that contain a bike” and then you do it in every conceivable way AND IT STILL WONT ACCEPT YOUR SUBMISSION. Stupid
Have you considered that you may be a bot?
G A S P I’d never thought about it that way before. I need to go see a mechanic now to check
Cease all motor functions, enter Analysis mode, turn the other cheek.
Have you ever questioned the nature of your own reality?
Well, apparently, statistically speaking he is more likely to be human according to the original post. We'll never know.
I’ve had a Captcha that required an image like this one to be classified as a bike
Ceci n'est pas un vélo
*un vélo
Fixed, thanks
Well, to be fair, it is a depiction of a bicycle.
Incredible, honestly. Whatta crock 🤣
I have a suspicion that it might depend on a few factors that might be related to Google’s ability to track, such as the browsers you use and whether you’re on VPN. I saw a post on Lemmy the other day (can’t remember where) that they suspect captcha is more annoying on Firefox, whereas fine with Chrominum-based browsers, esp Google Chrome. I think I sometimes experienced this as well, but I didn’t keep track.
Yep when using a VPN I get way more captchas on Google web searches (regardless of browser).
It makes it really annoying to search on Google, to the point I want to turn off my VPN or use other search engines.
It's also likely that it is because they are seeing a large number of searches from one IP (the VPN exit server), and that can look like bots.
For search engines, I’ve never used it but there’s whoogle that’s supposed to be proxy for google search.
Here's the thing about all search engines that aren't Google (including the ones "based" on Google results). They all suck for regional results in many non-Latin languages.
I had one today where it was click on all the squares that contain a bus. One square very clearly contained a truck, not a bus. When I didn't select that square, it told me I was wrong.
They're so infuriating.
Sounds about like me every fucking time lol
I’ve never in my life failed a captcha. Maybe you’re just bad at identifying bikes?
I just failed five by doing them correctly. It was sure there were more bicycles, but it was motorbikes.
sigh Could be 😔
Very simple. You solve it too fast, it thinks you're a bot and you fail it. Click slowly and you will pass it every time.
So does anyone know if there is a bot I can load on my phone and PC to auto solve these and save me the annoyance?
I've always thought that CAPTCHA also checksfor time, e.g. if the correct solution is entered too fast, it will determine that you are a bot. This would mean that a fast bot would not really "win". Not sure if this is the case or why I've thought it is, but it came to my mind from this 😅
The captcha with the checkbox tracks a lots of items like mouse movement, browsing history etc. Full list is not public but they use all their spying tools on that one
When I added the contact form on my website, I wanted to add reCaptcha so it would protect me from spam emails and make it easier for people by not having to click on pictures of fire hydrants. Reading a little about the privacy concerns about reCaptcha, I dropped it all together. I am in the EU, have a small numbers of visitors, it'a B2B thing so it was not worth it.
I wonder if Cloudflare Turnstile is any better
I always had the impression that only checks for ddos and bot activity. Might be wrong tho.
The entire point of a captcha is to stop bots, so a Turnstile verification stopping bots would qualify
Makes sense. Thanks for the info!
I've seen those puzzle piece captchas (common on crypto exchanges) that always say I completed the puzzle faster than 99% of users (even if I didn't do the puzzle particularly fast). I presume those 99% are all bots attempting to hack into crypto wallets.
Or, maybe, the 99% is a lie
For any podcast fans, the latest episode of Secretly Incredibly Fascinating is all about CAPTCHA and very interesting.
Off topic, but it may also interest people to note that their previous episode was on bras.
I haven't listened to either yet, but apparently I need to set aside 2 hours of my time and not just 1.
Is there a Bra conspiracy out there or just history and information.
Just history and info, still a good podcast though!
Both are good! Hosted by Alex Schmidt for any old Cracked fans.
Plot twist: Getting it wrong proves you're human. Now tell me which pictures have bridges.