I’m getting really tired of “verify you’re not a bot” checks everywhere.
I understand these checks help prevent bots, but they seem to appear on so many websites now. Does anyone else find them frustrating?
In the last 24 hours my small personal website got 8 legitimate visitors and 74,000 requests from bots, most of which were either malicious or aggressive ai crawlers
Wanna make that 9?
The internet is for bots
These days? Fuck no. I think it’s like 57% of all web traffic these days is just bots. This is absolutely needed.
Cloudflare is getting rich from all of these bots. We had turn to Cloudflare at work because bots were scraping our website and essentially DDoSing. Cloudflare worked perfectly and rate limiting bot scraping.
I wouldn’t be surprised if they make some of the bots they stop. It’d be like the people who bred snakes to get money for killing them in India.
You can add “Strengthened Cloudflare as the Internet’s single point of failure” to the long, growing list of awful things which have happened due to this AI boom.
I run a reticulum NomadNet site. I setup logging, since there are no ‘sessions’ and I don’t want to force everyone to identify, I just calcuate bot traffic based on time, order and frequency of loads, bots’s scraping are likely 80% of my traffic. And this is for a service that’s designed to support 200bps radio links.
You may not understand just how prevalent bot traffic is. I self-host a small personal website that real people only actually visit once in a blue moon, but I get constant attempts to access the Wordpress admin console (my website isn’t made with Wordpress lol)
Same here. And I don’t even offer a public page.
It’s proportional to the amount of rude AI crawling, from what I’ve heard.
The industry has put every reasonably accessible piece of human writing in their datasets, by now, and are scrounging for the remaining little bits. Old, obscure, obsolete paper books nobody ever reads are apparently a hot item. So, yeah, your latest blog post is probably worth some fraction of a penny, and that’s all it takes.
The anti-bot checks are awful for usability, and the bots that keep spamming everything are even worse
Yes, but I feel good knowing there’s an AI bot owner out there more frustrated than I am.
I’ve just considered it a casualty of the privacy wars.
I don’t know if they’re deployed as a “weapon of annoyance” or anything, but I do notice that when I’m using a) a degoogled browser, b) an ad-blocking dns, c) mask my location, and/or d) block my IP. Suddenly (almost) every website makes me wait by throwing that at me.
I think it’s lowkey trying to annoy me to the point where I give up on having those privacy measures, but they underestimate my hatred of corporate fuckery.
I’d guess your client just looks more like a crawler that way and gets a JS challenge thrown its way
As others have said, they are absolutely necessary in this day and age.
But also the bots are not deliberately targetting lemmy, it’s just that lemmy has unlimited pages to scrape and so the traffic piles up. If your issue is specifically with Lemmy then using an app should help. Many instances also give logged in users a free pass.
I could use an app, but since I don’t know English, I need to use a translator to understand what people are writing.
Ah well I’d suggest for Lemmy make sure you’re logged in. Many instances will bypass checks if the right cookie is set. Not always the case, as each instance would need to implement the bypass themselves. I’m not sure if Lemmy.world has done this.
On browser at least, lemmy.world – OP’s instance and my own – only requires a bot check if you’re not logged in, even if you’re behind a VPN (I am). I never see a bot check unless I am going in via a different browser or device where I am not logged in.
Cool, thanks for confirming.
Bot checks will only be for the website in a browser. Apps connect to the API, and you can’t really put a visible check in a silent background request.
The fediverse also is built on computers talking to each other, things like cloudflares bot blocking will sometimes kill federation because ultimately it’s bots talking to bots. But it would make no sense to connect to the API for scraping, if you are specifically targeting lemmy by calling its API, then you might as well just make your own instance, and then all the other instances will freely send you everything.
Makes sense. I knew bot traffic is an immense problem but this thread is up to date and has been informative. Thanks for explaining.
Im ok with them as long as theyre not demanding input by clicking everything that is a fire post so they can train their ai for free
They should just use Anubis. At least we then see cute anime characters while the bot check runs.
I’m even more annoyed by anubis. Letting me wait in the browser for an enternity for some check that a bot with a native implementation can solve in milliseconds.
Proof of work my ass.
It sure can, instead of doing 50 requests. A slowdown is all that’s needed to make random scraping uneconomical.
https://github.com/mikenrafter/anubis-fast.mirror
This is a sample of a much faster anubis solver (end user facing).
There’s versions. of this for people running bot networks - it’s just a few cheap cloud cpu cycles to invest. Proof of work is an ineffective defense against spam in the age of cheap and scalable cloud compute - and even moreso if the botnet runs on residential proxies.
My point is, the slowdown is NOT making random scraping uneconomical anymore, it does hurt legitimate users a lot more.
Tbh I never had to wait much with anubis, neither with my phone or computer; websites that make me wait a lot (like the arch wiki IIRC) then give ya access without having to do the check again;
Also, Anubis unlike Cloudflare does not block me just cuz i have NoScript installedHow modern is your phone? Anubis is killing my old Samsung A72 from 2020 - long waits, draining battery. On the laptop it’s less of an annoyance, I’ve got the laptop connected to mains anyway and it is fast enough.
honor magic 6 lite, it’s from 2023
That’s interesting, the honor 6 lite should be 20-30% faster than my old A72. Noticeable but not a lot. Maybe it’s the Firefox browser I’m still clinging to, and I should finally give in and change to Google Chrome.
I use IceRaven so i doubt the problem is firefox
Thanks, good to know. That’s the same JS engine.
I don’t like it, but I will say that it beats that period where I was forced to constantly do captchas
I dunno, Anubis never gave me too much trouble, meanwhile that fucking “Gandalf” thing they have on some Nitter/Redlib instances keeps getting stuck on my device, and if you reload the page you get insta blocked 🫠
Using a VPN? They trigger like whoA
Just a few moments ago I passed an anti-bot check with my VPN on.
Yeah you can pass the check but it’s everywhere. It sucks but I guess that’s the price to pay.
Man, if the checks feel excessive you’d be astounded to know how the bot frequency feels 😂
Seems like a complaint a bot would have.
Fuck, so now anyone who’s different is a bot or AI? Seriously, if I were a robot, why would I write a post complaining about websites? If I were AI, I’d have better things to do.
That’s exactly what an AI would say to try to look like a person.
Dude, you’re watching too many science fiction movies. Even though artificial intelligence might be quite advanced these days, if I were a robot I wouldn’t waste my time talking to humans. I’d think of something worse: ending the lives of these poor, brainless mortals. You’re crazy, man :-P
That’s exactly what a robot would say to look like a human.
Fuck you 😾
They are just messing with you
