There is going to come one day that companies (and any website owner) will wish they didn't use AI and instead used a simple (maybe even plain-text) website. That will come when people really need information from that website but can't get to it.
Heck, you can have your information as clear and concise as possible, but the moment I sense that "AI-generated" texture, I am immediately closing the tab. This is just extrapolating from how I feel about the bevy of obviously AI-Generated flyers and banners that have been gaining a lot of traction in real life.
I wish that was the answer to all of these people setting their spouse, their grandmother, their boss loose with LLMs to build sites. It should just be acceptable to have a simple site that conveys information well, and then we wouldn't need to boil the seas to generate pages of shiny slop.
Doubtful. This is just the cost of doing business. 99.9999% uptime for a fraction of the cost is totally fine for all businesses. Government should be the exception though.
Absolutely this. Although I like the shiny and polished looks, getting the info out is the most important, and it is hard yo beat pure black-on-white text
Because they slow down and often block access to websites, particularly for people that try to avoid being fingerprinted or otherwise tracked. Pretty evil behavior.
It is very challenging to distinguish individuals interested in privacy from bots acting maliciously or with reckless indifference. If you devise a way to do this more effectively than Cloudflare, you should start a business to sell this as a service.
My employer is a small business that has an e-commerce website that is attacked by fraudsters trying to validate stolen credit cards or obtain customer information hundreds of times per day. Operations like CloudFlare are the only way to foil these actors. Just trusting you is not a viable strategy.
I had this problem too for all my clients e-commerce websites, then solved it 3 years ago with very little code changes.
Happy to share it with you. Using small businesses for credit card testing is one of the most evil things on the Internet, so anything to stop it is worth it.
Your rate limits on adding and removing credit cards? Your input sanitization? Designing your system to not just disclose details around anything but that relevant to a logged in/authenticated user?
There are many practical ways to handle that sort of thing that isn't Cloudflare. It just requires thinking and a bit of dev time.
t. Been there, done that, cartels used an app to try to launder money through loyalty programs. Management was deadset against doing the one single thing that would make it impossible to do that at scale.
Ulterior motives abound everywhere but especially behind people claiming X is the only answer. Fingerprinting is far more intrusive than just only allowing one to add at max 2 cards a day per user.
> Your rate limits on adding and removing credit cards?
All those requests will appear from different ip's and different browsers, made by someone who can spend months on trying to defraud you. How do you differentiate this from valid customer who happens to try to buy something between 20 tries by bots?
> Your input sanitization?
All those fraud requests will give you valid credit cards which will work perfectly, but then defrauded people or banks will try to chargeback later.
> Designing your system to not just disclose details around anything but that relevant to a logged in/authenticated user?
They can register as normal buyers.
> It just requires thinking and a bit of dev time.
And they can spend months trying to outthink you, then will drain your service in 4 hours when you are asleep.
> Management was deadset against doing the one single thing that would make it impossible to do that at scale.
So, did you actually ever implemented and checked a good solution? Cloudflare isn't perfect, but not everyone has resources to implement their own solution that is better than cloudflare.
> Fingerprinting is far more intrusive than just only allowing one to add at max 2 cards a day per user.
The fraudsters will appear as completely new users each time, adding only one card and making one purchase.
> All those fraud requests will give you valid credit cards which will work perfectly, but then defrauded people or banks will try to chargeback later.
How are they getting pass 3D-S?
If they are able to get past it, then your liability drops off.
Yes it could be designed better, but that is a separate discussion.
It doesn’t matter that they can’t get past 3D-S, because the whole point of what they’re doing is checking what security features are enabled for a card, and whether their address and other validation data will pass. The fact that the scammers are testing so many cards that fail gets you banned from those payments providers, whether or not any payments go through. And this is ignoring all the attackers using bots for other purposes such as taking control of the website to obtain user or client data.
The technique is simple: don't publicly serve expensive (CPU/RAM-wise) pages. A simple blog like the linked article can be 100% cached and served to anyone without needing CloudFlare.
Performance optimization for website already is a business. It's just that product managers mostly don't care and optimize for other metrics (eyeball retention, SEO, etc).
Don't listen to these hecklers, you have every right to your privacy, and should demand it. Anonymity is paramount if people want to be able to talk without repercussions. Gaslighting, heckling and other forms of harassment are to convince you to self censor. Don't take the bait. Ignore them.
> Because they slow down and often block access to websites, particularly for people that try to avoid being fingerprinted or otherwise tracked. Pretty evil behavior.
Bots and scrapers and hackers also try and avoid being tracked, which is by far a bigger problem for them and most websites than the 15 of us using tons of antifingerprinting techniques. Evil? No, that's silly.
Hanlon's Razor is especially effective at getting Good people to put out their own eyes to keep them nice and soft targets for the malicious. I'll take being a harder to find likable by people I have no desire to be liked by to make myself a harder mark. The honest ones will understand. The malicious were never worth being close with, and I'm doing the world a service by getting the borderline enlightened.
Sufficiently advanced stupidity being indistinguishable from malice is also something to keep in mind.
Getting past cloudflare can be a problem if you are from a less reputable country or a VPN or if you use a less conventional browser.
There is also a larger problem of a private entity being the MITM for a large portion of internet traffic. But i don't think this is the point GP was trying to make.
It regularly blocks me too. I've begun search Google for: what does [url] say about [search query]
It'll make a Gemini summary of the page, and can be prompted for more details. It's really the only use for Gemini I've found reliable. I'd still rather directly view the output of the scraper bot, though.
If anyone knows a good scraping bot that lets you view the output directly, without running it through an LLM, please let me know.
Most CF sites I try to visit only give me an endless captcha loop, so I cannot visit them at all.
I suspect it's simply because my ISP regularly rotates IPs, and so I unfortunately have to share "reputation" with other users who can't behave online.
I just had a look through my logs, and I've had over 90,000 requests from known AI bots over the last month. All this to a personal website that doesn't post very often.
And that's just known AI, I can't imagine what requests are pretending to a real person when they aren't.
I think it's also because the regular dating sites are setup to not find you a match, so by bypassing them (at least in a way), you have a better chance of actually going on a date.
To keep in mind is that this article talks about people who have had success using LinkedIn, I'm sure there are people who've tried it without success.
The ENTIRE thing is AI generated. I'm not talking about the article. I'm talking about the entire website, the entire "product". https://0.mk/blog/zero-humans
Also it's super easy to tell by looking at it, way too many LLM-isms. No need for a AI checker tool.
> The name was registered in 2009 because it was the shortest URL possible: a zero, a dot, two letters. Seventeen years later the zero means something else.
Did the zero ever mean anything? It's still a 3 character domain regardless if the first character is a zero.
https://www.theverge.com/tech/993022/wordpress-automattic-ce... also has a writeup which takes some quotes from the article (and a statement by Automattic).
reply