Relentless AI scrapers pushed DNA Lounge's web hosting bill from $800 to $3,500/month.
That is a $32,000/year cost shift caused entirely by companies training models on their content without permission. AI companies are externalizing their massive training costs onto independent web operators, effectively taxing the independent web to build their proprietary models. 📉
I love DNA lounge. In my 20s you'd find me there dancing, high as a kite on E, until the wee hours of the morning when the sun rises. Then my best friend and I would take Muni home and every ambient sound morphed into music. We'd be in our rave clothes surrounded by commuters buttoned up tight. Grateful to not be one of them. Young people should all experience freedom and joy. I hope DNA lounge is able to protect themselves from these AI scrapey bots and keep going.
@Pikapods Those are the same #Clankers that performed a multi-month #DDoS attack on my site!
Same excuse of 'AI scraping', but they used so many IP addresses, that fail2ban couldn't catch up.
Even when I reluctantly deployed Anubis on it, Anubis was struggling to handle all those requests. Yes, it was that bad! It eventually stopped after that.
These #AI companies are cybercriminal enterprises. DDoS attacks were illegal last time I checked, and so was the hacking that they did to Huggingface.
It's abuse. It should be prosecuted, as the DDoS attack that it is.
@JeffGrigg @alice_pea_3526 @Pikapods Sadly, in the post-democracy late capitalism we now inhabit, nothing done by huge corporations that donate millions to politicians is prosecutable. The fact that they're literally killing people as well as engaging in massive theft is unimportant.
@lmgenealogy More, the Tech Bros are cornering root infrastructure, like energy markets, to complete the extinction level metrics to data centers, and AI.
Oct 9, 2026, 5:28 PMEsp for entertainment/lifestyle/restaurants etc. having a website is quickly becoming a liability, not a value-add. If I owned a company I'd probably shift to a social media model where the external costs of the scraping are taken on by the tech conglomerate that owns the network.
Which, come to think of it, might be the point lol
Oct 10, 2026, 8:42 PM@Pikapods A site I am involved with had to finally close its forums from being publicly visible after more than 15 years because we were having bots hammer us at 1Gbps of requests 24/7. It's insane.
@Pikapods After being a web designer for 31 years, I decided to retire this year because I was spending too much time defending against AI bots, spammers, and hackers. 90% of the traffic to my server was unwanted.
@Pikapods at this point I've resorted to a "shoot first, ask questions later" approach; blanket banning ranges, ASNs, etc. obviously not ideal but idk what else to do when these shitty bots fake their user agent and overload our web server. Cloudflare's ai bot blocking feature helps some but doesn't get all of them.
In conclusion whoever is making these scrapers should be banned from going within 15 feet of a computer ever again
@Pikapods Huh. "We spent 10m to solve Navier-Stokes" doesn't take into account $32,000 multiplied by 1m small businesses.
@Pikapods I knew this was going on back when #AdobePhotoshop switched to online only licenses. More features = more things to input from users!
@Pikapods Website thinks Firefox ESR153/#SailfishOS is a bot.
Oct 12, 2026, 1:22 AM@Pikapods
It's not even the cost of training, they're just programmed poorly
People have been scraping the web for decades by this point without causing these sorts of problems