Also Cloudflare: let's send normal humans who are trying to go about their daily lives into endless Turnstile spinner loops with absolutely zero recourse, grievance, or support infrastructure.
Turnstiles per minute.
I went for a personal newsfeed, agent pulls news form ~100 feeds related to my interests. Then reads all articles for me and orders them by how interesting they might be for me. I specifically asked for vector embeddings, up/down votes (-2..+2), visited status, LLM content evaluation. Probably there are some other mechanisms I didn't even bother to check. It's a work in progress but I can see myself replacing most of my news reading with it. For many news the AI summary, which contains main idea behind the item is enough for me. As a bonus it resolves clickbait and is quite good at it. Also no ads, ever. For sure I need to implement some grouping because when the popular story breaks I have many stories about the same thing with mostly overlapping details. AI merging them would be quite cool.
I also asked AI to extract my interests from my browsing/watching histories of my all accounts. V2 of my newsfeed might utilize that somehow for better results.
Silly thing is I made it in one afternoon with my only motivation of being slightly more annoyed with the web on that day.
When I touched it the last time I was a fan of gemma4 https://ollama.com/library/gemma4
Since then I grew really fond of Qwen3.6 (it's super capable in coding against the DOM) so I'll probably try to move to it with the next iteration. https://ollama.com/library/qwen3.6
When I move it to the server I might consider waking it up periodically to pull and analyze new stories and perhaps notify me if something absolutely great shows up.
There are so many possibilities for tuning it. And I don't need to think how to make it secure (beyond the basics), ultra performant, fitting other people's tastes and so on because this program has audience of one.
How do the vector embeddings fit into the picture?
Turnstile isn't something Cloudflare put up to annoy you. It's what the website owners decided to put up, for many different reasons.
In the same vein, Anubis has a default configuration that lets honest scrapers and crawlers through, because those can easily be rejected by basic web server configurations. Only scrapers pretending to be browsers need to solve the proof-of-work puzzle. You can disable that feature, of course.
Cloudflare may play this smart: force bots to pay for access, then take 30% of the cut and give the rest to the website owners. That way, websites get paid when the AI slop machine digests their content. Normal visitors get in for free, turn the scraper hellscape into a sustainable model. Bonus points for letting websites set their own rates (pre-declared to scrapers, of course) to dissuade all but the most interested scrapers.
Except for the unfortunate minority of normal visitors who always get misclassified as bots and get denied access regularly.
I wouldn’t be complaining if Cloudflare’s misclassifier bit any user with the same small probability. But it keeps biting the same users over and over again.
Website owners that use Turnstile and other such services choose to exclude these users. A tiny margin of false positives isn't going to dissuade most website owners, I imagine only the ones that themselves have issues with Cloudflare will bother to add the necessary rules to permit uncommon users (and the bots that copy them).