3 ms·
On the other hand, if the heuristic is "pay a bunch of Mechanical Turkers five cents per website to classify typical pages from websites as 'articles' or 'not a
by mechanical_fish 16y ago
On the other hand, if the heuristic is "pay a bunch of Mechanical Turkers five cents per website to classify typical pages from websites as 'articles' or 'not articles'" it's going to be pretty difficult to defeat.
Why bother with artificial intelligence when the real thing is so inexpensive?
Indeed, you might not even have to pay. Just build a system to report bugs. People may happily self-report when their adblocker fails to function on pages where they want it to function.
I suppose there might be tricks, like chopping your page up into misarranged divs and then unshuffling them with CSS and/or JS, that make it really hard to scrape the content for Articles, even if you know that the page should be an article. That class of hacks will cost you in accessibility, though, and they can't be great for SEO.
- chaosmachine 16y agoYou can avoid any major SEO problems by only serving the modified version to iPads (As long as it's providing proper user agent information, that's easy, but there are other ways to detect it as well). In fact, if you really want to go all out, you could just serve the entire website as a rendered image to iPad users.
- joe_the_user 16y agoYou could even cut your site into tables or divs but with the table-boundaries not corresponding to the text/ad boundaries. The possibilities are endless. I think the reason folks don't it now is that add blockers are not ubiquitous. The only way web filtering works is if the display mechanism and the meaning of page displayed roughly correspond. If someone decides to start massively filtering the web in a way that doesn't serve the original site owner, the site owner can certainly start serving content that displays correctly but is hard to filter.