This blogger lets the reader know they don’t know wtf they are talking about in the first fucking paragraph.
Making “anti-ai” fonts that obfuscate or scramble text is a fools errand. For fonts that scramble text, it’s an accessibility problem, since the scrambled text is what screen readers and other accessibility tools will parse.
The text scambling techniques to stop AI do not affect the normal readable font on the page and doesn’t stop text to voice readers. 🤦♂️
Visually… it just functions like a copy blocker, which news sites have tried forever to fight plagarism.
The original text is still in the website’s source, in plain text.
…I don’t even understand what its supposed to accomplish. Just make copy-paste more difficult for humans, or agents looking at the page visually? The raw page source is the thing that’s scraped, AFAIK.
I just did. To allow screen readers which is what AI would do if this became a common block it uses this:
" Pressing it sets the reader’s browser to solving a compute-heavy puzzle: JavaScript and a few seconds of processing, more than most mass scrapers are willing to spend."
Which is exactly the method already addressed in the article:
" requiring computationally expensive systems to access the content"
It’s 100% scrappable. Nothing there stops me for scrapping the true text. It’s a very minor inconvenience at most. I would even argue that’s more of a pain for users and developers than for scrappers.
This blogger lets the reader know they don’t know wtf they are talking about in the first fucking paragraph.
The text scambling techniques to stop AI do not affect the normal readable font on the page and doesn’t stop text to voice readers. 🤦♂️
[citation needed]
Gonna need more than a confident declaration to convince people potentially impacted by this. Provide it… or shut up.
Try looking up how ShieldFont actually works. Or look at the code yourself; it’s open source.
I just tried the demo.
Visually… it just functions like a copy blocker, which news sites have tried forever to fight plagarism.
The original text is still in the website’s source, in plain text.
…I don’t even understand what its supposed to accomplish. Just make copy-paste more difficult for humans, or agents looking at the page visually? The raw page source is the thing that’s scraped, AFAIK.
I just did. To allow screen readers which is what AI would do if this became a common block it uses this:
" Pressing it sets the reader’s browser to solving a compute-heavy puzzle: JavaScript and a few seconds of processing, more than most mass scrapers are willing to spend."
Which is exactly the method already addressed in the article: " requiring computationally expensive systems to access the content"
I’m looking their demo website https://shieldfont.org/demo/
It’s 100% scrappable. Nothing there stops me for scrapping the true text. It’s a very minor inconvenience at most. I would even argue that’s more of a pain for users and developers than for scrappers.
What stops you is the javascript puzzle, I agree it’s not worth it for the inconvenience but the idea is at least interesting
OK cool, thanks. What I saw in demos a week or so ago seemed “fuck those screen reader folks”. Would be good if it’s not as bad as that.