On: Princeton’s Ad-Blocker May Put an End to the Ad-Bl...

[takeda]

Allow me to be a bit skeptic about it.

This reminds me a lot of Bayesian spam filters. Those things worked great very few false positives and negatives, once Google started using it on a large scale (GMail) spammers started playing the system and often they are successful.

[takeda → takeda]

Also, as an example, I noticed a spam today with a subject: "Magic pill lets Trump perform better" which is an obvious way to fool bayesian filters.