I have just started to learn differential privacy (https://en.wikipedia.org/wiki/Differential_privacy), and am wondering if the following might be feasible in principle: on the browser level, instead of blocking the trackers, add a certain level of noise to the submitted data. This might form a truce between the end users and the trackers. Through statistics, the trackers might still be able to learn something about the end user group as a population; at the same time, each individual user's privacy isn't breached much more than they are completely offline. Admittedly this might be ridiculous and is just me under Dunning–Kruger effect as a beginner in this field.