One of the main type of algorithm used, especially by advertisers, is
classification algorithms. They are used to classify target classes.
Taking your inputs and interactions, from social media, links and likes,
algorithms classify you into groups, which companies can then choose
to target, allowing maximum profit on their half. No point wasting
money on adverts from people within the wrong target group.

Although sometimes correct, try
experiment by putting different
objects in front of your webcam,
this classifier is often wrong.
Algorithms aren’t always correct,
and we can’t always trust
something just because we don’t
understand how it works. When
should you trust a machine over
your own judgement?