Move fast and break things indeed.
Facebook, Twitter, Reddit, etc. all need to curate so that their platforms stay family-safe, in order to please advertisers and the general public (imagine how bad it would be for e.g. Facebook if the general media started portraying them as supporting e.g. "terrorism").
Actually, mine does, if you have a certain mobile plan, people who call you get an ad for the plan when its ringing.
The general public has to accept there are some things they don't like and still have the right to be there, and stop acting like a spoiled brat. IMHO. Of course that wont happen.
I understand your point, but I would add some qualifications:
1) Let's remember that algorithms are human actions. For example, if Facebook created an algorithm that deleted accounts based on user skin color in their photos, it would be a human action.
2) Facebook is a highly sophisticated software company, not someone who downloaded a library without realizing quite what it did. They are especially responsible for what their algorithms do. (To think of it another way, imagine how they would look if they claimed otherwise, 'we don't really understand what our software does'.) EDIT: It's also a consequence of priorities: If the algorithm's quality had a billion dollar impact on their bottom line, there would be no errors.
3) Facebook isn't just a private business. With so much power over the world's political and social communication, they have an obligation to ensure their members freedom of speech.
4) Transparency and remediation is necessary for #3 (and maybe Facebook already has it; I don't know). Such things always can be blamed on an algorithm or computer error.
malicious: intending or intended to do harm.
bigoted: having prejudiced intolerance of the opinions of others.
I was only merely pointing out that reporting an atheist page is simply the result of intolerance and not of an intention to harm.