Just having one service ban you will screw up your life, such is the case of one programmer who uploaded a picture of his toddler to let the doctor check for infection - and got permanently banned by the AI, losing access to his Gmail and other Google tools registered to him.

And services rarely act alone, when they do decide to target someone, they usually act in unison to deplatform, whether justified or not.

In some ways big tech is far more powerful than governments.

This is a case for regulation* and for promoting competition and diversity in the marketplace. It's not a case for treating them as governments. If they're governments, who chooses their leadership? Do they have one board, selected by the US president/according to the rules of the US Congress, that is applicable in the US, and another board, selected by the Canadian PM/according to the rules of Canadian law, that is applicable in Canada? How much are we going to pay to their shareholders for confiscating their property?

* Banning moderation is a form of regulation. It falls far short of treating them as governments. It is possible to say "I want to regulate Twitter" without having to make weird statements like "We should treat Twitter as a government".

Nobody said that they are national governments. They are organizations with governance policies.

Even if corporations are not governments, they are far more similar to governments then they are to private individuals.

Moderation is a form of regulation- it is a way of regulating speech. Banning moderation is deregulation.

"More platforms" is not the answer. In a democracy, when people are publicly discussing matters of public or political importance, other citizens have the right to participate in that conversation.

> Nobody said that they are national governments. They are organizations with governance policies.

Under this definition, practically any organization "is a government". My college debate team, Tesla, your local Starbucks. All "governments", because they are organizations have a governance policy (which really just means "leadership").

> Banning moderation is deregulation.

No. The government controlling how corporations can act is, definitionally, regulation. You can argue that is good regulation, but it is regulation. If I am allowed to censor you, that is less regulation than if I am legally prevented from censoring you.

Can you prove collusion between services?

That person was banned because of automatic child pornography detection, you think that should be removed?

Maybe the detection could stay, but the decision making (= acting on gathered intelligence) should absolutely be not outsourced to dumb robots. In this particular case, the context of the e-mails between the father and the doctor would immediately indicate that the photo was legit.

How could a large service have enough people to review material in a timely manner?

CP is, in reality, a fairly infrequent crime. It is also a crime, meaning that someone human should definitely pay attention to it - unless you propose robotization of sentencing as well.

If your automatic detection is tuned well enough not to produce shittons of false positives, you should have enough people to review suspected cases. Real positives ought to be escalated to actual police.

(Also, actual police will not be happy about an automatic mechanism that locks out real CP producers out of their mailboxes, thus alerting them to the fact that there is something going on, and possibly giving them time and opportunity to destroy some incriminating evidence. There is a reason why suspected criminals are staked out carefully, then suddenly raided.)

do we really expect underpaid and overworked jannies who get to spend 12 hours a day checking gore and cp will have a better error rate than bots? customer service & support is dogshit even for the users that these services are ostensibly trying to keep. potential liabilities will never get anything short of kafka.

We are talking about a real black letter crime here, not just a vague Terms&Conditions violation. If actually detected, humans will (or should) be involved way beyond mere suspension of an e-mail account. As in "actually running a criminal investigation".

Which will either clear or implicate the person involved, but either way cannot be done robotically.

And you know what happened after?

Nothing. The programmer tried to open a lawsuit to gain back his rights but he couldn't afford a decade lawsuit against a mega corp.

So the programmer remains banned today, for Google's mistakes.

Google won't admit wrongdoing, because it reveals how flawed their system is and opens them to cases.

Should such an entity with no responsibility be allowed so much power over our lives?

Lots of government services post on Facebook - often only on Facebook. You can't comment on that or read it elsewhere. There are no alternatives unless the poster picks it for you.

Alternative to what? Protest? Why does it have to be the Facebook post?

That doesn't make Facebook like a government though. If twenty years ago the government chose to advertise a job in the three largest newspapers in a city, and not in the Dah00n Times, that doesn't make the newspapers like a government or incorporate them into the government. It might mean that people should question whether the government is too closely tied to certain private corporations. But that's a different question.