That's not an argument against moderation. That's an argument against monopolies and oligopolies. Banning a guy from the store isn't even in the top 10 worst things that will be going wrong in that scenario.

Okay, that's another way to fix the problem then: break Twitch up into 100 baby Twitches, and don't let anyone have ownership/stake/control of more than one of them. If all 100 decide to ban someone, they probably deserve to be banned, and if they don't, then the person can keep streaming on one that didn't.

No because the same social and economic pressures may apply to all of them. That's exactly what used to happen to black tenants when all of the "100 baby landlords" decided to ban them. It doesn't mean they probably deserved to be banned, it's that there was a systemic problem that couldn't be fixed simply by having competition. We already see this with politically unpopular people being banned from all popular social media sites, despite them being separate companies with different owners.

I think it's hilarious that you're using the suffering of black people to give weight to your fears about what's happening the sort of people who are in fact fond of or indifferent to the suffering of black people.

Back here in reality, a poster child for "politically unpopular people being banned" was Milo Yiannopoulous, who was banned for leading a wave of anti-black abuse: https://www.vox.com/2016/7/20/12226070/milo-yiannopoulus-twi...

One could argue that Milo is the one who started the shift we've seen over the last 5 years in social media moderation decisions. He still has a platform, of course. There are approximately infinite platforms, and some of them are happy enough to host open racists. So we aren't at the "100 baby landlords decided to ban" stage yet. As you say, he's just been booted off the popular platforms. Because it turns out a lot of people are no longer interested in supporting open racists.

And it seems like that's where things will stay: Users and companies have unimpinged freedom of association. Racists, etc, have the freedom to build sites that support racists, which are mostly ignored by everybody else. This seems like a fine outcome to me. If it doesn't to you, maybe think about why.

Of course that supermarket should be entitled to ban the misbehaving person from their store.

To turn the story around, why should a business that takes it on itself to provide for a remote community be punished for it by forcing them to accept unruly customers?

The point is that they're banning people for things not done in the store.

What if there's a single major provider of groceries in a state and they decide to ban everyone convicted of a felony for life. Would that be ok?

In the case of the supermarket, the person who is banned could pay somebody else to buy stuff for them. In an extreme case, that somebody would have to be the government preventing the person from starving.

In general I think private businesses should be allowed to decide who they want to do business with.

In the case of Twitch, the question to me is what is the contract the users enter with Twitch - what promise does Twitch make? If it doesn't promise anything, users invest at their own risk.

I had no problem with hosts kicking off Parler, and I don't really have a problem with this Twitch policy, but those are very different things than the grocery example. Social media companies, webhosts, etc. are directly enabling speech as a core part of their product. As was mentioned above it would be against the principle of not coercing speech to force one of these companies to continue serving a customer.

But I really do not think allowing someone to buy lettuce should be considered a form of speech, or even an "association". Banning someone from using a grocery store - or perhaps more relevant to this example Amazon.com - for actions completely unrelated to the service is wrong IMO, even if it were somehow restricted to 100% accuracy rate of only blocking very bad people.

I haven't seen it happen yet, but if it becomes in vogue to worry that selling a random material good to a customer is some kind of endorsement of that customer, that will lead to a very concerning situation down the line. It is now all too easy to gather information on people, potentially dating back decades. I don't see any reason grocery stores should know anything about their customers outside of what can be observed when shopping.