Yeah, you're going to find that most people are very much biased against enabling a violent coup on the government that is being really good to them. A coup that, if successful, has the potential to endanger their entire business, their wealth, and possibly their lives.

BTW, "you're biased" is not, in and of itself, a reason to entirely shut down an argument as invalid.

The bias isn't in the response, it is in the faulty interpretation:

  A coup that, if successful, has the potential to endanger their entire business, their wealth, and possibly their lives.

Not only was this an unsupported at the time, but unless something really flashy happens it is only going to become more obviously wrong. 1 violent riot in 2021 compares not so badly to the widespread peaceful arson and looting in 2020 that was brushed off as routine political expression.

> BTW, "you're biased" is not, in and of itself, a reason to entirely shut down an argument as invalid.

No argument on that particular point, but it is a decent reason not to try and use a certain infrastructure provider for communication tools or artistic content. Most people are looking for a dumb pipe when they pick an infrastructure provider, not a back-seat moderator with a twitchy trigger finger.

> What you are presenting is the worldview of someone who is biased.

"This site we host was just used to plan a violent revolution. It's more trouble than it's worth, let's dump them."

Is this what you mean by "bias"?

No, the bias is assuming that is true and acting on it without sufficient evidence.

What do you consider sufficient evidence here? There were multiple documented posts on Parler that have been documented running up to the events on Jan 6'th that show people organizing and preparing for it. The evidence was present before the decision by AWS. I know folks who were collecting that events weeks before the event and predicting violence based on what they were seeing. The evidence for it is only increasing now but there more than sufficient evidence on the day of the event to indicate that Parler was going to be a liability to anyone hosting them.

Bias in this case works both ways. It makes some ignore the evidence that was already present. Businesses aren't courts that need to prove guilt. This was about getting rid of a high risk client who had made clear that they will not be changing their high risk behavior. There is no reason AWS needs to share in that risk after giving them the opportunity to change.

> What do you consider sufficient evidence here?

If I became convinced the expulsion had nothing to do with Jan 6 and was part of a standard process of review that AWS conducted for all their customers then that'd be sufficient.

There were, generously, a few hundred people involved in the Capitol storming. It isn't obvious how organised they were organised since they lacked guns and riots sometimes start spontaneously from protests. AWS didn't have time to work out who they were, what they were trying to achieve, how they were organising, what there motivations were or what was going to happen next. Using that as a trigger to ban Parler - when they could easily have been organising on Facebook or in person - would be strong evidence of political bias.

> Businesses aren't courts that need to prove guilt.

I don't mind if AWS is biased, but part of the fun of kibitzing is assessing whether decisions are good or bad. AWS making decisions on bias rather than rational thought process would be bad. People who use evidence and can think in a consistent manner have a big advantage.

Twitter made the mistake of trying to rationalise their bias, giving us a memorial of them trying to rationalise "To all of those who have asked, I will not be going to the Inauguration on January 20th." as a call to violence. It is a legitimate concern that the same sort of evidence-free thinking could take root in AWS's leadership - they might easily be drawing their leadership from similar pools of people to Twitter since the big tech companies are acting in the same direction at the same time.

> If I became convinced the expulsion had nothing to do with Jan 6 and was part of a standard process of review that AWS conducted for all their customers then that'd be sufficient.

Whoosh. The entire point of the evidence that exists is that Jan 6'th dramatically changed the risk calculation. Without Jan 6'th I can imagine that Parler might still be online. But if you ignore Jan 6'th and it's impacts you ignore a large part of the calculation going on for a business. Jan 6'th took Parler from hypothetically risky to proven risky.

You keep attempting to dodge or downplay those events but they happened and they had a very dramatic impact on the political landscape.