[hackerlight]

> also have never faced the challenges of moderation at million people scale. So, it kind of cuts both ways

Not really, given that I can see the billions in cash on hand of these social media companies, which they could use to hire real people instead of intentionally neglecting customer service because it isn't a driver of profits, and instead choosing to dump these costs on the taxpayer of the countries that have to clean up the mess via the legal system. Negligence or even incompetence is not a valid excuse for actively facilitating a spectrum of behaviors ranging from harassment (this case) to fomenting populism (US elections) to outright murder (Myanmar).

[deltree7]

Again, you have no clue how to handle this.

It's not about the money. Assume you can hire 100,000 people. How would you maintain consistency among all those 100,000 people. You'll get huge variance in the kind of decision making from each of those people.

Then you are going to say, codify it and don't allow variations, which means a program / AI can do a better job, which is what most of these firms are optimizing far.

Unless you have personally solved that issue or have an example of someone solving it, it is literally arm-chair critiquing

[hackerlight]

It's cute that you assume good faith. Here is evidence they are operating in bad faith. They allow users who are being harassed to disable replies but not quote tweets. These quote tweets are then used as a vector of further harassment and dogpiling. This happened with Steven Pinker for a time period and Amber Heard. They are choosing virality over preventing harassment.

  "Assume you can hire 100,000 people. How would you maintain consistency among all those 100,000 people. You'll get huge variance in the kind of decision making from each of those people."

This is the perfect solution fallacy. Perfect solutions are not the bar. Timely response to reports and human review is the bar that I expect of them. And they're failing to meet that bar.