Can I kindly ask you to explain what is the problem with this kind of message?

I see this all the time - journalists critique pseudononymous internet writers. All through these pieces there's references to "who they really are" and "could they get away with saying this in public with their real name".

Why is this a bad thing? I know the NYTimes isn't the most moral upstanding institution, but they are hardly stochastic terrorists?

The implication of ST is you aren't targeting a specific individual, so this analogy falls short there - "Chris2048 on HackerNews" is definitely targeted. People rail against HN and its "tech bros" all the time.

The definition is widely accepted now to include targeting either a group, or an individual[1]. The "stochastic" component is that no individual or group is being specifically directed by command to carry out the attack, but the speaker is knowingly or negligently adding rhetoric to incite someone to do so.

But you're also conflating terms: criticism is not an incitement to violence. An incitement to violence is specific: "I think it would be great if some implied violence happened to that group I don't like".

It's specifically not considered ST to say something like "I hope some legal consequences happen to this person who I think is committing crimes" because amongst other things, that's a call to the application of the justice system, not terrorism.

It's also worth noting that other incitements, like encouraging someone to commit suicide has been found to be a punishable crime - and this is the sort of action Kiwi Farms specializes in and openly supports against their targets.

[1] https://en.wikipedia.org/wiki/Lone_wolf_attack#Stochastic_te...

Perhaps, but then "individual or group" seems to be pretty relevant to me, especially if you provide information specific to that individual/group versus a "tribe" of distributed individuals.

Plus, the "terrorism" aspect seems to apply to a wider group, or who else is terrorised? Presumably, individuals or groups with similar principles.

Your example also doesn't incite violence, just heavily implies it.

This is called fedposting and it's basically the message board version of what the FBI does to entrap Muslim terrorists and white supremacists IRL. The problem you're going to have with making it a crime is, where do you draw the line?

If someone gets in front of a big crowd and says "Donald Trump is a fascist, he's the next Hitler", and then one of those people takes the matter into his own hands - do you think the speaker should be found guilty of a crime?

The speaker is pretty obviously guilty of incitement to violence if they were inciting violence. Which my parody above is demonstrating pretty clearly. Now, is one incident sufficient cause? Probably not.

What if I launched a sustained campaign across multiple media against that HN user? What if that user starts receiving harassment and death threats? I am running a sustained campaign of making specific, targeted references to how desirable "doing something" about that user is with specific violent imagery, but stopping short of directly declaring that I want something to happen.

You're far too comfortable pretending this can't be a crime, or shouldn't be a crime, because in reality you're pretty comfortable it won't happen to you.

If someone actually does something, then yes, they should be guilty of a crime. I don't think it's possible to come up with a definition of "stochastic terrorism" in the way that you want, which doesn't result in a massive curtailment of civil liberties. How many acts of speech are necessary to make it into a crime? How many listeners / readers?