Dude, that's so wrong.

People in the US can promote and even advocate violence. They just can't do that when there's both the intent and likelihood of triggering imminent lawless action. See also Hess v. Indiana, 414 U.S. 105 (1973), where someone who advocated lawless action was allowed to do that because the action wasn't imminent.

On the one hand, I think it's awesome and hilarious that Russia's strategic bomber fleet got so smashed.

But that fleet hasn't been moving the front lines in Ukraine. Troops on the ground have been doing that, and apparently those front lines are still moving in Russia's favor. Also, strategic bombers being destroyed shifts the calculus of Russia's engagement with third countries. I'm worried that it increases the risk of broader war without significant progress in Ukraine. Maybe I'll be wrong, and Ukraine will use the same kind of tactic in other situations that do motivate Russia/Putin to come to a ceasefire and peace -- but so far it doesn't look likely.

Similarly, in October 1903, an editorial in the New York Times claimed it would take between one and ten million years for humans to develop heavier-than-air machines that could fly. The Wright Brothers proved that wrong just under ten weeks later.

https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fen.wikipedia.org%2Fwiki%2F...

The paper linked above says they gave the model a fixed number of the most recent tokens as context, with N usually 30,000. Assuming they also repeated the high level directions for the task, that doesn't seem unreasonable for current context windows, but it also suggests why models went off the rails sooner or later (and usually sooner): they have to re-run the model with a window reset occurring frequently, so there are lots of opportunities for it to go awry. Models will need a lot more sophistication before they can keep consistency over long tasks.

You continue to miss my point with the discussion of prose vs non-prose. Learn to read!

Generally it requires some level of effort to get non-public information that they know to be sensitive, and to use it to the detriment of the country they got the information from. This information was public, therefore not espionage.

Repressive regimes might have different definitions of espionage. Were you using one of those definitions?

See my response to the guy above. You're absolutely wrong on the law here.

Courts have held that non-prose behavior of software represents protected speech by the company creating or running the software. A fortiori, prose output of software represents protected speech as well.

The LLM isn't the entity that has speech rights here, it's whoever runs it -- just like search engines, online platforms with moderation, and so on, all the other examples that you pretend are not speech but that precedent says do represent speech. See https://globalfreedomofexpress... for discussion, including reference to other relevant cases:

That's just begging the argument. The output of LLMs is obviously speech -- and in the US, speech that isn't protected by the First Amendment is defined by quite narrow exceptions. Which one do you think it fits into? How do you distinguish it from, say, automated decisions about content moderation (which lots of people here and elsewhere argue is protected as free speech by the First Amendment), or search results, or a wide variety of other output from software that is a lot less like traditional English prose? If a program outputs very speech-like prose, why is it less speech than non-speech-like outputs from other software?

It's been in the news recently, but in the form of organized crime going after cryptocurrency high rollers: https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fwww.npr.org%2F2025%2F05%2F28...

A few governments will use wrench attacks readily. More will use it in what they think is an emergency. Using cryptography is basically never going to trigger it, though: the government will have already decided the target deserves that kind of treatment regardless of cryptography use.

On the other hand, using a cryptography back door leaves almost no evidence and may be entirely invisible to its target. That makes governments a lot more willing to use it.

Why is that blasphemous? Tolkien wrote about what was essentially the divine right of kings to muster their nations and lead militaries against an evil empire and a quisling. The whole bit at the end about Wormtongue taking over the Shire was a way to say that destroying the One Ring by itself wasn't enough to defeat evil.

No, you don't. Like most of what you post, you made this up.

He paid $0, just like almost all contract winners, and his name is Zuckerberg, not Zack.

It must be nice to have the CCP pay you to post such incoherent bullshit. You don't have a never ending well of money, and nobody's is taking $800B from the US healthcare system. That's yet another thing you made up.

This being Amazon, I believe the obligatory description is "they lose money on every sale, but make it up in volume". That's why people assured me Amazon was going to go out of business in the late 1990s.

A billino can't be very big if it only costs $3. I'm surprised that anyone bothered prosecuting a lawsuit all the way to its conclusion over such a small dollar amount.

They've been moving to a model where the number in the product name reflects when you pay for it, thus Microsoft 365, which you pay for every day.