bidirectional inverted commas, negative parallelism, short punchy prose, dot point listing ... these language techniques are consistently present throughout. whether this article was written by AI or not, its structure and style utterly screams of AI.
it is understood that LLM style is derived from human style, and that LLM style is now influencing human style. this doesnt make it any more obvious how to type an em dash or bidirectional inverted comma. nobody outside of academia or publishing uses these techniques of their own accord, only LLMs do. thusly i optimize out your chicken/egg conditional as it seems like a redundant branch - if something looks like AI, it's probably AI, so why does it matter what the AI was trained on?
I was thinking more like the expectation we used to have that the police didn’t just start shouting in a surprise ambush. For example, the U.S. navy knew that the ship was unarmed and returning from an event hosted by the Indian government which the U.S. had also participated in, that it was attempting to dock rather than attack, and that even fully loaded it posed an insignificant threat, so there’s an argument that the Navy could have one of their thousands of aircraft to give them the opportunity to surrender or evacuate before the boat was sunk. We did that for actual Nazis, who posed far more of a peer-level threat than Iran does.
The ship posed the only plausible threat, not the sailors.
> there’s an argument that the Navy could have one of their thousands of aircraft to give them the opportunity to surrender or evacuate before the boat was sunk. We did that for actual Nazis
I think there is absolutey an argument that good decorum would have provided a nearby surface vessel to assist with rescue. But not being nice in a war isn't the same as a war crime. And expanding the notion of war criminality to cover even breaches of decorum fundamentally waters down a term that has already started being seen as meaningless because people want to make it apply to any act of war.
No, but the U.S. navy has more than a single submarine. Given the vast power differential and the fact that they knew the ship was unarmed having participated in the same International Fleet Review exercise, there’s an argument that, say, an aircraft radio message giving them an opportunity to surrender first.
The “shot across the bow” phrase comes from a relevant naval tradition, such as when the U.S. Navy captured ships from far more serious threats like Nazi Germany:
Looking through that article, one of the examples is "The wife of a conservative politician was sentenced to 31 months in prison for what police said was an unacceptable post."
"The wife of a Conservative councillor has been jailed for 31 months after calling for hotels housing asylum seekers to be set on fire."
This is pretty clear incitement to violence.
The UK has problems, but it's not very useful to throw all of these cases together to make a big number, it really rather undermines the point.
(edit - looking at the video posted in a sibling comment is enlightening. The number actually convicted of anything is around ~400 and this includes a lot of direct incitement to violence, stalking and all sorts. Which are similarly illegal in the US. The US right-wing talking points are as usual a load of crap.
None of which is to say I think the UK has things right, and that number of arrests is clearly a problem in itself, but as usual the "OH MY GOD look at what's happening over there! Muh free speech!" from the US commenters is hypocritical and myopic)
In this podcast episode[0] he does talk about this kind of model and how it "learns about physics" through experience instead of just ingesting theorical material.
The way I see it, the "world models" he wants to train require a magnitude more compute than what LLM training requires since physical data is likely much more unstructured than internet data.
He raised $1b but that seems way too little to buy enough compute to train.
My bet is that OpenAI or Anthropic or both will eventually train the model that he always wanted because they will use revenue from LLMs to train a world model.
I meant during the two world wars, which should have been obvious. The idea you think I know what NYC is but forgot about 9/11 says more about you than I.
Iran had nothing to do with 9/11. If that was the point you were attempting it is incorrect. Not even the current administration is attempting that line of reasoning.
-G. K. Chesterton
reply