Hacker Newsnew | past | comments | ask | show | jobs | submit | puttycat's commentslogin

Yes. And in combination it completely rules out the suggestion that the Netherlands attack is some kind of local activist protest. (The France attack resulted in several grave injuries and could have just as easily resulted in dozens of deaths.)

that's a huge leap

We should see some claim of responsibility if it was indeed anyone other than the usual suspects.

Why? Russia gains nothing from taking responsibility.

No I agree with that, but if it's "activists" I would expect them to claim credit otherwise it's all pointless. The same doesn't follow for Russian fuckery, IMO.

Quite appalling to open a "journalistic" report by ridiculising the people you report on:

> Not since Socrates complained that the written word might make people lazy about remembering things has there been such an outpouring of angst against a new technology in the world of academia.


It's also sort of a misleading gloss of what is going on in that dialogue, which, it is important to note _is in a book_ -- one of many that Plato wrote. Plato has Socrates recount a story (supposedly) from ancient Egypt in which an Egyptian god claims that books will make people forgetful. He never makes that claim himself, as a character, and certainly Plato didn't believe it, because he wrote lots of books before and after that dialogue. The general thrust of that section of the dialogue is that writing can be a useful tool for remembering, but _not_ sufficient for learning, which requires person to person interaction and dialogue.

Intelligence is actually correlated with the ability to simplify complicated things. Occam's razor. Compression as comprehension.

We're not asking the model to simplify something, we're asking it to perform a task. Its subtle preferences show up as an overcomplicated path to the goal.

In some cases, there are also nuances that we don't pick up on. Here it's our preference for simplification that's showing up. We set the lossy compression factor higher than it does.


This is really impressive. Can you say a bit about the underlying process? I'm guessing this is post-training qantization? Isn't PTQ also resource-intensive? (Ie might not work on any machine)


Not sure that's correct. You're replacing a simple visual recognition problem with something similar to the halting problem (but maybe less fatalistic, depending on the formalism you're working against ie DOM vs js).


Opus 5's language is just terrible


What's the business model of these startups?


> he brought these attacks with his war

I'm no Trump supporter and this war was a big mistake, but justifying a nation poisoning another's civilian water supply is a bit upside down.


Yeah, what could we possibly have done[0] to justify an attack on our water supply?

[0]: https://www.commondreams.org/news/iran-water-desalination-pl...


We don't follow "rules" in war anymore, according to "Secretary of War" Pete Hegseth. We show no quarter and take no prisoners, we pursue "maximum lethality, not tepid legality."

That's American policy now. The gloves are off, no holds barred, everything and everyone is on the table. So I guess we reap what we sow.


Ask the people of Flint, Michigan?


> Video models can certainly predict what stirring cream into coffee does.

Not sure that physicists and professional animators would agree.



Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: