saagarjha
saagarjha on Hacker News
I wonder if one could use this to efficiently abliterate models and identify guardrails ;)
I am very amused that the agent speaks exactly like an agent in this world too
No most of the projects I've contributed to don't require it anymore
I mean, it seems like you are probably not submitting patches to anything at all with that attitude
Unfortunately she had the gall to retire beforehand
> You're misunderstanding, probably intentionally. GP is not upset on behalf of the cold lunch or the person receiving it. They're saying this is a stupid way to protest.
I understood this
> It had no benefit to anyone, brought no attention to what she was upset about
“Stopping in front of the robots, Marvin said, led to productive conversations with other passersby who asked what she was doing.”
> which is moronic anwyays - being mad that robots are taking the "jobs" of DD/UE "workers" is completely misunderstanding the issue
I intentionally did not express an opinion on the merits of her complaints
You have greatly misinterpreted history if you believe this to be the case.
When people protest by bringing out the guillotine it is of course too violent and unbridled. When they block bridges then there are ambulances people might die in. When they pour paint on art it destroys priceless works. Now, when someone delays the Uber robot, it causes someone to get their lunch cold. One wonders if effective protest can only be done by sitting quietly at home, so as not to harm anyone or anything.
Then provide a link to your submission and we can see this for ourselves.
In this case you're not doing much better.
You may have trouble catching good attempts then.
I would not be surprised if what happened here is that Apple had hardware to release this week and needed to get 27.2 out the door with support for them, and it was "too late" to merge in the certificate store changes. Of course, I think most people here would argue that this is completely unacceptable and I would more or less agree with them, but unfortunately Apple has never been capable of validating a release quickly or pushing small patches. They've tried and, well, most of those efforts have not worked at all. I wonder how they will deal with a new full chain targeting their OSes every week or whatever AI is going to do.
XProtect removes malware, not certificates
Yes because that thing is completely useless and provides basically no value
This isn't technical documentation though, it's just notes the agents are keeping for each other. I think that is a very different task
I assume having a person write words for other people to read is not feasible?
You should make another index is my point. The database layer properly implements this extra index instead of just being a stupid dictionary that you have to massage with the one key value you have
Running an internet forum does not make you better or less impartial at commenting
Let's also try to figure out if smoking causes cancer, yes.
I don't actually care that much for Grok the model because I think they are more or less homogenous with like, a small house spin on how "smart" they are and some other features like their behavior in a harness etc. Personally I have no problem believing that it might be good at doing what you ask directly instead of wandering or trying to be independent or whatever. I wouldn't call it out as paid marketing and I also just don't really think about this topic very much.
However, the comment I did reply to seemed to think that just because you ran a forum that you would be unlikely to be biased. This is like obviously false, if anything you have major internet brain where you spend enough time on the internet where you come up with a whole complex about how you are contributing the uncommon opinion the site needs to hear or whatever, when it's just the same dreck from the rest of the internet sprinkled in here when it hasn't been posted enough by someone else. That is not directly "I used Grok and thought it was good at x" but, like, it's clear that it's something that is at the periphery of a whole network of positions you hold that are aligned this way.
I don't necessarily think this is bad and in fact I think most people on the internet at least basically post in this way but I do think it's important to be aware of it at least. I'm not sure you are. I can tell you more or less where my slop opinions come from and I don't think it would be accurate in any capacity to describe them as bucking "the prevailing HN orthodoxy". I, uh, think you should probably describe your comments differently.
If you were a little slower to jump to conclusions you might see that the blog post does not actually make the elementary blunder of not understanding how Rice's Theorem works, and that the claims made are qualified correctly as they come from people who are quite good at understanding what formal verification is and can do ;)
What is the intended goal of this website? The doctor told you smoking is bad so you're going to give your child a pack a day and see if they die by age 20? If they don't, obviously, your doctor is a quack who cries wolf, and you shouldn't listen to that killjoy anyway :/
C'mon guys, I get that OpenAI is involved and they want to make ten trillion dollars next year or whatever, but I thought we had put the idea of random nonexperts trying to judge how "hard" something is to exploit to rest years ago. The guys who used to try doing this lost repeatedly, publicly, until we finally realized that we have to secure our stuff. Sure it's annoying to update your computer every week and sometimes some guy at Forbes is going to tell you that if you don't install iOS 67 today you will get hacked by the Mossad which may not have been entirely true but when you have memory corruption in a reasonably common configuration that faces the public internet, you should patch your software. End of story.
We can argue all day about the cost to exploit something, whether it is "hard for humans" or what it would cost on the black market or in this case how many tokens are required, but that is mostly just a distraction. Just patch your stuff. The people involved here are (currently) paid by AI companies just like ten years ago they were paid by the ads companies but despite the annoying way our economy works they mostly know what they are doing. I get that this is annoying, I understand that nobody wants to hear the words memory safety again, but this is anti-intellectualism. "Ask HN: Can you explain how the PHP bug could be exploited?" would be far more productive than having a bunch of weary sysadmins duke it out in the comment section against security guys.
No, the claims they made are well supported.
I don't really know what to tell you but the comments are definitely Claude. The result is pretty good but AI was definitely involved in the creation of this.
Sure but at least you are being honest about it. For what it's worth the database can handle this internally without you needing to hand this back but if you have a bad one then yes you should probably have it give you back a location hint that you use for future queries
Yes, which means that you should probably be able to provide your timestamp to the database so it can use it for optimizing its index, rather than trying to shove it in your ID