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

> this feature is very technically complex, and totally useless

Now, to break your confidentiality, Signal would have to have a relatively complex system setup for trying to match up messages and deanonymize people. You could imagine many scenarios where a bad actor (agency) attempts to trick Signal into logging metadata. This now requires a lot more information, and if nothing else would give you a level of deniability.


Not public, but some of the factors are listed here https://news.ycombinator.com/newsfaq.html (q2)


> In 2023, Microsoft's Bing chatbot famously adopted an alter-ego called "Sydney,” which declared love for users and made threats of blackmail. More recently, xAI’s Grok chatbot would for a brief period sometimes identify as “MechaHitler” and make antisemitic comments. Other personality changes are subtler but still unsettling, like when models start sucking up to users or making up facts.

Funny that they managed to call out all of their competitors without mentioning any of Claude's bad behavior


The only bad behavior I can think of from Claude is how it used to be so ethical it'd just refuse to do anything.

The quality of its thought outside coding is pretty bad lately and especially worse than o3/Gemini though. It really feels like they've forced it to short answers for cost control.


What bad behaviour of Claude was as famous as Sydney, or MechaHitler, or GPT' sycophancy? I've not heard anything.


Here's the paper that they wrote: https://arxiv.org/pdf/2506.20639

It's notable that this was an intern project.


Either this or https://huggingface.co/apple/DiffuCoder-7B-cpGRPO should be replaced with the current article.


I mean tbh industry research labs pump out a lot of good research due to them being intern projects (as in you have an army of passionate interns)


I wonder if this could be abused by purchasing 2+ refundable tickets, and then canceling all but 1.


Fully refundable tickets are an entirely different fare (and much more expensive)



I imagine that they'd get blocked by CORS


Good point. CORS somehow didn't cross my mind.


Do you find this to be worse than googlebot somehow?


Someone already thought of this and uploaded all of the contents of the box to a website! You can find it at wikipedia.com


That's missing the point. Full Internet access is just too broad. Going to wikipedia and aimlessly browsing about is fun, but a more educative approach can narrow the focus for students and especially for younger learners.

How to market it in developed countries is going to be a tough nut to crack though.


Well there is nothing stopping any school in the developed world from loading this on to a pi or something and having everyone use it too. It's free and open source (from what I can tell).

It's aimed at places with little to no, or unreliable, internet. So if you have normal internet speed there is nothing you can't get that's on the box. Also it seems that its not even a curated Wikipedia, it's just a full clone of it (assuming for whatever language your downloading)


Plenty of schools have network control over the devices that are used in schools, meaning that you can indeed narrow the focus by only allowing a few websites to be accessed.

My kid's school uses a software called GoGuardian, which allows individual teachers to whitelist specific websites for the students in their class during their class period.


Great video on how this works: https://www.youtube.com/watch?v=mvvVSTlbqEI


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

Search: