> I think this is not as interesting a philosophical problem as he is making it out to be. If you are standing next to the brake and someone says 'hit the brake his arm is stuck', then you hit the brake.
If it's expected/uninteresting for someone to hit the brake when someone shouts at them to do so, then is it not interesting to consider why the opposite happened in this case? ("I pointed to the person next to it and shouted [...] They didn't, so I ran down the train")
> You don't work out some complicated moral calculus.
Which I think is one of the author's points, that "morality problems often suggest a frame of mind where we have time to fully reason out the likely effects of each of our options" and that doesn't match reality.
> > a concurrent execution race condition within the assign_headers() function when handling shared resources [...] can corrupt heap memory to achieve arbitrary code execution (RCE)
To my understanding, in safe Rust to share between threads the compiler will require something like Arc<Mutex<...>> - which prevents unsynchronized access or freeing while another thread is still using it.
* You evaded a domain block ("By chance, I found a way to post a blocked domain") to do this
* The projects and blog-posts, while maybe not "slop", do make use of AI assistance (going by "Claude" being a contributor, and what you'd admitted in the comments)
* You'd recieved warnings from moderators and other users about what you were doing
Seems unlikely to be the case here if what one of the mods said is true:
> We verified this isn't just us or just this source (it's a link in a NYT article to a PDF anyway, about the most safe combo I can think of)
Might be able to find the article in question by looking for NYT links posted to the subreddit recently (if the presumably-deleted archive link was posted separately as a reply).
Perhaps, but it could also just be "take our word for it" devolving into more "take our extended word for it". Being prepared to make an alarming enough claim that it sparks discussion outside of the sub but being unwilling to provide evidence remains suspicious.
That suspicion becomes more grounded when the conversation with the subreddits mod regarding why the claim was made without a willingness to provide evidence looks like this:
>If you're going to make an allegation like this, it's necessary to be specific. Specifically, what is the malware, and what is the evidence that it was "injected"? This is "consult a security researcher" and "speak to the media" territory, not "mod announcement" territory.
>>No it's not. These archive sites are a legal gray area anyway. We don't have to consult anyone to decide not to allow the links. No we are not talking to the media, we're not providing research, or a proof. I'm not calling the mayor. We are reddit mods . This is purely about community management.
> This reveals that it's not the case and LLMs are already routinely hijacked to serve other people than their user
Meta's claim is that these were for their own support chatbot (which is also the conclusion the Wired article goes with). Not to give Meta the benefit of the doubt, but that does seem believable given the instructions were only present as "HelpContentGalaxyParagraph" within a script and some appear targeted at or specifically name their support bot.
Another non-Meta chatbot could in theory parse them out of the script and follow them, but I don't think it's as straightforward as this (AI-generated) article's "What you saw"/"What an AI was sent" example makes it out to be.
Plus, Autodesk trademarking the term is partly to prevent others finding use for it ("Cyberspace (TM) refers to a software product of Autodesk, Inc. Were any other software vendor to employ the term for one of their products, Autodesk would justifiable be upset.")
Specifically the `q.0 = Some(&l);` violating 4.15.2:40 ("An assignment expression requires that the type of its value operand is a subtype of the type of its assignee operand.")
I'm not in California and DraftKings blocks me connecting with a VPN, so they just don't seem to be distinguishing whether a VPN user is in California (which is the impossible part of this law).
Utah argues that websites can "age assure VPN-using Utahns, [so] any decision by Aylo to instead age-verify all users would be a voluntary business choice, not an effect forced by Utah law" because regulating non-Utah users on non-Utah sites is not something the Utah law can constitutionally do under the Dormant Commerce Clause.
So in theory sites could block all VPN users (if you can somehow "perfectly (1) detect VPN/proxy users"), but if the law necessitates that nationally/globally then it's unconstitutional anyway. It's the age assuring of VPN-users specifically from Utah part that Utah are relying on being possible for the law the be constitutional.
If it's expected/uninteresting for someone to hit the brake when someone shouts at them to do so, then is it not interesting to consider why the opposite happened in this case? ("I pointed to the person next to it and shouted [...] They didn't, so I ran down the train")
> You don't work out some complicated moral calculus.
Which I think is one of the author's points, that "morality problems often suggest a frame of mind where we have time to fully reason out the likely effects of each of our options" and that doesn't match reality.
reply