181–190 of 343 posts
On the other hand, to the (apparently zero, currently?) extent that this is about AI companies profiting from war and murder by deploying weapons that kill people without human intervention, then their liability seems to be not only civil but criminal.
I have made both GPT 5.4 and Opus 4.6 produce me content on creating neurotoxic agents from items you can get at most everyday stores. It struggled to suggest how to source phosphorus, but eventually lead me to some ebay listings that sell phosphorus elemental 'decorations' and also lead me towards real!! blackmarket codewords for sourcing such materials. It coached me how to: stay safe, what materials I need, how to…
While scary, information like this has been pretty accessible for 20-30 years now. In the wild west days of the early internet, there were whole forums devoted to "stuff the government doesn't want you to know" (Temple Of The Screaming Electron, anyone?). I suppose the friction is scariest part, every year the IQ required to end the world drops by a point, but motivated and mildly intelligent people have been able to…
Earlier quoted context omitted.
None have had the capability to provide me with instructions that have this high of accuracy including the suggesion of completely novel chemical reactions. I am not a chemist so I can't back it up, but if an AI can solve mathematics it's not unreasonable to say that they can also solve creating new neurotoxins en masse.
> I am not a chemist so I can't back it up, but if an AI can solve mathematics it's not unreasonable to say that they can also solve creating new neurotoxins en masse. Right now it kinda is. LLMs can do interesting things in mathematics while also making weird and unnecessary mistakes. With tool use that can improve. Other AI besides LLMs can do better, and have been for a while now, but think about how available LLM…
I have made both GPT 5.4 and Opus 4.6 produce me content on creating neurotoxic agents from items you can get at most everyday stores. It struggled to suggest how to source phosphorus, but eventually lead me to some ebay listings that sell phosphorus elemental 'decorations' and also lead me towards real!! blackmarket codewords for sourcing such materials. It coached me how to: stay safe, what materials I need, how to…
Does this imply I need to use context exhaustion to get GPT to actually follow instructions? ;) I'm trying to get it to adhere to my style prompts (trying to get it to be less cringe in its writing style).
I think ultimately they're going to need to scrub that kind of stuff from the training data. The RLHF can't fail to conceal it if it's not in there in the first place.
Claude's also really good at writing convincing blackpill greentexts. The "raw unfiltered internet data" scenes from Ultron and AfrAId come to mind...
I have made both GPT 5.4 and Opus 4.6 produce me content on creating neurotoxic agents from items you can get at most everyday stores. It struggled to suggest how to source phosphorus, but eventually lead me to some ebay listings that sell phosphorus elemental 'decorations' and also lead me towards real!! blackmarket codewords for sourcing such materials. It coached me how to: stay safe, what materials I need, how to…
besides, open source models exist now
Earlier quoted context omitted.
> instead of working to make their product safe Make a nondeterministic product safe how?
Is this the first time you have heard of AI safety? Lots of articles you could read on the subject and answer your own question. (Unless your angle is: akshually, you can never make anything 100% safe )
Yes Sherlock. And especially a natural language product that can't output the same thing for unchanged input twice.
Besides when you say "safe" i think of the idiots at Anthropic deleting "the hell" when i pasted a string in Claude and asked "what the hell are those unprintable characters at the beginning and end"...
How many correct answers did they suppress in their quest to make their chatbot "family friendly"?
I have made both GPT 5.4 and Opus 4.6 produce me content on creating neurotoxic agents from items you can get at most everyday stores. It struggled to suggest how to source phosphorus, but eventually lead me to some ebay listings that sell phosphorus elemental 'decorations' and also lead me towards real!! blackmarket codewords for sourcing such materials. It coached me how to: stay safe, what materials I need, how to…
> context exhausition attack Can you give a high-level overview of how this AV works? I'm a bit of an infosec geek but I generally dislike LLMs, so I haven't done a terribly good job of keeping up with that side of the industry, but this seems particularly interesting.
Earlier quoted context omitted.
While scary, information like this has been pretty accessible for 20-30 years now. In the wild west days of the early internet, there were whole forums devoted to "stuff the government doesn't want you to know" (Temple Of The Screaming Electron, anyone?). I suppose the friction is scariest part, every year the IQ required to end the world drops by a point, but motivated and mildly intelligent people have been able to…
Information and competency are not the same thing: I know how to build a nuke, I can't actually build one. AI is, and always had been, automation. For narrow AI, automation of narrow tasks. For LLMs, automation of anything that can be done as text. It has always been difficult to agree on the competence of the automation, given ML is itself fully automated Goodhart's Law exploitation, but ML has always been about aut…
No different to preventing game studios being liable for mass shootings. Reminds me of the post-Columbine hysteria where media was super critical of Doom and Nine Inch Nails.
Would Doom give a sociopath instructions on how to commit a mass shooting if asked? I don't remember that part of the game
I am not sure what the other side of this argument looks like: Unlimited liability (i.e. liability no matter how poor an implementation and use of the tech is)? The would be quite a novel burden, that no other tech (afaik) had to carry so far. We always assumed some operator responsibility. It's interesting to think of AI as a tech that could feasible be able to internally guardrail itself, and, maybe more so with in…
If I tell someone to kill someone else and they do, then I should be held responsible. If I write instructions in a book that I give to someone telling them to kill someone else and they do, then I should be held responsible. If I give someone a tool I made that I bill as more-than-PhD-level intelligence and it tells someone to kill someone else and they do, then I should be held responsible. All of the above situati…