OpenAI wants to not be responsible for "accidents" that kill more than 100 people, despite some critics arguing that their current actions are likely to cause such harms.
OpenAI backs Illinois bill that would limit when AI labs can be held liable
171–180 of 343 posts
Re: OpenAI backs Illinois bill that would limit when AI labs can be held liable
#172I 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…
Yes fortunately it is really bad at actually making novel bioweapons or syntheses in general so whatever you made probably wouldn't do more than give someone a mild headache.
I am not so sure about that, trial and error can produce very dangerous results especially over the span of years or even decades.
Re: OpenAI backs Illinois bill that would limit when AI labs can be held liable
#173I 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…
> All these findings were reported to both openai and anthropic and they were not interested in responding Let’s dive into why. When we run normal bounty and responsible disclosure programs there’s usually some level of disregard for issues that can’t / won’t be fixed. They just accept the risk. Perhaps because LLMs don’t have a clean divide between control and input that’s makes the problem unsolvable. Yes. You can…
LLM providers are not obliged to only use LLMs to guard against hazardous output. They could use other automated and non-automated techniques. And they ought to do so if they are given good evidence that existing safeguards are inadequate. Loss of product quality or additional cost should be secondary.
Re: OpenAI backs Illinois bill that would limit when AI labs can be held liable
#174No 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
Re: OpenAI backs Illinois bill that would limit when AI labs can be held liable
#175Re: OpenAI backs Illinois bill that would limit when AI labs can be held liable
#176Earlier quoted context omitted.
Countless downloadable models (including de-aligned mainstream models) can do this.
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 LLMs in software development (so, not Claude Mythos) are still at best junior developers, and apply that to non-software roles.
This past February I tried to use Codex to make a physics simulation. Even though it identified open source libraries to use, instead of using them it wrote its own "as a fallback in case you can't install the FOSS libraries"; the simulation software it wrote itself was showing non-physical behaviour, but would I have known that if I hadn't already been interested in the thing I was trying to get it to build me a simulation of? I doubt it.
Re: OpenAI backs Illinois bill that would limit when AI labs can be held liable
#177I 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…
If someone were inclined to attempt producing nefarious agents in this category, is this not also available on the plain web? I would search to answer my own question, but I'll defer that task for obvious reasons.
Re: OpenAI backs Illinois bill that would limit when AI labs can be held liable
#178Earlier quoted context omitted.
I think you've just summed up late stage capitalism.
People on both sides seem to give capitalism a lot of credit for human traits that existed long before capitalism.
Well the point of capitalism (going back to Adam Smith) is that the invisible hand converts locally selfish behavior to globally good outcomes. The argument is whether or not that emerges. So if your implication was that human trait was selfishness, yes, that is quite the point of capitalism.
Re: OpenAI backs Illinois bill that would limit when AI labs can be held liable
#179Is this for like military scenarios or like, ChatGPT designed a drug that seemed to work, but people died by the millions 5 years later? Because they should 100% be liable for the latter. The former, good luck trying to prosecute an AI company for something the military does. To an extent, the military would probably want their AI models to be behind their private network, completely firewalled from any public networ…
Shouldn’t the pharmaceutical company be held liable for insufficiently understanding the drug before releasing it? I don’t think I understand blaming a tool used in the process of designing it and not those who chose to release it.
Pharmaceuticals are heavily regulated, the "we vibecoded a therapeutic and released it without testing" hypothetical has no basis in reality
Re: OpenAI backs Illinois bill that would limit when AI labs can be held liable
#180I 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.
Presumably they mean the fundamental failure mode of LLMs that if you fill their context with stuff that stretches the bounds of their "safety training", suddenly deciding that "no, this goes too far" becomes a very low-probability prediction compared to just carrying on with it.