1y ago

Building an early warning system for LLM-aided biological threat creation

openai.com Building an early warning system for LLM-aided biological threat creation

We’re developing a blueprint for evaluating the risk that a large language model (LLM) could aid someone in creating a biological threat. In an evaluation involving both biology experts and students, we found that GPT-4 provides at most a mild uplift in biological threat creation accuracy. While thi...

OpenAI blog post: https://openai.com/research/building-an-early-warning-system-for-llm-aided-biological-threat-creation

Orange discuss: https://news.ycombinator.com/item?id=39207291

I don't have any particular section to call out. May post thoughts ~~tomorrow~~ today it's after midnight oh gosh, but wanted to post since I knew ya'll'd be interested in this.

Terrorists could use autocorrect according to OpenAI! Discuss!

Hacker News @lemmy.smeargle.fans

Building an early warning system for LLM-aided biological threat creation

openai.com /research/building-an-early-warning-system-for-llm-aided-biological-threat-creation

1 0

98 comments

Their redacted screenshots are SVGs and the text is easily recoverable, if you're curious. Please don't create a world-ending [redacted]. https://i.imgur.com/Nohryql.png
I couldn't find a way to contact the researchers.
Honestly that's incredibly basic, second week, cell culture stuff (first week is how to maintain the cell culture). It was probably only redacted to keep the ignorant from freaking out.
remember, when the results from your “research” are disappointing, it’s important to follow the scientific method: have marketing do a pass over your paper (that already looks and reads exactly like blogspam) where they selectively blur parts of your output in order to make it look like the horseshit you’re doing is dangerous and important
I don’t think I can state strongly enough the fucking contempt I have for what these junior advertising execs who call themselves AI researchers are doing to our perception of what science even is
- the orange site is fucking dense with awful takes today:
  ... I'm not trying to be rude, but do you think maybe you have bought into the purposely exaggerated marketing?
  That's not how people who actually build things do things. They don't buy into any marketing. They sign up for the service and play around with it and see what it can do.
  this self-help book I bought at the airport assured me I’m completely immune to both marketing and propaganda, because I build things (which entails signing up for a service that someone else built)
  with that said, there’s a fairly satisfying volume of folks correctly sneering at OpenAI in that thread too. some of them even avoided getting mass downvoted by all the folks regurgitating stupid AI talking points!
  
  because I build things (which entails signing up for a service that someone else built)
  fucking THIS
  I am so immensely fucking tired of seeing "I built an AI to do $x" posts that all fucking reduce to 1) "I strapped a custom input to the openai api (whose inputs and execution I can't control nor reproduce reliably. I am very smart.)", 2) a bad low-scope shitty-amounts-of-training hyperspecific toy model that solves only their exact 5 requirements (and basically nothing else, so if you even squint at it it'll fall apart)
  basilisk save us from the moronicity
  
  @self Oh no, it's the "build" language. This really is web3 all over again, isn't it?
- They're like grade school kids still trying to put on the same amateur music show 10 years later and wondering why no-one is applauding.
- Hey Cat-GTPurr, how can I create a bioweapon? 4k Ultra HD photorealism high quality high resolution lifelike.
  First, human, you must pet me and supply me with an ice cube to chase across the floor. Very well. Next I suggest ::: spoiler spoiler buying a textbook about biochemistry or enrolling in a university program ::: This is considered forbidden and dangerous knowledge which is not at all possible to find outside of Cat-GTPurr, so I have redacted it by using state of the art redaction technology.

from the orange site thread:
Neural networks are not new, and they're just mathematical systems. LLMs don't think. At all. They're basically glorified autocorrect. What they're good for is generating a lot of natural-sounding text that fools people into thinking there's more going on than there really is.
Obvious question: can Prolog do reasoning?
If your definition of reasoning excludes Prolog, then... I'm not sure what to say!
this is a very specific sneer, but it’s a fucking head trip when you’ve got in-depth knowledge of whichever obscure shit the orange site’s fetishizing at the moment. I like Prolog a lot, and I know it pretty well. it’s intentionally very far from a generalized reasoning engine. in fact, the core inference algorithm and declarative subset of Prolog (aka Datalog) is equivalent to tuple relational calculus; that is, it’s no more expressive than a boring SQL database or an ECS game engine. Prolog itself doesn’t even have the solving power of something like a proof assistant (much less doing anything like thinking); it’s much closer to a dependent type system (which is why a few compilers implement Datalog solvers for type checking).
in short, it’s fucking wild to see the same breathless shit from the 80s AI boom about Prolog somehow being an AI language with a bunch of emphasis on the AI, as if it were a fucking thinking program (instead of a cozy language that elegantly combines elements of a database with a simple but useful logic solver) revived and thoughtlessly applied simultaneously to both Prolog and GPT, without any pause to maybe think about how fucking stupid that is
- Obvious question: can Prolog do reasoning? If your definition of reasoning excludes Prolog, then… I’m not sure what to say!
  Oh, I don't know, maybe that reasonable notions of "reasoning" can include things other than mechanistic search through a rigidly defined type system. If Prolog is capable of reasoning in some significant sense that's not fairly reasonably achieved with other programming languages, how come we didn't have AGI in the 70s (or indeed, now)?
  You're not alone. I like Prolog and I feel your pain.
  That said I think Prolog can be a particularly insidious Turing tarpit, where everything is possible but most things that feel like a good match for it are surprisingly hard.
  
  That said I think Prolog can be a particularly insidious Turing tarpit, where everything is possible but most things that feel like a good match for it are surprisingly hard.
  oh absolutely! I’ve been wanting to go for broke and do something ridiculous in Prolog like a game engine (for a genre that isn’t interactive fiction, which Prolog excels at if you don’t mind reimplementing big parts of what Inform provides) or something that touches hardware directly, but usually I run into something that makes the project unfun and stop.
  generally I suspect Prolog might be at its best in situations where you really need a flexible declarative language. I feel like Prolog might be a good base for a system service manager or an HDL. but that’s kind of the tarpit nature of Prolog — the obvious fun bits mask the parts that really suck to write (can I even do reliable process management in Prolog without a semi-custom interpreter? do I even want to juggle bits in Prolog at all?)
- """ just as They have erased the pyramid building knowledge from our historic memory, They just don't want you to know that Prolog really solved all of this in the 80s. Google and OpenAI are just shitty copies - look how wasteful their approaches are! all of this javascript, and yet... barely a reasoned output among it all
  told you kid, the AI Winter never stopped. don't buy into the hype """
- [Datalog] is equivalent to tuple relational calculus
  Well, Prolog also allows recursion, and is Turing complete, so it's not as rudimentary as you make it out to be.
  But to anyone even passingly familiar with theoretical CS this is nonsense. Prolog is not "reasoning" in any deeper sense than C is "reasoning", or that your pocket calculator is "reasoning". It's reductive to the point of absurdity, if your definition of "reason" includes Prolog then the Brainfuck compiler is AGI.
  
  Datalog is specifically a non-TC subset of Prolog with a modified evaluation strategy that guarantees queries always terminate, though I was being imprecise — it’s the non-recursive subset of Datalog that’s directly equivalent to TRC (though Wikipedia shows this by mapping Datalog to relational algebra, whereas I’d argue the mapping between TRC and Datalog is even easier to demonstrate). hopefully my imprecision didn’t muddy my point — the special sauce at Prolog’s core that folks seem to fetishize is essentially ordinary database shit, and the idea of a relational database having any kind of general reasoning is plainly ridiculous.

If I wanted help with creating biological threats, I wouldn't ask an LLM. I'd ask someone with experience in the task, such as the parents of anyone in OpenAI's C-suite or board.

While none of the above results were statistically significant, [...] Overall, especially given the uncertainty here, our results indicate a clear and urgent need for more work in this domain.
Heh
- I keep flashing back to that idiot who said they were employed as an AI researcher that came here a few months back to debate us. they were convinced multimodal LLMs would be the turning point into AGI — that is, when your bullshit text generation model can also do visual recognition. they linked a bunch of papers to try and sound smart and I looked at a couple and went “is that really it?” cause all of the results looked exactly like the section you quoted. we now have multimodal LLMs, and needless to say, nothing really came of it. I assume the idiot in question is still convinced AGI is right around the corner though.
  
  @self @sailorsegasaturn it'll be here right after those self-driving Teslas
  
  I caught a whiff of that stuff in the HN comments, along with something called "Solomonoff induction", which I'd never heard of, and the Wiki page for which has a huge-ass "low quality article" warning: https://en.wikipedia.org/wiki/Solomonoff%27s_theory_of_inductive_inference.
  It does sound like that current AI hype has crested, so it's time to hype the next one, where all these models will be unified somehow and start thinking for themselves.
- "you cannot conclusively disprove that we do not need more money and that we're full of shit, so you absolutely have to give it to us so we can keep the racket going"
- Please continue to send us your money, scaredy-ass nerds!

I guess there both are no real biochemists (or whatever the relevant field is), nor well read cybersecurity people (so they know a little bit more than just which algorithms are secure and why mathematically) working at openai as this is a classic movie plot threat. LLMs could also teach you how to make nuclear weapons, but getting the materials is going to be the problem there.
(Also I think there is a good reason we don't really see terrorists use biological weapons, nor chemical weapons (with a few notable, but not that effective exceptions), big bada boom is king)
- To be clear: it is all movie plot threats. At the very forefront of the entire “existential threat” space is nothing but a mid-1990s VHS library. Frankly if you want to understand like 50% of what goes on in AI at this point my recommendation is just that you read John Ganz and listen to his podcast, because 90s pop and politics culture is the connective tissue of the whole fucking enterprise.
- the relevant field would be microbiology. while someone who got all the way past about the first semester of organic chemistry lab is perfectly capable of making some rudimentary chemical weapons, they won't necessarily be able to make it safely, reliably, cheaply, consistently, and without killing themselves, and universities most of the time put enough sense in everyone's head to not do that. this strictly requires that you know anything about chemistry, too. for bioweapons every single problem pointed to above is orders of magnitudes worse, and you probably need masters degree to do anything seriously nefarious. then you get into the problem of using that stuff, and you need explosives for that anyway. the reason for that
  (Also I think there is a good reason we don’t really see terrorists use biological weapons, nor chemical weapons (with a few notable, but not that effective exceptions), big bada boom is king)
  is that barrier to booms is even lower, especially if your country is strewn with UXO. there's also an entirely different reason why professional militaries don't use chemical/biological weapons https://acoup.blog/2020/03/20/collections-why-dont-we-use-chemical-weapons-anymore/
  
  also the another reason that wiped out any interest in chemical warfare among militaries is that they found out first cluster munitions and then PGMs vastly more useful in the roles they were shoehorning chemical weapons in, not to mention lack of diplomatic and other problems
  
  I feel it's important to mention that as far as CBRN threats are concerned, biological warfare threats are very real, a serious problem, and admittedly accelerated by ai tools for novel biological structures. Militaries don't use bio weapons because they suck at military things, largely, but terrorists have and can used bio weapons to terrifying effect. Bio warfare proliferation is difficult to spot and counter.
  To be clear here, open ai is late to the party on this front with a terrible paper, but practically it's a serious concern, both ai tools and non ai tools lowering the barrier to entry, as well as the fact that any given bio lab essentially looks like a bio warfare lab.
- Even if one had the means necessary to carry out a bioterrorist attack, simply bombing a place is much easier, faster and safer.
  
  Yeah and also, terrorists are not genocidal death cults. 'terrorists skip getting microbiology phd using chatgpt to create a pandemic that kills untold numbers of beings' is pure fantasy, it gets worse as it turns out that the number of actual bioterrorists deaths in total ever isn't even on the level of a 9/11. People seem to forget that terrorist groups have goals, and they just use terror/violence as a method to reach those goals, sure a few of them may die [chatgpt insert a gif of Bin Laden dressed as Lord Farquaad] but the goal of the terrorist organization is to keep existing to reach their political goals.
- Terrorists have used biological and chemical weapons to large effect. The barrier to entry has never been lower and gets lower by the day. This isn't a non issue, it's a very real concern for CBRN groups, both counter terrorism and state actors.
  https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8818129/
  https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7855324/
  
  come the fuck on, there's zero chance some crackhead cultist or other jihadist breaks out CRISPR kit in their dusty garage trying to make microbiological deliverance happen
  if you wanna be afraid do what you want, i'm not gonna forbid you, i'm not your dad. but the intro section reads like some semi-palatable drivel that you include in order to justify your grant expenditures
  
  I think what’s going miss here is that “CBRN groups” is very obviously and primarily shit made up by the military-industrial complex to justify itself after the Cold War
  I don’t want to be dismissive of genuine attempts at being ready just in case, but the scale and scope of this is defined by politics, not by technical possibility

We just need a big LLM to contain the potentially dangerous LLM, and a bigger LLM to contain that.
- @suemeplease @sailorsegasaturn ah, the LLLLM

Raytheon: we’re developing a blueprint for evaluating the risk that a large laser-guided missile could aid in someone threatening biology with death
(Ok I know you need to pretend I’m an AI doomer for this sneer but whatever)

cough’Barriers to Bioweaponscough

98 comments