Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Large language models in national security applications

arxiv.org

Post reply on HN
Print view View original

31–40 of 60 posts

Re: Large language models in national security applications

#31

Post by ben_w »

psunavy03 wrote: post #30
ben_w wrote: post #25

Earlier quoted context omitted.

> Can LLMs get in an enemy's head? That may be much easier for an LLM than all the other things you listed. Read their socials, write a script that grabs the voices and faces of their loved ones from videos they've shared, synthesise a video call… And yes, they can write the scripts even if they don't have the power to clone voices and faces themselves. I have no idea what's coming. But this is going to be a wild dec…

Creating chaos and confusion is great, but it's only part of what a military campaign needs. You have to be able to use all levers of government power to put the other government or the adversary organization in a point where they feel compelled to quit or negotiate.

Aye.

FWIW, I hope all those other things remain a long way off.

Whoever's doing war game planning needs to consider the possibility of AI that can do those other things, but I'm going to have to just hope.

Re: Large language models in national security applications

#32

Post by recursive »

exe34 wrote: post #29
JoshTriplett wrote: post #26

Earlier quoted context omitted.

That's not inherently true. If there were a way to reliably destroy all the sock puppets, we should, and the world would be better off. For instance, reliable bot detection, or mechanisms by which major social networks could detect and prohibit bot-like activity that isn't labeled as such.

charge one penny for every post. most people can afford it. bots become less cost effective and you'd be able to trace the source of funds.

Even if they could afford it, they won't. The UX friction of charging money would send engagement off a cliff, even if the nominal charge was $0.

Re: Large language models in national security applications

#33

Post by exe34 »

recursive wrote: post #32
exe34 wrote: post #29

Earlier quoted context omitted.

charge one penny for every post. most people can afford it. bots become less cost effective and you'd be able to trace the source of funds.

Even if they could afford it, they won't. The UX friction of charging money would send engagement off a cliff, even if the nominal charge was $0.

that's why it should be mandated by law for national security reasons!

Re: Large language models in national security applications

#35

Post by paganel »

nyrikki wrote: post #14
gmaster1440 wrote: post #4

The paper argues against using LLMs for military strategy, claiming "no textbook contains the right answers" and strategy can't be learned from text alone (the "Virtual Clausewitz" Problem). But this seems to underestimate LLMs' demonstrated ability to reason through novel situations. Rather than just pattern-matching historical examples, modern LLMs can synthesize insights across domains, identify non-obvious patter…

You are using a different definition of strategic than the DoD uses, what you are describing is closer to tactical decisions. They are talking about typically Org wide scope, long-term direction . They aren't talking about planning hidden as 'strategic planning' in the biz world. LLMs are powerful, but are by definition past focused, and are still in-context learners. As they covered, hallucinations, adverse actions,…

> but are by definition past focused,

To add to that, and because the GP had mentioned (a "virtual") Clausewitz, "human"/irl strategy itself has in many cases been too focused on said past and, because of that, has caused defeats for the adopters of those "past-focused" strategies. Look at the Clausewitzian concept of "decisive victory" which was adopted by German WW1 strategists who, in so doing, ended up causing defeat for their country.

Good strategy is an art, the same as war, no LLM nor any other computer code would be ever able to replicate it or improve on it.

Re: Large language models in national security applications

#36

Post by JohnMakin »

psunavy03 wrote: post #22
gmaster1440 wrote: post #10

Earlier quoted context omitted.

The "second year university student" analogy is interesting, but might not fully capture what's unique about LLMs in strategic analysis. Unlike students, LLMs can simultaneously process and synthesize insights from thousands of historical conflicts, military doctrines, and real-time data points without human cognitive limitations or biases. The paper actually makes a stronger case for using LLMs to enhance rather tha…

But the aide won't have deeply analyzed every military campaign in history; it will only spout off answers from books about those campaigns. It will have little to no insight on how to apply principles and lessons learned from similar campaigns in the current problem. Wars are not won by lines on maps. They're not won by cool gear. They're won by psychologically beating down the enemy until they're ready to surrender…

The person you are responding to seems to be promoting a concept that is frequently spouted here and other places, but to me lacking sufficient or any evidence - that AI models, particularly LLMs, are both capable of reasoning (or what we consider reasoning) around problems and generating novel insights that it hasn't been trained on.

Re: Large language models in national security applications

#37

Post by idle_zealot »

bilbo0s wrote: post #16
perihelions wrote: post #5

The most obvious way the US national security industry could use LLM's right now is simply to spam foreign adversaries with chatbots. That's their greatest strength right now—a use-case they have amply proven themselves for. This paper comes off as eager to avoid this topic: they (briefly) talk about detecting foreign LLM spam, which is called propaganda, but sidestep the idea of our own side using it. If we were con…

Or, to rephrase this as a question: Is it ethical to spam another democracy with millions of chatbots pretending to be citizens of that country—if the effect is to manipulate those citizens to [take any action advantageous to US national interest]...? Just, Devil's Advocate, but ethical or not, that's what we should be doing and what we are doing. Every nation has its sock puppets out there, our job is to stop everyo…

I don't think you understand the Devil's Advocate rhetorical device. It's not just a thing you say before you state a morally reprehensible opinion you hold.

Re: Large language models in national security applications

#38

Post by alisonatwork »

exe34 wrote: post #29
JoshTriplett wrote: post #26

Earlier quoted context omitted.

That's not inherently true. If there were a way to reliably destroy all the sock puppets, we should, and the world would be better off. For instance, reliable bot detection, or mechanisms by which major social networks could detect and prohibit bot-like activity that isn't labeled as such.

charge one penny for every post. most people can afford it. bots become less cost effective and you'd be able to trace the source of funds.

This is a nice idea, but we already know that organized criminals and state actors have no problem spending money to make money (or influence public opinion).

On top of that, social media isn't like email where a potential transaction fee increases the cost per person reached - a single post on a hot topic could get millions of views.

Re: Large language models in national security applications

#40

Post by itishappy »

bilbo0s wrote: post #16
perihelions wrote: post #5

The most obvious way the US national security industry could use LLM's right now is simply to spam foreign adversaries with chatbots. That's their greatest strength right now—a use-case they have amply proven themselves for. This paper comes off as eager to avoid this topic: they (briefly) talk about detecting foreign LLM spam, which is called propaganda, but sidestep the idea of our own side using it. If we were con…

Or, to rephrase this as a question: Is it ethical to spam another democracy with millions of chatbots pretending to be citizens of that country—if the effect is to manipulate those citizens to [take any action advantageous to US national interest]...? Just, Devil's Advocate, but ethical or not, that's what we should be doing and what we are doing. Every nation has its sock puppets out there, our job is to stop everyo…

Why stop there? Send missiles! Heck, send bioweapons and dirty bombs. It might not be ethical but if others doing it we might as well join 'em.

(This is sarcasm. I don't want to live in this world.)

Post reply on HN