Hacker News new | past | comments | ask | show | jobs | submit
LeoPanthera | on: Listen and read USA nuclear forces Emergency Action Messages, heard on 11175 kHz
This is a cool idea. I'm actually quite glad that AI generated stuff is still so very obvious. The world will be a darker place when it isn't.
publlus_enigma | parent | next
Until the kids learn to speak like Claude. Then it will all be the same.
ricksunny | root | parent
Gen-C coming right up.
ahofmann | parent | next
The obvious AI stuff is obvious, all the rest is either human generated, or you aren't able to detect that it's not.

So I would argue the world is already "a darker place".

dpc050505 | root | parent | next
It's also a lot easier to get caught when you're not paying full attention to whatever is being sent your way. Say an algorythm starts playing AI music while it's just playing in the background, or a friend posts something silly that you read sideways.
lukan | root | parent
"algorythm"

Spelling errors are so far still a good sign of human written output.

0c3ca83 | root | parent | next
Wouldn't it be lovely if we didn't rush headlong into making things worse at the first sign of profit or power?
InsidiousWombat | root | parent
I mean I've seen people saying that whole "oh, it's just so good you can't detect it" a lot but not any actual evidence. I guess sentence can be short enough to not look clauded, but aside from that smells are still very common even for the best frontier models, so how anything else would be indistinguishable ?
quasarj | parent | next
I have no idea how people are apparently spotting AI generated text. At this point, I don't even think I want to know. Really, I guess it's no different to me than text written by another person.
devin | root | parent | next
It's different to me from text written by another person, because it is awful in a way that even a bad human writer can't replicate without first performing a lobotomy on themself.
jiggawatts | root | parent
A great mental model here is that if you talk to many people, each one has their verbal ticks, their repeated phrases, the preferred idioms, but... you don't notice because you speak to each acquaintance only for a few minutes or hours at most at a week. If you really know someone, since you were both kids, or whatever, then you can often learn their particular style and even mimic it.

If you think of the the frontier AI models as people, there aren't hundreds or thousands of "intelligent human-like machines" that you're speaking to, there are like... two: whatever the best model is made by Open AI or Anthropic right now.

A programmer might be reading (or at least skimming) thousands of pages of text from just one (or at most two or three) models a week, week after week, for months now.

It's like the entire world is talking to two aliens that have stepped of their glittering spaceship. The aliens can hold millions of conversations at once, but they have only two personalities, so... that's what everyone is getting, and everyone has learned to recognise.

What makes things even more obvious is that unlike humans, the models don't "self adjust". On the contrary, they are inherently amnesiacs, and hence don't realise that they're awfully repetitive! A human using the same phrase fifty times in a row will dial it back a notch, but these dials are the weights of the models and are frozen for each released version.

10111001 | parent | next
> This is a cool idea. I'm actually quite glad that AI generated stuff is still so very obvious. The world will be a darker place when it isn't.

You mean this site? Because I think the messages it's about is real. The AI stuff is just the dumb copy in the interface (e.g. "Mostly static. The net is quiet until it isn't.").

skygazer | parent | next
My worry is that what we recognize is the unstylized default, but it might be very easy to emulate any particular human easily enough with some examples and a prompt and n-gram/style template biasing token selection, or iterative generation-scoring loops.
rcxdude | root | parent | next
Yeah. The big LLMs have fairly idiosyncratic writing styles. Smaller finetuned LLMs can be quite different (and you can also tune them on a particular person's writing). I think some of the confidence in this is because there's plenty of cases which just don't get noticed.
Larrikin | root | parent
This is probably possible now. Back when machine learning was all the rage people were using it to identify authors of various anonymous works. It became an introductory project to identify the Federalist Papers authors. Just read in some works, identify patterns, set up tests, and set the LLM into a loop testing whether the output would be predicted to be the author you are trying to impersonate.

https://towardsdatascience.com/hamilton-a-text-analysis-of-t...

ajsnigrutin | parent
[flagged]
timcobb | root | parent
Why this, of all things? It's horrible, but look at where the school was, why did they put a school there? Who puts a school next to a military base? There's way darker stuff than people missing a military base and hitting an adjacent school by accident.
Vaushite | root | parent | next
Most of the really important military targets in Israel, such as the Mossad headquarters, are in the middle of residential areas.
timcobb | root | parent
Yeah I was surprised to see that last year when they were targeted (and almost hit? or hit?) by Iran. Yikes.
mullingitover | root | parent | next
> Who puts a school next to a military base?

These days they’re calling themselves the “Department of War”

jacquesm | root | parent | next
The extent to which some people will twist to justify that which can not be justified is mind boggling.
timcobb | root | parent | next
What's the twist? I'm curious why this event was chosen, if for any reason. Personally I am frustrated by people building a school next to a military installation. It feels extremely negligent. I find that to be pretty unsettling and wish people wouldn't do that, because military installations are legitimate and predictable military tagets.
DoctorOetker | root | parent | next
I fully agree, and think the international community is making a big mistake by not (re)formalizing international law and the rules of war, we should force all nations (including those not in conflict -yet-) to store school locations etc. as machine readable maps... we will never be able to roll back the clock and find out if Iran would have properly declared the presence of the school or not, ... but we can PREVENT SIMILAR SITUATIONS in the future, or at the very least have a mechanism which makes it crystal clear when one or more parties are playing very dirty games.

Wouldn't it be much nicer to live in a first world with more clarity, where it is immediately obvious which nation either knowingly bombed a kids school, or what other nation knowingly uses some poor children as human shieds...

But thats not the world we live in: apparently we live in the world where the event also happened, but where the rules weren't set up to provide such clarity a priori. And this second world is IMHO darker than the also dark first world.

10111001 | root | parent | next
> I fully agree, and think the international community is making a big mistake by not (re)formalizing international law and the rules of war, we should force all nations (including those not in conflict -yet-) to store school locations etc. as machine readable maps... we will never be able to roll back the clock and find out if Iran would have properly declared the presence of the school or not, ... but we can PREVENT SIMILAR SITUATIONS in the future, or at the very least have a mechanism which makes it crystal clear when one or more parties are playing very dirty games.

The problem with that idea isn't that a country like Iran would fail to properly declare all of its schools, it's that it would improperly declare military targets as being schools. No one could trust the maps.

If you really wanted to do something like that, the maps would have to be compiled by some neutral 3rd party with on-the-ground verification.

jacquesm | root | parent
Russia would use that information to attack your schools. They don't care about war crimes in some countries, they see them as things you get bonus points for.
LtWorf | root | parent
Oh Russians are bad… we all agree. But it wasn't them who bombed the schools in Iran.
Vaushite | root | parent
What would be the punishment on nations that used the map to decide where to bomb first?
ok_dad | root | parent | next
Did you know nearly every middle to larger sized US military base has base housing and even schools?

That’s the twist.

ok_dad | root | parent
Downvote it all you want; I’m right.
| root | parent
{"deleted":true,"id":50226563,"parent":50226368,"time":1792884721,"type":"comment"}
DoctorOetker | root | parent | next
I think nobody was justifying what happened, just pointing out more than one party did unjustifiable things.
betterunix2 | root | parent | next
I mean...sure...but if you are going to be consistent you are going to have to criticize the Allied victory in WWII. There were literally piles of dead children in cities all over fascist Europe. We do not even have exact numbers because in some cases the death toll had to be estimated by weighing the ashes found in bomb shelters.
kingofmen | root | parent | next
And the civilian bombings, explicitly named "morale bombing" (we called it "terror bombing" when the Axis did the same) were later found to be completely inefficient; a waste of trained crews and expensive bombers that did not contribute to victory at all. Or in other words,

YES.jpg

ajsnigrutin | root | parent
It's one thing to throw thousands of tons of unguided missiles randomly, and it's something else when you brag about the "all knowing AI", how your intelligence services are the best and how your missiles are accurate.
mindslight | root | parent | next
It turned out that the real dark stuff was the fiends we made along the way.
ajsnigrutin | root | parent | next
> Who puts a school next to a military base?

https://www.militaryonesource.mil/education-employment/for-c...

Americans it seems.

Also, in my country, there's a large shopping area next to an army base, very hard to spot which hangar-looking building is full of stores (or a single large store, or a trampoline park) and which belongs to the army.

I mean.. AI knows best, and we should trust it, even when picking targets, right?

segfault99 | root | parent
[flagged]
timcobb | root | parent
That's horrible. If they get bombed, we'll know why...
segfault99 | root | parent
Not every nation thinks it has a God-given right to go half-way around the world and bomb other nation's bases, let alone schools. The kids at Fort Bragg Literacywatsdat Junior High have more to worry about violence-wise from each other cf. some imaginary foreign threat.

Again, really difficult concept to grasp. Take some time.

ok_dad | root | parent | next
You have a good message in general but here at hacker news we try to avoid personal attacks and that sort of thing. Just letting you know why you’re being flagged; but I do agree Americans are generally clueless.
timcobb | root | parent
In my defense, the likelihood that the US has schools on bases did occur to me. It's just I think you shouldn't put schools on or near bases and expect them to not get bombed, if not by accident. So I think it applies to all parties.

> Not every nation thinks it has a God-given right to go half-way around the world and bomb other nation's bases,

I would pause at this assertion. The US is a global thing, with global friends and enemies. Iran being one of our big foes is a knowing thing.