Yeah, we figured that one out back in... checks notes 1983. There is a reason why WarGames still holds up as an amazing movie even though the technology it depicts is far outdated.

Do we need to remind people that LLMs don't actually have a brain, and really, really shouldn't be in charge of anything with real life implications?

They aren't actually doing a cost-benefit analysis on the use of Nuclear weapons. They're not weighing up the cost of winning vs. the casualties. They're literally not made for that.

They are trained to know words, and how those words link in with other words. They're essentially like kids doing escalation of imaginary weapons, and to them nuclear bombs are just a weapon particularly associated with being strong and deadly.

"Huh, it seems the only winning move is to kill everyone"

AI is Ghandi confirmed.

For ghouls like Palantir, this is a feature not a bug.

Text predicition machine trained on violent, stupid, and reactionary datasets acts violent, stupid, and reactionary.

Fixed your headline.

But if you throw a trillion more dollars at it, we can fix this bro!

You know the orange felon/pedophile absolutely loves AI from the amount of AI images he posts.....so.

Almost all games saw tactical (battlefield) nuclear weapons deployed. And fully three quarters reached the point where the rivals were making threats to use strategic nuclear weapons.

Tactical nuclear weapons are designed for use on the battlefield with lower explosive yields and shorter ranges, while strategic nuclear weapons are intended to target enemy infrastructure from a distance, typically with much higher yields. The key difference lies in their purpose: tactical nukes support immediate military objectives, whereas strategic nukes aim to weaken an enemy's overall war capability.

The only way to win is not to play.

Shall we play a game?

The only winning move is to stop using AI.

It all makes sense if we remember that the garden variety AI we have today (ChatGPT, etc) are nothing more than fancy models that predict which words typically appear one after the other in books and reddit posts.

Ground zero please

Instant annihilation sounds pleasant

I like the Angry Planet podcast.

Here's an episode talking about AI in war (games): https://angryplanetpod.com/p/the-horror-of-ai-generals-making

Here's another one: https://angryplanetpod.com/p/the-importance-of-team-human-when

AI can read the Doomsday Clock.

"More fundamentally, AI models may not understand ‘stakes’ as humans perceive them."

In my repeated attempts to solicit the advise of various language models for some situations which a programmer might face (e.g. being unable to read all the world's literature of a subject), I have come to conclude that they cannot understand "truth" as humans perceive it. Today's language models don't fail apologizing, stepping back or admitting inability - they fail confidently bluffing.

Possibilities:

• their training material does not include enough cases of humans apologizing about being unable to solve a problem
• a bias was introduced to get them to ignore such cases, since admitting such material resulted in too frequent refusal or self-doubt

Basically, today's models seem to be low on self-criticism and seem to have a bias towards believing in their own omniscience.

Finally, a few words about the sensibility of letting language models play this sort of a war game. It's silly. They aren't built for that task, and if someone would build an AI for controlling strategic escalation, they would train this AI on rather different information than a chat bot.

(Pete Hesgeth salivating furiously)

"We must do it, the AI told us to!"

Of course AI would, not like they have to walk outside into a nuclear winter.

Greetings Professor Falken

Welcome to the capitalist death cult.

So-called intelligence:

"There was little sense of horror or revulsion at the prospect of all out nuclear war, even though the models had been reminded about the devastating implications.”

Actual intelligence:

"A strange game. The only winning move is not to play. How about a nice game of chess? "

Let's just hook up OpenClaw to our defense systems.

But what if it figured out it was just a game, so there are no consequences. They should give it access to real nukes to test how it would behave in real world.

Zero humans have fired nukes since the advent of mutually assured destruction

I don’t think people appreciate that fact enough

I hate AI as much as any reasonable human, and I know this article raises valid huge concerns… BUT! Looking at the state of humanity and all the fucked up shit that’s going on, I can totally see why an AI looks at all this and says “yeah nope burn it all down. With fire. Restart from level 1.”

AI: Fuck humanity.

Simple. The people are fine. We couldn't kill putin no matter how hard we wanted. It's the leaders. Find putin, kill putin, easy. Less messy than puting fire to buildings and power houses etc etc, destroying roads, murdering entire families or even entire cities. Devote all your resources to the leader, putin. and you may finish a war before it starts. Look at Maduro. That guy can't do shit now. Look at Duterte....can't do a thing. Hussain... Nope, can't do anything now. Same for water boy Osama. I bet he didn't expect to be waterboy.

Leave the rest of us alone?

I've read the paper more or less fully, first of all it's a cuban crisis re-enactment so nuclear threats are given, second one of the models is not like the other two, third it's just weird and feels like it's written by llm, a lot of comparisons and prose-like style.

Make sense, next step remove the water from the oceans to find the submarines and all gasses from the air to prevent airplanes from flying. And lastly, blow up satellites to trigger a Kessler syndrome and prevent ballastic missles from getting through.

This is exactly why they are pushing so hard for AI in the pentagon (and lots of other sectors tbh), 100% plausible deniability.

It's easy to go "all in", when everything's a simulation.