Hacker News Reader

Comment by vasco | Hacker News Reader

So at some point most war machinery will be automated and not carry humans inside, I think this is easy to accept. A little bit harder to accept, but what would be really cool is to just realise that if there's no humans, and the whole thing is automated, why not just simulate it?

The same way mythical armies of literature used champions to fight it out - causing less mysery, why not just simulate the war itself and figure out who wins?

I guess the main issue with this is that right now there's multiple components that help you win a war that are still super relevant, logistics, manufacturing and so on. Once algorithms become the dominant aspect and those other physically bound ones are no more, could we move war fully to the digital realm?

View on HN

> why not just simulate it?

The same reason we don’t simulate Facebook Farmville plots to produce all of our food.

War isn’t some hobby soldiers play at to earn points for their side, it’s the continuation of policy by other means. There is no “simulating” it anymore than you can force someone to kill themselves and all their friends and family by asking nicely.

It's not the same thing. The outcome of war is "who gets to impose their will", which is something that can change without any physical change. We could declare tomorrow that there's a new king of the world and if everyone accepts that, nothing actually needs to change. Producing food is not like that.

The point is eventually all that matters is algorithms. If everything is automated, if a lot of war is cyber war etc, you know already that the best algorithms applied to physical stuff will win, so why not just pit them against each other virtually?

As for the argument of "why wouldn't they just go at it in the physical world next", well they could, but after 10 or 20 wars where you see the outcome is going to be aligned to the algorithmic winner anyway you see there's no point other than to kill your own population.

For what it's worth I don't think this happens in 10 or even 100 years - it's stretching it to a much longer timespan.

If we could all agree to go along with who the leader is, then we're all living in the same society.

But that's not the point: the point is "who's going to make me?" when our disagreements are significant.

And significant is minor details like "I believe I own this land" and "I have a right to exist".

EDIT: it's also worth noting that we resolve wars by simulation all the time - that's what diplomacy and it's supporting institutions are. Both sides present their objectives, tend to outline their military forces and escalation priorities to each other, and then go away and decide if they believe the outcome is substantially worse then what could be achieved by negotiation - which with war it usually is.

> The outcome of war is "who gets to impose their will", which is something that can change without any physical change.

That’s not how that works. The nation that gets to impose its will is the one that can put boots on the ground and hold them there. Doesn’t matter if they’re cyborg or robot boots, but they have to physically be there. That’s a lesson as old as history.

I understand what you’re getting at, and maybe someday smaller powers will participate in “simulated warfare” (which means something else) in lieu of total destruction like some hunter-gatherer and monkey tribes do, but it will always be in the physical world and involve death or destruction.

Otherwise, why would anyone go to war even now? If any other mechanism other than physical violence would have worked, we wouldnt go to war.

Agree, I think countries already run their simulations, and when they believe they are on the winning side, they invade. Doesn't matter if the simulation is rolling a dice.

If you don't invade, you haven't taken control.

Or putting it from other perspective, you could run the simulation between 2 parties, be the winner and "impose" your will. But I don't think that the other side will now leave their houses and leave their own city voluntarily, at that point you need to invade anyways.

People don’t fight wars they don’t think they might win. High fidelity simulations might actually help with that.

Do we not have a pair of counterexamples going on right now in current world events.

How so?

> why not just simulate the war itself and figure out who wins?

This only works if the consequences of losing are no big deal.

If all that changes is which set of leaders get to collect taxes vs having to swear off politics and go into industry or whatever, then sure.

If instead the losing population is removed to provide more land for the winners? Maybe expecting them to do as the simulated result says is a bit much

Also on a vaguely related topic, I'd like to note that bank robberies are still a thing despite being well known to be approximately impossible to get away with.

It cannot be simulated because many aspects of military capability are a secret by design and as a matter of functional necessity. The available information is supposed to be inaccurate. Uncertainty around capability and causality is a fundamental part of the deterrent and battlefield shaping effects during conflict.

How do you prove the fidelity of the simulation with respect to actual military capabilities? How do you prevent a party from claiming capability in the simulation that they do not actually possess? Countries like Russia constantly churn out supposed super-weapons that are routinely found out in real combat.

If war became all about simulation engineering rather than actual capability, a country is going to recognize the arbitrage opportunity and take advantage of it in the physical world.

It is a bit like conflating e-sports with the real thing. One is not a substitute for the other because the constraints and incentives are fundamentally different, optimizing for very different things in execution.

1967 Star Trek presented one take on this idea: https://en.wikipedia.org/wiki/A_Taste_of_Armageddon

Awesome, thanks for the share. What a gift to come up with a funky idea and having a ready to go Star Trek episode about it! If only it was like this with all the cooky ideas we come up with.

> If only it was like this with all the cooky ideas we come up with.

It probably is :) Any other ideas you'd like us to come up with a Star Trek episode for? There are nearly a thousand to select from these days!

<https://news.ycombinator.com/item?id=10270273#10270607>

Interesting thought. I guess war is a kind of "proof of work" which identifies the society that is more able to productively employ its resources. This includes strategically and tactically on the battlefield but also all the logistical and economic support. Simulations fall short on that. Cyberwarfare might come closer though as we become more digitalised so maybe that's the direction things will move in?

How do you simulate attacking someone who is not expecting to fight a war with you? What do you do when the simulated loser refuses to concede?

From what i understand this is already mostly the case. The military excercises by USA and China produces reports that shows an attack is too expesive or leads to a standstill. This is works as a deterrence to the other side, unless the other side acts irrationally, which is known to happen…

> So at some point most war machinery will be automated and not carry humans inside, I think this is easy to accept. A little bit harder to accept, but what would be really cool is to just realise that if there's no humans, and the whole thing is automated, why not just simulate it?

War is effectively just an escalation of using force to impose one's will over another. You don't need a war if you're China and trying to get your shipping passage rights through Suez or Panama. You'd need a war it you're Russia trying to obtain land illegally from your neighbors.

What’s to stop the side that loses the virtual war just having another go in reality? Can you really imagine Putin or Saddam Hussein just accepting a loss and giving up?

Meanwhile a delay in real time while the virtual war is prepared and fought would almost alway be significantly more beneficial to one side over the other.

So many other factors, but that’s a start.

What’s to stop the side that loses the virtual war just having another go in reality? Can you really imagine Biden or Zelensky or Netanyahu just accepting a loss and giving up?

I'm dead serious here.

No? What is your point?

Of course, absolutely. It’s a serious question.

There is the issue that Biden or any other elected leaders are answerable to their electorate, and other institutions in their country, in a way that authoritarians are not, so it’s a matter of the commitment of the nation too, but that’s not directly an issue with the simulated war concept. It’s nonsense for purely practical reasons.

War games stop would-be belligerents all the time. It's probably the only thing keeping Taiwan a democracy.

> could we move war fully to the digital realm?

Not if we can't even get belligerents and defenders to uphold their treaty agreements.

I can’t cite any but there were several old (i.e. 70s era) science fiction story where disputes were decided by having, essentially, Champions of sorts battle each other using war simulations. Such as re-fighting some famous ancient battle using a war game console.

Early eSport if you will.

There was a YA book series about exacty this topic: https://www.goodreads.com/book/show/87427.Virtual_War

There was even a Star Trek: TOS episode about this: https://en.m.wikipedia.org/wiki/A_Taste_of_Armageddon

This is fine until the other side flies a plane into your skyscrapers.

A really good read along those lines:

https://en.wikipedia.org/wiki/Peace_on_Earth_(novel)

Imagine the money and human cost that could be saved if nations would agree to settle international disputes over the board game of "Risk."

Reply on news.ycombinator.com