Comment by gmueckl | Hacker News Reader

Comment by gmueckl | Hacker News Reader

Ender's Game is an interesting story. If we were to simplify the scenario to a lab, in which a monkey gets trained to play an abstract game to get food, the resolution is trivial: if soneone wired the game to act as a trigger for an A bomb under the enemy's capital, hoping to blame the monkey for pushing the final button, they would clearly be in the wrong. The monkey would not be capable of understanding his choice or his actions. The moral responsibility and guilt lies entirely with the people planning and executing that plan.

Ender is different: he knows about the war. He knows that he is being trained to fight in it. His fighting game gets ultimately wired up the the actual battlefield. The book is actually surprisingly ambivalent on whether Ender truely believes that this is still a simulation or something else. And when he is presented with the trigger (the MD device) he decides to press it. If I remember, he even kills half of his own fleet with that thing just because he does not see any other way to win.

I don't quite buy that Ender is complete innocence in this. He was trained to command these battles and knew what consequences his decisions have. Yet he makes the conscious decision to act as an intelligent being who is fundamentally able to decide otherwise if he wanted to. The alternative is set up to have terrible consequences, but it exists and Ender understands it. And the most damning aspect of all in my opinion is that Ender seems to suspect that it is not just a simulation anymore. Yet he continues as he always has.

Then again understanding your choices and actions relies heavily on being aware of their consequences. So a soldier trained to kill cannot be judged by what he does in a simulation (or what they believe to be one). The consequences of your actions are what really define them.

So an Ender playing in a simulation is no different from a monkey being taught to press a button. One of them doesn't know why it's doing that and can't fathom the consequences but there's a reward at the end, the other concludes there are no real consequences but there's a reward at the end. He has not yet been faced with the real situation so it's impossible to to decide whether he would act the same way or not.

But if you are that soldier and you develop reasonable doubt that whatever is said to be a simulation is in fact reality, what do you do?

As a civilian I am certain my reactions, instincts, and decisions are vastly different from those of a soldier. I am not trained to follow orders and take lives regardless of my opinions. I am allowed to have doubts and act on those doubts as I see fit without fear of a court marshal.

At this time in real life any soldier should have reasonable doubt about whether their next target is an enemy combatant or a civilian based on an almost endless record of mistakes and mishaps. But questioning the order or entertaining that sense of doubt could be the difference between surviving and getting your whole squad killed.

Now take that soldier and put them in a simulation and you raised the bar 1000 fold on whether they should worry about real life consequences. They're trained not to worry when they know it's real, let alone when it might not be real at all.

The thing with soldiers is that the whole military culture instills a sense of following orders, the "reasonable doubt" bar is set a lot higher before you can expect them to oppose or even entertain that doubt and investigate.

I was a conscript in the German army and there the emphasis is definitely not put on creating soldiers that follow orders mindlessly. The way the German army is organized allows (and to some extent demands) that soldiers get a degree of freedom about how to arrived at their set goals (in contrast to micromanagement by commanding officers). Attached to that is a responsibility for every soldier to recognize unlawful orders and disobey them.

Speaking as a veteran and someone who was in an OPFOR unit, we trained in "simulations" as close as possible to the way we fight.

> So a soldier trained to kill cannot be judged by what he does in a simulation (or what they believe to be one).

Interestingly, EXROE (exercise rules of engagement) violations are subject to article 15 non-judicial punishment, so they are judged by what they do in a simulation in the US military.

Our simulations are more realistic than the ones depicted in Ender's game, which were pretty strange looking back at it. In ours, you're in full kit, you're carrying your service weapon with a MILES laser attached and a blank fire adapter, there are helicopters flying overhead, and we have villages with imams and such, etc. And there are consequences... if the OPFOR wins the town will go "red" and demonstrations will break out, if BLUFOR wins the town goes "blue" and OPFOR is forced to leave.

The big misconception people have is there are all these instant decisions you have to make or ironclad orders that have to be executed in an instant. The reality is you're responsible for what you shoot at, and that's why you're trained in target identification to know what you're shooting at. You generally don't want to be in a tactical situation where you have to make instant decisions; that's why the cordon and search tactics (similar to SWAT tactics) were so controversial precisely because going room to room meant making large numbers of snap decisions. The results were pretty predictable.

Well developed rules of engagement and tactics help soldiers distinguish clearly between enemy combatants, friendlies and civilians. For example, if you are defending a gate, you're going to set up barriers to slow vehicles down so you have plenty of time to identify a potential VBIED. And you mark out different zones... this is where you flash lights, this is where you fire a warning shot, and if they get past this point you open up.

That's not the kind of sim I was thinking about. That's the case because the rules of your sim precisely replicate real life and that's the only point of it. You're "playing" following a very clear rule book. And they are still "real life", you are not put in a virtual training/learning environment where you write the rule book and can expect mistakes to have 0 consequences.

The question was what if the simulation didn't replicate real life? What if you were expected or requested to do more by being told "it's a simulation and your ONLY goal is to win". It's GTA, Counter Strike, StarCraft, etc. Would knowing the person you just ran over while trying to win is a real person that just died or that you sent a real squad of marines to die so you can save one tank change your decisions? It's a hypothetical exercise, just like the Ender story.

My main point is that even if you're trained to kill in real life the decisions you take in a sim with "fictitious" rules and scenarios can't be considered a breach of morality.

Anyway, there is no justice or real consequences even for mistakes made in real life when it comes to the military. As evidence stands the mountain of innocent civilian victims. And to make the parallel with real life, Ender is not a soldier, he's a general. And just like in real life a decision that costs millions of lives on any side can still be rewarded with medals, glory, fame, as long as you're on the winning side.

It is barely even ambiguous. Bean absolutely knew, and showed it; Ender probably also knew, but intentionally hid the fact that he knew from the adults. The whole battle school was filled to the brim with brilliant, command-trained, engineered psychopaths, as a design feature of the program. They needed someone just empathy-deficient enough to sacrifice most of the human combatants, if it meant annihilating the enemy.

Ender may have suspected that the command would turn on him once the war was won, as they turned on Rackham, so established a public facade of a personality that would struggle morally with winning a total war by any means necessary. He had to be a good guy, rather than a captive, monster-fighting monster, and then keep up the act until he could escape the ensuing post-victory political meltdown on Earth.

I'm not sure if Card even knew at the time he was writing it, or if he left that intentionally ambiguous, to be filled in later if necessary. But as I read it, Ender knew, and was already playing at least one move ahead, past the victory against the aliens.

Empathy deficiency is definitively not what the battle school was looking for. It's the reverse - they wanted someone capable of emphathizing with his enemy while still willing to destroy them with no hesitation. The obvious counterpoint in the story is his brother Peter: he's just as brilliant as Ender, but gets rejected by the academy explicitly for being a sociopath.

Bean knew but we don't know that he knew until Ender's Shadow.

Reply on news.ycombinator.com