Ethics in Strategy Games

Since A.I is rapidly evolving, sooner or later we'll be able to simulate a human brain. If you're a computer gamer, and play any strategy game whatsoever, you know that each unit has a sort-of programmed intelligence. They can only get smarter, and what will it be like when you actally know that you're sending (in starcraft) twenty marines to their deaths at the claws of the nefarious zerg? Will it be allowed? Will people be tried for murder or even war crimes? The possibilities are endless.

 

Log in or Sign up to hide all adverts.

Only if you program consciousness into the AI... which in terms of computer games wouldnt have any practical application. i suppose someone would do it if they could, just to see the results.. but i cant see why it would be done for market release.

You do raise an interesting thought though... I suppose a form of consciousness could be used to simulate the effects of certain types of warfare on actual soldiers. In that case... umm. I suppose it comes down to whether you believe the "soul" (if such exists) is present in the brain as an extension of thought or if it resides elsewhere independant of intelligence.

 

Log in or Sign up to hide all adverts.

That's scary... we'd see so many new psychological problems, new types of marraiges, social structure... a loss of our freedom in gaming or using technology as monitors would be set into place. Plus if they can develop it in computers they can develop robots, which leads to other fears.
With neuro-technology this could be achieved, too.
I think it would divide the population.
Psychologists could practice on virtual patients... but once something has feelings, is it right to just discard it?
Very mind-stirring topic!

I don't think humanity would allow it. In cyberspace perfection is more attainable than real life, and the breakdown of true human interaction would surely happen.

 

Log in or Sign up to hide all adverts.

Technology has always resulted in the 'breakdown of true human interaction'. Onward we march.

 

"Dungeon keeper" did have some method of working out how many of your creatures had been killed and if the gods were happy with your sacrifices.

But all of that and other such games are working on preprogrammed systems, not ones that have "Programmed themselves" like artificial intelligences are suppose to be capable of.

This means that for a game like that to have an intelligence that decides over the moral dilemma of sending troops in to be slaughtered by zerg in starcraft, it would have to be given information to get an understanding of wether it's right or wrong to.

In a sense we know that if that was reality, Casualties are unacceptable, And civilian ones are more so, but since these soldiers are just sprites on a screen with now families, or real lives, they can be sacrificed.

(It of course would be different if those soldiers took 20 years to grow into established fractal macro's for your to fight with as then you would decide that their lives have value.)

The worst case scenario is if the computer has an accident like "Wargames" Namely it thinks it's playing a sim, a computer game or strategy war game, when in reality it's sending thousands of people to die because it thinks and gives them value like they are sprites

 

But as they become more intelligent and grow they become more human, stryder we could be the computer in 'wargames' sending thousands of men into the meatgrinder and not really know it. Even now, some sprites may have extremely, extremely basic feelings that gives them a soul. Feelings just like:attack, defend, hold position etc. The more complex they become the more likely they are to...rebel? That'd be funny, ordering your troops around the map to places where you think the enemy is hiding and not having them comply because they're afraid.
Here's an example: I actually just finished a multiplayer game of Brood War and I was playing against one other guy. He had attacked me several times with some heavy units and wiped out several bases I had established on the map. The map isn't 3-d, but the troops I have could see in 3-d, and so I assume they saw these units destroying my own and putting me in a pesky situation. I ordered them to the areas already destroyed and extinguished the attackers without force but just because my men were cloaked, you can't see them without special equipment. Now, if my men thought the necessary equipment was nearby they probably wouldn't have gone willingly into battle, I might've had to pay them minerals to do so, and they might decide to take over the whole operation. Again, like I said the possibilities are endless.

There is a practical application of human a.i in strategy games: To outwit your enemies! It'd be awsome if they got smarter as they won more battles and developed their own strategies so I could just send them into the area and watch, and sometimes take control if things got out of hand.

Yes, I bet some of them would fight and talk and stuff. If there were male and female gendered-soldiers you might even get a surplus of children spies or something.

 

There are certain points about how the AI would develop the sprites feelings, after all in the real act of war between two civillised countries we stick to rules that were created, e.g. The Geneva convention

This means that, If soldiers surrender they must be taken prisoner and treated fairly.
No soldiers should be tortured.
Injured soldiers should be given medical treatment.

Now if you placed those rules and a few other rules...

Ammo does run out (along with food rations and medical supplies)
When a platoon takes heavy casulaties they might surrender.
Different injuries would reflect on if they can continue fighting or not.

(The last point there is most notible in the game DARK COLONY)

Now your sprites don't run into battle to die aimlessly, they can be taken prisoner, run out of food, or medical supplies, or worse Ammo. They don't shoot unarmed soldiers (which is what occurs when you run out of ammo) they take them prisoner, and if you have a field medic, he will not just patch your soldiers but also the fallen enemy.

Also there is the point that you would be able to do like how a mate of mind use to command his troops to test his weapons... (coughAdamcough) No longer can you run your little infantry man into range of a nasty flame tank, and roast him to see what firepower it has.

How about doing such war atrocities would envoke a tribunal or commision to look into your orders, where you could be tried for war crimes. (Just like real life)

Of course this takes us away from the real use of Artificial intelligences in the role of decision making like "Friend or Foe" targetting or automated responses to rocket attacks etc.

(Of course with recent developments, how would you tach an AI to stop planes being hijacked and anthrax being sent through the post?)

 

yeah!! That sounds kewl!

People would still die though. Computer people.

 

The Stim-Drop and Other Bloody Things

First, props to a haunting question.

As a frequent commander of epic Starcraft battles myself, the thought of those marching Marines having a soul and a will and a fear behind that scarred machine helmet is curious indeed. Take one of my favorite tactics: the Stim-Drop. It seems to be a painful, even ultimate sacrifice on the part of our loyal Marine.

Two dropships (mayhaps the most dangerous employment), eight marines. Find a young and potentially fruitful enemy base - you know the kind - an overwhelming amount of harvesters, a new building and hasty defenses. Sneak your dropships up onto the ridge and behind just so you can see the frenzied spark of a harvesting Protoss. Then push your dropships forward, up, and just into view of the Nexus. Drop those loyal and bloodthirsty marines to the ground. Command them to plunge the needles into already-swollen forearms as some kind of wicked stimulant courses through bulging veins. Perhaps a painful surge of energy as they holler and scream and whip bullets into the machinery around them. Eight splashes of blood later, the harvesters destroyed, a Nexus certainly in blue flame, and enemy base crippled, you return the dropships back to base.

For a soulful Marine trained in preparation for a Stim-Drop, it must surely be a dreaded thing to anticipate. And if you push the "ifs" further, what kind of thoughts slip into the minds of men (as they sit and shift and yawn uncomfortably in the belly of a dropship)who will surely die a most glorious and painful death? What kind of ethics am I employing when I send Marines to die as regularly as a mouse-click?

And then again, haven't you ever been playing a furious battle, online perhaps, and in sending a hapless unit to its doom you find much later it has gotten itself stuck? Perhaps circumambulated over a mountain ridge and sent itself back to base? Was it really just a glitch of 0s and 1s? Or was it a sudden and painful flash of the will to survive, spawned from those "extremely, extremely basic feelings" and quite urgently they find a desperate need to live longer than your mouse-click?

I wonder . . .

 

Hmmmm... most interesting topic, perhaps a burning topic in 3000 A.E (after earth, 200000003000 A.D)

Please Register or Log in to view the hidden image!

What about stradegy games of larger scale for example Master of Orion 2 type of games in which millions of people/aliens live in single planet and if you accindently shoot a bit too much or inaccurately and a few million people/alien die? Or what if you use nerve gases or genetic plagues in Sid Meier´s Alpha Centauri and kill 50000 citizens?

What about something a bit more personal, for example 1st person shootings... what about when a wounded enemy soldier starts to beg for his/her/its life?

Will there actually be time in which even a single unit (out of hundreds) will be consciousnes... in a PC (or some other future machine)?

Also... what if these consciousness units realize that there is a whole other universe out there and that even those who created them are just a WERY small part of that universe? What if they find a way to influence this world... somehow and start for example making robots in this world and transfer theyr consciousness into the machines (I know, we´ll just start to make same kind of robot machines and clone humans to operate them... then we´ll have Total Annihilation {PC game} in real life

Please Register or Log in to view the hidden image!

)?

If so... then we will be GODS... because we will have someone to pick on

Please Register or Log in to view the hidden image!

Perhaps in the future we will see a same kind of struggle that Africans and women had to fight to get some rights. Perhaps we will start to discriminate machines. I

 

So here's the post! I forgot all about it!!

Supir, I'm planning a sciforums starcraft game for next week. Check it out under free thoughts.

Continue!

 

Luck exists, just bad luck!

 

Ughhh. no, because the computer just restarts the ai into a new marine.

An ai that compicated would be an utter bitch to recreate every time you make a marine

So more than likely it has literally 100 different ai programs for your marines, on a digital book case, so when a marine is built, it takes one off the "book case" and puts it into the marine, when the marine dies, the ai goes back to the bookshelf until it gets selected for a new marine.

This way the computer does not have to create an utter bitch of a program for everytime you make a marine, it just recycles old ones.

 

Please Register or Log in to view the hidden image!

NECROMANCY!

You know when you pull an old thread up like this, it's like looking at what was shown in the 1950's at the ideal home exhibition for "The Home of the Future"....

Now where's my robot and why aren't my dishes done?

As for the topic. AI has improved over the years with programmers "Specialising" in just creating the responses. In recent games like Valves Left4Dead, they've actually created a "Director" for the game which randomises what enemies you are going to be up against and where they are going to be placed in the game, so ever game is unique.

In regards to RTS's, the main consumption of CPU power is really down to Pathfinding. Obviously if you program you'd have to have a predefined limit for the resources available for pathfinding a single troop or soldier in any such game, where if it goes above the resource limit, it will take a simple approach to pathfinding a route. The simplest option is to obviously have invisible roads/highways, where the pathfind algorithm just has to find it's way to the predefined paths. Of course this in turn can undermine free roaming action as well as generate predictable routes the AI will take to allow exploitation.

(This point is of course based upon pathfinding using collision detection)

It's already been suggested that Warcraft characters(and other such MMO games characters) can be classed as an 'Asset' and there have even been custody battles over who owns the character. Obviously people are putting value on those sprites now.... too much value and too much time.

 

I have a confession.

When I'm playing "Medieval - Total War 2" I really get uncomfortable marrying off 16 year old princesses to 40 year old men because I want them to pop out future leaders asap.

Particularly when they're spunky little things with attributes like "Brave" and "Intelligent", and can handle weapons. Makes me think they deserve better.
Fancy considering the feelings of unmarried Princesses when marrying them off in a computer game.
I'm not right in the head.

 

Strategy games do not depend on ethics, but depend on rules, just as pure nature. Having the sense of justice, rights and consciousness are human and partly animal faculties. A.I. development will not be one sided; but humans will also put their own brains into the A.I. environment. So certain human rules may probably be applied within those games. But here it is: This may not be even necessary...

There is such an incredible idea that one day, we (our consciousness) will be able to survive through computerized brain simulation. We may still continue to depend on matter and energy, but obviously we will not need any mortal physical bodies and spaces. We can die and come back to "life" endlessly, just as we do in computer games.

Try to connect these ideas and imagine how could it be:

- Apart from some arts and maths, we all depend on organic bodies. Our justice, understanding of life, and every other thing are heavily occupied by this reality. We have to sleep, get sick, eat and drink (not too much, is it too much, ohh!), wash ourselves, be carefull of passing cars and many others...
- If consiousness, ideas, dreams, and other complicated mechanisms of human brain could be simulated, we would use almost any matter as hardware. You can even save your own life to come back later.
-You don't even need to make zombie Terminator-like robots, or to invade other bodies for our software; take the idea one step further: Nanotecnology may provide solutions to store our consciousness within atomic level. And in quantum level time works differently. An average life time of an atom is not less than billions of years.
- We also know that, as any software, simulated brains will need an environment to perform, to experience and situmulate itself. Just as internet, this will be other brains. And these brains do not necessarily have to be 100% humans. You can populate your own planet with as many A.I. as you like. So your own understanding of ethics, justice etc. will be in charge without harming or affecting other souls. If you want to treat your A.I. as gentle creatures, this will only be only the reflection of your own brain. If you do otherwise, that will not make a great difference for other humans.
- Todays' technology can provide devices to translate mind signals into computers and vice versa. If we keep going in this pace in same direction, the planet and its living organisms, as well as everyday humans, will no longer suffer from stupid economic crisis, conflicts, medical problems, disasters, poverty, hunger, violations, and other monkey businesses.
- You will be able to taste, smell, fly, die, born, have sex, live under water and in space,
Because life and reality are going to be an ultimate endless game played in this cosmos and maybe beyond. This type of existence will create a better superficial heaven than any religion or ideology have ever promised. Welcome to the Intermatrix Universe! Knowledge must prevail.

 

That's a funny way of looking at A.I. As long as the A.I. has no flesh and blood, you can treat it however you'd like.