AI Interaction and Psychological Effects
I've been wondering about the mechanics that can be applied to the AI of NPC's we will be interacting with in the game. For the game to work well, there will need to a lot of work done for the AI. Our companions, (and our enemies,) will be coming under a lot of fire, doubtlessly much of this will be supressive fire. So what are the options for these AI should lose their bottle and try cut and run? Should all the friendly AI be under our control? (IE: Dragon Age and Mass Effect.) Should they be totally independant of our control, only following where we go? Or should we be able to give orders, but the AI can choose to ignore them? basically, I am talking about self-preservation. It is classed as one of the fundamental requirements for proof of life, it is that critical. So should the AI try to show some self-preservation instincts?
I for one say yes; and past that, I also think that different NPC's should have different levels of 'courage'. Now I don't want to go into this game and find out that ever member of the crew I enlist will be a die-hard selfless individual that will carry out every single action I instruct them to. That is not realistic and that is not cyberpunk. If that happens, I will be greatly disappointed. I want to see some variation. I want to see everything from the guy who 'thinks' he can handle it but quickly finds out he can't right up to the psychopath who just doesn't know when to stop, (and is just as likely to get you all killed as anything else.) When I recruit companions, I want to find out how they handle themselves under fire. I have been in situations in real life where the shit has hit the fan and in the back of your mind your watching the new guy to see if he's going to stand his ground next to you or if he's going to go hug a toilet. That is real.
Lastly, we will need our friends to know when and how to keep us alive. They need to know when to rush over and dose us up on stims to keep us alive and friendly medics will need to know when to start giving us adrenalin shots and nanite cocktails. But more importantly, these same AI will probably have to be programmed to weigh up the risk to themselves while they try to save our bacon. I don't expect these mental states to remain solid for the entire playthrough. People change; they adapt and they evolve, both emotionally and mentally. Go into enough fights with someone who flat out will not stand and fight and 99% of the time they will grow a spine. Alternatively, When someone has a close call that is just that bit too close, they can break. It takes time for them to get their confidence back and not everyone regains it.
So, the big question is:
"Should AI feel or show fear and if so, how should it manifest itself?"
I've been wondering about the mechanics that can be applied to the AI of NPC's we will be interacting with in the game. For the game to work well, there will need to a lot of work done for the AI. Our companions, (and our enemies,) will be coming under a lot of fire, doubtlessly much of this will be supressive fire. So what are the options for these AI should lose their bottle and try cut and run? Should all the friendly AI be under our control? (IE: Dragon Age and Mass Effect.) Should they be totally independant of our control, only following where we go? Or should we be able to give orders, but the AI can choose to ignore them? basically, I am talking about self-preservation. It is classed as one of the fundamental requirements for proof of life, it is that critical. So should the AI try to show some self-preservation instincts?
I for one say yes; and past that, I also think that different NPC's should have different levels of 'courage'. Now I don't want to go into this game and find out that ever member of the crew I enlist will be a die-hard selfless individual that will carry out every single action I instruct them to. That is not realistic and that is not cyberpunk. If that happens, I will be greatly disappointed. I want to see some variation. I want to see everything from the guy who 'thinks' he can handle it but quickly finds out he can't right up to the psychopath who just doesn't know when to stop, (and is just as likely to get you all killed as anything else.) When I recruit companions, I want to find out how they handle themselves under fire. I have been in situations in real life where the shit has hit the fan and in the back of your mind your watching the new guy to see if he's going to stand his ground next to you or if he's going to go hug a toilet. That is real.
Lastly, we will need our friends to know when and how to keep us alive. They need to know when to rush over and dose us up on stims to keep us alive and friendly medics will need to know when to start giving us adrenalin shots and nanite cocktails. But more importantly, these same AI will probably have to be programmed to weigh up the risk to themselves while they try to save our bacon. I don't expect these mental states to remain solid for the entire playthrough. People change; they adapt and they evolve, both emotionally and mentally. Go into enough fights with someone who flat out will not stand and fight and 99% of the time they will grow a spine. Alternatively, When someone has a close call that is just that bit too close, they can break. It takes time for them to get their confidence back and not everyone regains it.
So, the big question is:
"Should AI feel or show fear and if so, how should it manifest itself?"


