9:16am - Well it's a new day and I still feel worky, so I think I'm going to try and polish off the last of that book. I can use it for reference once I've at least skimme the last bits and figured out what use I can get from it.
10:04am - And that's the rest of the book down. Good book. A little old, but still relevant.
Now that I've got it scoped out and read the interesting parts, I can use it as a reference while I build my AI Experiments.
I'm feeling a bit off kilter now though so I'm not sure I'm ready to outline what I want to actually make next. I'm going to take a litle break.
11:21am - Alright so I've taken a bit of a break, and now I'll try to sort out what I want to do next.
There's really a couple different things I'm thinking about. The first involves the standard Movement AI. I found that great little Movement AI Library for Unity the other day, but I've been thinking more and more that it would be most appropriate to use Unity's builtin NavMesh solution for my next project, and so I may need to make my own Movement AI Library based on that NavMesh after all.
The other thing is much bigger. I want to make a prototype of the AI type for the game I want to build next. See, I want to make a game where the player is a member and perhaps leader of a community of very detailed AI townsfolk. Specifically a colonization themed game where the player builds their community and with the help of the townsfolk, gathers and erects all of its resources.
I'm not sure yet if I want this to be multiplayer, but probably not. It's just one person here and I'd like to finish the game in my lifetime.
As for the townsfolk, I want each of them to make their own high-level plans based on their goals and interests. They'll generally want to work and contribute to the community, and I'm thinking I won't shoehorn them into jobs. The player will do that, when building the community. They might have preferences, but everyone should be able to do any unskilled task, and maybe learn skills? That's getting ahead of myself though.
A single townsfolk will have life goals like "Become Wealthy" or "Marry" or "Master a Trade" and interests like "Fish" and "Flowers" that dictate the hobbies and distractions they'll get up to. I'm thinking they'll schedule each week in advance with general plans for what they'll do to achieve their goals and interests and civic duties that the player assigns them, and they'll narrow it down to plans for each hour of the day, which dictates where they'll try to be and what they'll try to be doing. Then during that hour things might happen to interrupt them or change their longer term plans, like maybe monsters attack the colony or something. I don't know yet if I'm going to have monsters but that's an example of an emergency that could happen.
So they have a week schedule, a day schedule, an hour schedule, and they react to stimulus as they go about their scheduled activities. But the most important thing is that they make up this schedule largely themselves. The player might schedule some activities but it's the Townsfolk's choice to do them depending on their goals and interests.
Relationships will also be important. The townsfolk will need to remember what's happened, at least somewhat. They'll probably have both short term and long term memories that dictate how they respond to other characters and the player.
Communication is the most difficult part of this though. How should the townsfolk communicate with eachother and the player? That'll be my biggest hurdle. So much so that I'll probably start with a prototype that's just the townsfolk and no players, and work up from that.
Natural language processing is kind of beyond what I can accomplish, I think, so I'll need to figure out some way to let the player tell people how they feel or want and share other ideas. It may be as simple as picking from a dropdown of options or maybe something like an unnatural language processing system like the old Ultima games used. Name, Job, Health? Heh. We'll see.
For now though, now that I've got those two ideas out there, I'm going to have to put my work on hold. Rob's up and he wants to watch One Piece. <3
10:04am - And that's the rest of the book down. Good book. A little old, but still relevant.
Now that I've got it scoped out and read the interesting parts, I can use it as a reference while I build my AI Experiments.
I'm feeling a bit off kilter now though so I'm not sure I'm ready to outline what I want to actually make next. I'm going to take a litle break.
11:21am - Alright so I've taken a bit of a break, and now I'll try to sort out what I want to do next.
There's really a couple different things I'm thinking about. The first involves the standard Movement AI. I found that great little Movement AI Library for Unity the other day, but I've been thinking more and more that it would be most appropriate to use Unity's builtin NavMesh solution for my next project, and so I may need to make my own Movement AI Library based on that NavMesh after all.
The other thing is much bigger. I want to make a prototype of the AI type for the game I want to build next. See, I want to make a game where the player is a member and perhaps leader of a community of very detailed AI townsfolk. Specifically a colonization themed game where the player builds their community and with the help of the townsfolk, gathers and erects all of its resources.
I'm not sure yet if I want this to be multiplayer, but probably not. It's just one person here and I'd like to finish the game in my lifetime.
As for the townsfolk, I want each of them to make their own high-level plans based on their goals and interests. They'll generally want to work and contribute to the community, and I'm thinking I won't shoehorn them into jobs. The player will do that, when building the community. They might have preferences, but everyone should be able to do any unskilled task, and maybe learn skills? That's getting ahead of myself though.
A single townsfolk will have life goals like "Become Wealthy" or "Marry" or "Master a Trade" and interests like "Fish" and "Flowers" that dictate the hobbies and distractions they'll get up to. I'm thinking they'll schedule each week in advance with general plans for what they'll do to achieve their goals and interests and civic duties that the player assigns them, and they'll narrow it down to plans for each hour of the day, which dictates where they'll try to be and what they'll try to be doing. Then during that hour things might happen to interrupt them or change their longer term plans, like maybe monsters attack the colony or something. I don't know yet if I'm going to have monsters but that's an example of an emergency that could happen.
So they have a week schedule, a day schedule, an hour schedule, and they react to stimulus as they go about their scheduled activities. But the most important thing is that they make up this schedule largely themselves. The player might schedule some activities but it's the Townsfolk's choice to do them depending on their goals and interests.
Relationships will also be important. The townsfolk will need to remember what's happened, at least somewhat. They'll probably have both short term and long term memories that dictate how they respond to other characters and the player.
Communication is the most difficult part of this though. How should the townsfolk communicate with eachother and the player? That'll be my biggest hurdle. So much so that I'll probably start with a prototype that's just the townsfolk and no players, and work up from that.
Natural language processing is kind of beyond what I can accomplish, I think, so I'll need to figure out some way to let the player tell people how they feel or want and share other ideas. It may be as simple as picking from a dropdown of options or maybe something like an unnatural language processing system like the old Ultima games used. Name, Job, Health? Heh. We'll see.
For now though, now that I've got those two ideas out there, I'm going to have to put my work on hold. Rob's up and he wants to watch One Piece. <3