New Information from VR on AI

Amathe

Article on AI from the official forums. 

Wizardry

They didn't really give any details..sadly.

Too bad because this is a VERY important aspect of a game,one i look at very closely and can ruin my entire experience if not done very well.

He says he worked on Vanguard,well guess what,i quit VG over the AI,so i really hope he is not going to do that same crap again.

This also happens to most often be a very weak area of most game designs/systems.I usually keep tooting FFXI's horn but unless i see equal or BETTER,i will continue to say most games are doing a crap job with AI and systems related to Ai.

This guy scares me when asked about AI changing depending on the group because i have a very good idea where that comment was leading.I HOPE this doesn't mean lame automated ideas like automatically targeting the weakest link/healer etc etc.Ideas like that would be a STRATEGIC attacking formation idea and NOT a realistic feature of a defending mobs AI and that would if we assumed the mobs had high intelligence.

I also did not like one bit VG...Wow...EQ all of them when comes to VERY simple ,lazy cheap proximity aggro.FFXI absolutely blows these other games out of the water with aggro systems.Another key area i did not like in other games is leashing 2...3 maybe even 6 or more mobs together on the same AI.Yet again in FFXI mobs would react to each other IF already inside the combat,other wise they would not,they acted individually.Once inside the combat together,they might heal each other try to sleep your party etc etc.

So hearing from this guy that there is not many systems in place ,really is sad news,i feel this very important area was left TOO late into the design phase and won't be done very well.

Amathe

Wizardry said:

So hearing from this guy that there is not many systems in place ,really is sad news,i feel this very important area was left TOO late into the design phase and won't be done very well.

To be fair, VR also said:

What challenges are you facing at the moment concerning the AI?

Not really any right now, which is what makes it fun. That part of the game is well ahead of other things and waiting for the rest of the game to catch up. We have zones to build, NPCs to place, art that needs to be rendered into the game, etc. There’s things that need to happen before there’s even any reason for me to add AI to the NPCs. Things are being done, but in general, programming is ahead of the creative team because without programming, the creative team can’t do anything. Programming has to get to a certain point first before the creative stuff can go in. And that creative stuff is a big deal - it’s a lot of work for a lot of people. Meanwhile, I don’t like to sit idly, so I’ve been working on improvements to the chat system or any bug that is a quality of life issue for our testers, because we don’t want them frustrated while they’re testing. There’s a lot of things going on, but not a lot about AI until the rest of things catch up. At that point, I’ll have a better idea of what I need.

Mendel

Amathe said:

Wizardry said:

So hearing from this guy that there is not many systems in place ,really is sad news,i feel this very important area was left TOO late into the design phase and won't be done very well.

To be fair, VR also said:

What challenges are you facing at the moment concerning the AI?

Not really any right now, which is what makes it fun. That part of the game is well ahead of other things and waiting for the rest of the game to catch up. We have zones to build, NPCs to place, art that needs to be rendered into the game, etc. There’s things that need to happen before there’s even any reason for me to add AI to the NPCs. Things are being done, but in general, programming is ahead of the creative team because without programming, the creative team can’t do anything. Programming has to get to a certain point first before the creative stuff can go in. And that creative stuff is a big deal - it’s a lot of work for a lot of people. Meanwhile, I don’t like to sit idly, so I’ve been working on improvements to the chat system or any bug that is a quality of life issue for our testers, because we don’t want them frustrated while they’re testing. There’s a lot of things going on, but not a lot about AI until the rest of things catch up. At that point, I’ll have a better idea of what I need.

That bit actually concerned me a bit.  If he's not got anything to do with regards to AI, then my initial thoughts are that the AI that Pantheon is attempting might not be all that aggressive.  It may be as he's said that the AI work is as far along as it can be at the moment, but that seems contrary to my own, admittedly limited, exposure to AI coding (Master's level CS coursework).  Waiting for the creative team seems a bit like he's pushing testing off until the in-game components are completed.

I hope that this isn't a case where the AI in Pantheon is less ambitious than it could be.

Lokero

Mendel said:

Amathe said:

Wizardry said:

So hearing from this guy that there is not many systems in place ,really is sad news,i feel this very important area was left TOO late into the design phase and won't be done very well.

To be fair, VR also said:

What challenges are you facing at the moment concerning the AI?

Not really any right now, which is what makes it fun. That part of the game is well ahead of other things and waiting for the rest of the game to catch up. We have zones to build, NPCs to place, art that needs to be rendered into the game, etc. There’s things that need to happen before there’s even any reason for me to add AI to the NPCs. Things are being done, but in general, programming is ahead of the creative team because without programming, the creative team can’t do anything. Programming has to get to a certain point first before the creative stuff can go in. And that creative stuff is a big deal - it’s a lot of work for a lot of people. Meanwhile, I don’t like to sit idly, so I’ve been working on improvements to the chat system or any bug that is a quality of life issue for our testers, because we don’t want them frustrated while they’re testing. There’s a lot of things going on, but not a lot about AI until the rest of things catch up. At that point, I’ll have a better idea of what I need.

That bit actually concerned me a bit.  If he's not got anything to do with regards to AI, then my initial thoughts are that the AI that Pantheon is attempting might not be all that aggressive.  It may be as he's said that the AI work is as far along as it can be at the moment, but that seems contrary to my own, admittedly limited, exposure to AI coding (Master's level CS coursework).  Waiting for the creative team seems a bit like he's pushing testing off until the in-game components are completed.

I hope that this isn't a case where the AI in Pantheon is less ambitious than it could be.

Well, I don't think his comment really means anything negative about the AI.  It'd be pretty hard to test out a lot of things if there's no genuine environment in place to test them.  You can't exactly go in and make the AI more "ambitious" without being able to test regularly and properly -- every step of the way.

I find that whole thing rather concerning, too.  But, for me, it's saying a lot of negative things about the rest of the development, and nothing about the AI.

It sounds very much like general development has gotten almost nowhere if there are so few assets available to him for testing code.
It makes wonder if they are familiar with the concept of placeholders  

Kajidourden

If you're expecting ground-breaking AI from this project then you were setting yourself up for disappointment from the beginning.

I think we all know exactly what we can expect from it.

Dullahan

Pantheon's success is about as contingent on groundbreaking AI as it is unicorn milk.

Amathe

I read the article as positive news. That is why I shared it. 

delete5230

My honest opinion.  

VR is ignoring everything and everyone in development, they're doing their own thing. 

This is good in story, game play and longevity.  This is everything I want.  

The end results I'm expecting are: 
- Nice graphics and draw distance 
- Poor animation  
- Poor coding  

I'm OK with that

Wizardry

Well Delete,i would say those few small areas of concern are likely going to be accurate but not nearly enough of a target design area to concern me.
I look at the ENTIRE package because if i am going to play for longevity which a mmorpg SHOULD be about,i wil end up crossing every single aspect of the game.

I am not one of these modern WOW era of gamer that ignores the entire game to race to end levels then just spend 24/7 hitting dungeon finder.I actually play the ENTIRE game,al of it and i want it al to be done very well.

I do NOT need great graphics,however if you are going low end np just don't make it hard on the eyes.Example there was something about Wildstar that gave me headaches ,lighting done badly idk exactly what it was.

One area that MOST of this new era gamer likely sees as trivial is huge for me.Aggro systems,i really despise trying to get past some mob and if i hit that 1 pixel area,he auto turns and attacks without ANY realistic hesitation.Mobs should act like a real mob,hesitate and think for a second before determining to attack me,it should not look automated.

Long story short,i rather have NO system if it is done poorly,using this example,i rather have ZERO aggro than the way it is being done in games.LAZY systems bother me,i want a game to play and look like a professional made it.

Kyleran

delete5230 said:

My honest opinion.  

VR is ignoring everything and everyone in development, they're doing their own thing. 

This is good in story, game play and longevity.  This is everything I want.  

The end results I'm expecting are: 
- Nice graphics and draw distance 
- Poor animation  
- Poor coding  

I'm OK with that

Let's hope they avoid poor coding, otherwise it will be quickly labeled as another Vanguard and many will avoid playing.