IA Logo

IA Information

Dave Mark's books

IA on AI

Posts Tagged ‘interactive’

We Can’t Think of Everything!

Friday, November 11th, 2011

I admit that, despite it being 11/11/11, I haven’t played Skyrim. I don’t even know if I will be able to get to it for a few weeks. However, that doesn’t stop the barrage of information from people playing it. I am trying to avoid most of the breathy reports from my friends and colleagues around the internet. However, this one kept popping up on my Twitter feed so I figured I would take a look.

The title of this YouTube video is “How to steal in Skyrim.” When I clicked on it, I really didn’t know what to expect. I figured it was going to either be a boring instructional video or a blooper reel. I suppose it goes in both categories, for whatever that’s worth. However, it is an instructional video for game AI developers and designers alike.

What you are seeing is an attempt by Bethesda to mitigate a problem that has plagued RPGs since their inception — that of rampant stealing from houses and shops. Usually, one can do this right in front of people and no one seems to care. One poor solution was to mark objects as being people’s possessions and alert that person when they are stolen. However, that pretty much eliminates the notion that you could steal something when that person is not around…. kind of a “page 1″ lesson in the Book of Stealing, really.

Every person in the game is completely cool with me just placing large objects over their head?

What Bethesda obviously did here is perform a line-of-sight check to the player. If the player grabs something that isn’t normal, react to it. In the case of the lady at the table, she simply queries the player about what he is doing. In the case of the shopkeeper, he reacts badly when the player takes something of his. All of this is well and good. However, when the line of sight is blocked (in this case by putting something over their heads), they can no longer see the player taking something and, therefore, don’t react.

But what about the reaction that should take place when you put a bucket over the person’s head? Are you telling me that every person in the game is completely cool with me just placing large objects over their head? It certainly looks that way!

The lesson here is that we either can’t think of every possible action the player could perform in the game or we simply do not have the resources to deal with it — for example, by having the player protest and remove the ill-designed helmet.

“But why would the player want to do that?”

In the past (and I mean >10 years ago), when the player’s interaction with the world was simpler, many of the faux pas would have stemmed from the former reason. We just didn’t bother to think about the possible actions. The pervasive mentality was simply, “but why would the player want to do that?” Of course, players did do things like that — but given the limited worlds that we existed in, the ramifications weren’t that huge. We were far enough away from the proverbial “uncanny valley” that we simply accepted that the simulation didn’t model that sort of thing and we moved on.

Adding one mechanic to the game could have hundreds of different possible applications.

More recently, as games have allowed the player even more interaction with the world, there is a necessary exponential explosion of possible results for those actions. That is, simply adding one mechanic to the game could have hundreds of different possible applications. When you figure that game mechanics can be combined so as to intersect in the world, the potential space of resulting interactions is mind-numbingly complex. The problem then becomes, how do I account for all of this as a game developer?

One game that began simulating this stuff on an almost too deep level was Dwarf Fortress. I admit going through a DF kick last spring and it almost killed me. (It was like experimenting with powerful drugs, I suppose.) Pretty much everything in that world interacts with everything else in a realistic way. The rulebase for those interactions is spectacular. However, pretty much the only way they can pull it off is because their world is represented iconically rather than in the modern, 3D, photo-realistic, way. For DF, creating a new visual game asset is as simple as scouring the text character library for something they haven’t used yet and making a note of the ASCII code. In Skyrim (and all modern games of its ilk), the process of creating an asset and all its associated animations is slightly more involved. Or so the rumor would have it.

Given the example in the video above, DF (or any other text-based game) could simply respond, “the lady removes the bucket and yells obscenities at you.” Problem solved. In Skyrim, they would specifically have to animate removing things from their head and hope their IK model can handle grasping the bucket no matter where the physics engine has placed it.

What occurred in the video isn’t necessarily a failing of AI.

So there’s the problem. What occurred in the video isn’t necessarily a failing of AI. We AI programmers could rather simply model something like, “your messing with [my body] and I don’t like it.” It just wouldn’t do us a lot of good if we can’t model it in an appropriate way in-game.

This bottleneck can apply to a lot of things. I could represent complex emotional states on finely graduated continua, but until the animation of facial expressions and body language can be modeled quickly and to a finer degree of realism, it doesn’t do anyone any good. No one will ever see that the character is 0.27 annoyed, 0.12 fearful, and 0.63 excited.

In the meantime, rest assured that the hive mind of the gaming public will think of all sorts of ways to screw with our game. Sure, they will find stuff that we haven’t thought of. It’s far more likely, however, that we did think of it and we were simply unable to deal with it given the technology and budget constraints.

And to the gaming public who thinks that this is actually a sign of bad AI programming? Here… I’ve got a bucket for you. It looks like this:


Dating Sims: A New Frontier for RPG AI?

Thursday, May 27th, 2010

It’s amazing what pops up in my daily Google Alerts for “game AI” (although I’m getting tired of Allen Iverson news). This one caught my eye, though. On a blog called “Win My Ex Back” (no thank you, by the way), there was a post titled Online Dating Sim Game. There wasn’t a lot of detail about a specific game other than to mention online that “dating simulation games are among the new genres of online gaming that depicts romance.” To give an idea of what the author (Andy Jill) has in mind, I quote his summary:

It’s a simulation game where the main character that you’ll play (commonly fictional characters) has to achieve specific goals. The most typical one is to date numerous and different women and to have high level of relationship and among them within specified time limit. Generally, the game character must have enough funds by either securing jobs or other income-generating activities such as business.

In the same manner, attributes of the character is important in the game. Such attributes can be improved by doing different task and accomplishing it within the time limit. Most of these tasks take time to be accomplished and games of this type have real-time to them.

The author goes on to describe what apparently was the first online dating Sim, “the Dokyusei or Classmates” from 1992. Again to quote:

In this classic dating sim game, you will be controlling a male avatar that is surrounded by various female game characters. The game play will involve conversations with a variety of artificial intelligence (AI)-controlled girls, in which you will attempt to increase your “internal love meter” by means of right choices of dialogues. The game usually last for a limited game time like a month or a year.

Once the game is finished, your character may lose the game if it failed to win the hearts of any girl. However, you may “finish” one or more girls, usually by having sex with them or by attaining eternal love.

18 years later and we are still relegated to dialog trees.

To me, this sounds like the Sims titles… or for that matter, how some people try to play Mass Effect 2. Anyway, the point is not the gameplay mechanism (which is somewhat standard). Even the personality and mood meter facets are not all that uncommon. Again, think Sims 3. What I find almost joltingly alarming is that this game from 1992 was based on a dialog tree. Sure, that’s fine. We’ve had dialog trees for a while. The problem is… that’s how we would still have to do things! 18 years later and we are still relegated to dialog trees.

The only reason that the Sims doesn’t have dialog trees is that there isn’t real “dialog”. At least not the intelligible sort. Sure, there are behavior selection trees for choosing what to do next, but there is a subtle difference. When you select an action in the Sims, the game designer hasn’t necessarily hand-crafted what the response (or potential responses) will be. In a dialog tree, you are always in a specific place in that tree until you exit it. In the Sims, all that happens when you select an action is that you vaguely change the internal state of the character you are interacting with. That character’s actual response is calculated in a ridiculously expansive set of state values, formulas, and then a stochastic factor tossed in for good measure. You aren’t really ever sure exactly what you are going to get… although you may have a good idea what it might be.

On the user input level, it is still reminiscent of Zork or the early Ultima games.

I have to assume that this dating sim — and all others like it — would rely on a representation of actual dialog, however. And that brings us back to the dialog tree. Natural language processing isn’t going to cut it. Emily Short does a good job of it in her interactive fiction but the root of it all is still keyword-based input. As amazing as her work is, on the user input level, it is still reminiscent of Zork or the early Ultima games. Translated into a dating sim, the user’s free-form input could very well be reduced to “ask job”, “tell age”, and “compliment boobs”. In effect, it wouldn’t be all that different than the chat room shorthand of “a/s/l” for “age/sex/location”. Even if the game then gave elaborate (yet pre-scripted) answers to your questions, it will still be annoying to have it reply “I don’t understand what you mean” when you don’t guess the right keywords to use. Additionally, using that sort of shorthand isn’t going to ever feel really “romantic” is it?

I’m halfway through Noah Wardrip-Fruin’s book, Expressive Processing, where he talks extensively about the history and state of interactive dialog and drama. Even with all the work that has gone into this field for over 40 years, I’m sorry to say that we are no where near being able to replicate something other than a stilted parody of a romantic courtship.

That being said, it doesn’t matter how deep the AI is behind the scenes. Until we can solve the input/output problem our AI is trapped inside itself. Animation has gotten a lot better of late—especially facial and speech animation. I know there are adult-themed games out there and I assume that they are taking full advantage of photo-realism (not to mention realistic physics). However, all that nifty facial animation and subtle gesturing would still have to be tied to canned, pre-written dialog. And that is our bottleneck, isn’t it?

I don’t know how to solve it, really. Noah’s book is my first foray into even thinking about this interactive dialog and fiction stuff. (On the other hand, I would be tickled to do the personality, mood, and emotion modeling. That is in my wheelhouse!) That said, I don’t know how to solve it. I just find it sad that we are still stuck in this world where we can’t really interact with our game characters on a meaningful, natural-feeling level. I do know, however, that when we find it, that will be one of the final cusps we need to cross over in games. At that point, there’s not a lot stopping us.

The Case for Procedural AI

Sunday, March 8th, 2009

Ok… this goes in the “Amen, brother!” category. Kris Erickson at PS3Informer.com wrote a column entitled Why Procedural AI is the Next Big Milestone in Gaming. In it, he smacks on the problem of ostensibly large-scale world with very repetitive content. He sums it up with one question:

How can we create realistic open world games where people that we meet in the street repeat more than the same 3 phrases over and over ad infinitum?

Can that be any more accurate?

At the upcoming AI Summit at the Game Developers Conference, I am on a panel “Characters Welcome: Next Steps Towards Human AI” where I hope to bring up this very notion. My observation is that, until we can solve the natural bottleneck of content creation, in-depth AI is going to be hamstrung. It doesn’t matter that we can create 100′s of subtle behaviors and interactions if our characters only have the voice acting and animations for 20.

With the reasonable success at procedural animation for Spore’s creatures, I feel that we may be able to leverage that for human character animation. Many games are already using varieties of automatic animation creation (which, not being my speciality, is completely beyond me). However, we are definitely up against a wall with regard to voice assets. Until we can do realistic generation of speech, we are going to be hurting for a way to accomplish dialog interaction without pre-written lines for voice actors.

Even if we could pull of natural-sounding speech, automatically generating content is a bit of a quandary as well. If you have time to read 200 pages, I’ve started muddling through my colleague Rob Zubek’s PhD thesis, Hierarchical Parallel Markov Models for Interactive Social Agents (pdf). I am only about a quarter of the way through, but I like where he’s going with it. By applying rational reasoning to interactive speech patterns, we are taking a big step forward in being able to process input speech as well as generate responses. Combine that with natural-sounding speech synthesis and our games will take a massive leap forward.

In the mean time, I believe we have to apply procedural concepts wherever necessary to be able to bypass the content generation pipeline such as it exists now. After all, GTA 4 had a $100 million budget and people still thought that the content was limited. Can we, as an industry, even afford to continue down this route?

The artwork of Chess AI

Saturday, December 8th, 2007

Ever wonder what it may look like to see the computer opponent in chess working on its moves? This site has done just that. By simply playing a game of chess against the computer opponenent, you get to watch as it itterates through the minimax decision tree.

One thing that’s interesting to see is how the flow lines change as the game progresses. I wish there was more detail about exactly what the color and strength of the lines meant. I also wish that you could compare how the flow lines changed from one move to the next so that you could see how you may have completely shut off certain strategies – or made them less valuable. Still, it’s kind of interesting to watch!
There is a gallery page that shows some mid-game shots as well so you don’t have to play through a full game to see how it looks toward the end.
Add to Google Reader or Homepage

Latest blog posts:

IA News

IA on AI


Content ©2002-2010 by Intrinsic Algorithm L.L.C.