Hi there,
I'm one of the guys working on Storybricks and wanted to address some of the comments plus share some info.
Storybricks started because AI in MMOs suck, and because all the money in AI was going into FPSs, where the lifespan of an NPC is about 5 seconds.
Also part of the problem is that AI middleware is built with single player console games in mind, which means that the game is supposed to last 20-40 hours so you can script a lot of stuff to give a consistent game experience. MMOs are persistent worlds that last year, and the interactions need to be meaningful but there are currently no tools to do that.
There is a reason why you have a lot of shooters, driving and city sims. It's because it's easy to simulate a spatial representation. When you start looking at emotions, no one has cracked how to incorporate them as a mechanic in a game without suffering from combinatorial explosion or generally not being fun in the first place.
Was a huge fan of games like Planescape: Torment. The problem was that there was no technology, just good design and elbow grease and that does not scale. Good design you cannot replicate but at least you can have tools that focuses on mapping relationships and having information as objects that affect behaviors.
This is one of the reasons Storybricks was born.
Now, to address some of your specific comments
J1M
- The tool creates some basic dialogue, but we do not want to wander into the procedurally generated dialogue beyond basic since it's an AI complete problem and cannot be solved in a satisfactory manner.
- Debugging _is_ a problem, we are trying to automate as much as possible
- This is the first playable, you normally see games at a much developed stage. Elder Scrolls MMO has been in development for 5 years, we've been going for 9 months. I hate the secrecy that permeates the games industry and have no shame in showing what we are working on early on or allowing people to play with broken unfinished builds. ZeniMax might care about your feedback, but reality is that after 5 years of development they have way beyond the design phase so nothing you say will be implemented.
Davaris
- we're using Unity
- the language is Scratch from MIT
- the reason why we are not doing a plugin is that it has been tried before and companies failed because there was no game that could show what you can do with AI, also a tool to edit AI behaviors is 5% of the problem, 95% is the actual behaviors themselves. We believe the market of people that want to build easily a game is an order of magnitude bigger than the Unity user base.
- Chris Crawford approached the problem from an interactive fiction angle, which is an AI complete problem and doomed to fail. We know Chris and respect his work, but I am trying to build something that works.
- Some parts are easy to replicate, some need an insane amount of money. With AI half of the problem is to reach a decision (easy) and the other half is how to convey that decision to the player (expensive). We have decided to use animation to do that, hence why Don Bluth is on board.
- there is a lot of stuff about Storybricks that is not yet in the release that is hard to develop, such as understanding context for actions, an NPC sees two PCs performing an action, understands his relationship with them, their action, infers their relationship, and changes automatically its behavior. that is very hard to build without scripting it.
- we're looking at adding a way to exporting in XML (easy) and importing (hard) Storybricks so you can share your stories with others
@Yaar
- the use of FSMs should be deprecated and the industry should move to behavior trees, the fact that middleware still uses FSMs is part of the problem that we would like to solve. AI in games is moving at a glacial pace (Halo 3 had this new blackboard architecture that was invented in the '70s!!!)
- we're looking at a way to curate games, as for a 3rd party, we are in early discussions to license our technology
Doc Savage
- we're not a wizard
Regardless of the content of your comment, I really appreciate that you guys took the time to express an opinion about Storybricks. While we have some famous backers, it's by no means guaranteed that we will succeed in this KS campaign as the target is insanely high.
Any help in spreading the word is appreciated. Anything you would like to see, please say so.
Would you guys be interested in playing with an alpha?
-Mughal