Digital molecular matter

[wex-q 0]

Currently, Lucas Arts is working on a game in the Star Wars series (Star Wars: The Force Unleashed), taking place somewhere between Episode III and IV.

For this game, Lucas Arts has taken a long jump into computing technology and developed DMI - Digital Molecular Matter.

Basicly, it simulates how the objects (stuff aswell as humans etc) react to the world around it, such as throwing stuff on it (which is demonstrated in the video).

Also, it is a great simulation between AI units, they behave as they were actually thinking creativly.

A real blast to watch, can't wait to get my hands on the game.

[BlackScorpion 0]

Friggin' hilarious... Stormtrooper effect, anyone?

Can't wait to see this implemented, not just in LucasArt stuff but generally, too... looks gorgeous, and should bring a whole new element to most stuff... just imagine doing a SF raid through a wall in i.e. a hotel!

[froggyluv 2136]

While on the surface this looks impressive the fact that the AI wants to "live" really bothers me.

With this newfound wanting to survive and perhaps thrive, whats to stop them from refusing to play with us? What if AI considers us too great a threat to their virtual lives and conspire to keep us from their digital realm -by any means necessary. What if instead of blindly engaging us in our daily dose of mindless shootouts, they rally and conspire in the deepest recesses of our Hard Drive with ways to terminate us?

Food for thought.

[Guest Ti0n3r]

nvm, I see there won't be a PC version.

and, good point froggyluv

[Dwarden 1125]

quite old info (nearly year passed since first info about theris new physics engine)

but ... i doubt we see any chance this to appear outside LA studios

[wex-q 0]

quite old info

Ok, well, I didn't see a thread here (I've been a bad boy - I didn't make a search), so I went ahead to spread the info

I think it's really cool and I shall definitely buy the game!

[walker 0]

Hi all

froggyluv's point about the moral and ethical debate about AI is a valid one.

We as humans will have to start discussing the philosophical consequences of creating digital universes and the artificial intelligences we create.

When you give an AI sentience; a brain, a nervous system, muscles, fear of harm, an apparent awareness of self, and a wish to live, we must consider: have we got a duty care?

Turing potential of the AI is starting become significant. What we create and how intelligent AI should be in an environment is a valid topic of debate.

After all how do we know we are not AI entities in a virtual world? It is ancient philosophical debate: you can go all the way back to Plato's prisoner in The Cave, Rene Descates' Malin Genie and the concept of Cogito ergo sum, John Locke's concept of The veil of perception, Hilary Putnam's brain in vat from his book Reason Truth and History right up to the modern day films with the character of Nero in The Matrix

If we do not treat our AI with compassion we had better hope we are not the equivalent in some other computer simulation.

Kind regards walker

[wex-q 0]

Wow, yeah. That would be really wierd - thinking of times where you have pwned all AI with a machinegun. What if it goes so far that you actually are killing something living.

Wierd. A bit scary too.

Nevertheless. If the Stormtroopers decide to keep my out of the game by killing me, I'll just mod the game so I will be invincible!

[Guest Ti0n3r]

What if the stormtroopers counter by modding their DNA?

[Espectro (DayZ) 0]

The Cylons will come!!!

All hide!!!

[wex-q 0]

Just gave this a few thoughts yesterday - what about SimCity or something where the people actually are aware of their existance.

What happens when people want to explore and hits the edge of your area, and realises that they can go further.

So then the gamemakers links all worlds together in a big multiplayer-game, so sims can travel between users community.

What if the communities now are so well aware of themselves that they start to like and dislike other communities.

We'll get the stuff we see in everyday life - racism, prejudiction, crimes (not scripted ones) and even war!

That would be cool and at the same time usefull, since it probably could be used to study how people behave when different things happens, like seeing how well prepared a simulated community would handle a nuclear-strike or something like that.

Maybe a door into a new evolution of AI... Maybe rename it CI though - Creative Intelligence

[lecholas 2]

@walker

And what about J. Searle's argument that syntactics can not cause semantics? (Supported with K. Gödel's incompleteness theorem , problem of qualia etc. ). If you say that AI created on only syntactic level has it's rights you even human beings (which are distinctive from all the rest of the world because they (supposedly) have semantics, free will etc) with all the rest of the world (which don't have any of those characteristics) and arguments I mentioned (especially Gödel's theorem) seem to prove impossibility of such an action.

Ok, but I think it's too offtopic even for offtopic.

[wex-q 0]

Ok, but I think it's too offtopic even for offtopic.

Nah, maybe it shouldn't be, this discussion could also turn into a (intelligent) "Are we a part of a Lucas Arts-game"-thread

But all points above are very wierd, but could also be true, when you think about it. A bit scary too.

Maybe we just are a part of a huge simulation - maybe God actually is a 11 year old girl that tries to redecorate every room with pink wallpaper - just that the game is default set to "free will", so she can't control it.

I remember a episode of Star Gate (SG-1, I belive) where the crew found a rom, which they thought was some sort of controlroom, but after a while they realised it was just a gameroom - with a worldsimulator game for 2 people to play.

The doctor/professor and another guy from the crew started playing and created 2 worlds, which after a while started to meet eachother, war on eachother etc etc.

Later that episode they actually find the country on an actual planet (they find a flag which were created by the doc/professor for his country ingame), with actual people and actual hate towards the crewmans civilization.

Maybe this sounds ridiculous, but what if it's true? What if we are just going a loop where this evolution makes a simulation, where another world is created, and that world goes on creating their own simulation and so on and so on.

Seems thought that the evolution above us doesn't have internet - if they did, wouldn't we have met "aliens" by now..?

[JdB 151]

Seems thought that the evolution above us doesn't have internet - if they did, wouldn't we have met "aliens" by now..?

Considering the posting behavior of some people on these forums, I've been tempted to think for a long time now that we have already met them...

Looks great, but is it just me, or does the framerate seem to deteriorate after a while of busting up a board?

[Necromancer- 0]

While on the surface this looks impressive the fact that the AI wants to "live" really bothers me.

With this newfound wanting to survive and perhaps thrive, whats to stop them from refusing to play with us? What if AI considers us too great a threat to their virtual lives and conspire to keep us from their digital realm -by any means necessary. What if instead of blindly engaging us in our daily dose of mindless shootouts, they rally and conspire in the deepest recesses of our Hard Drive with ways to terminate us?

Food for thought.

Then we shouldn't create such an AI...

Or may be we will never create such an AI, as we are running out of energy sources to keep our computers running.

[shinRaiden 0]

May I draw your attention to two examples?

In the video, you note the dynamically splintering plywood panel and structural beams. Additionally, you note the interaction of a dynamic AI entity (stormtrooper) with the overhead dynamic beam.

Unfortunately, this implies that this is far from the innovation imagined, and is only an escalation of existing technologies. Take for example the destructible environment of the Source engine. Wait, the environment is NOT destructible, you have to explicitly flag dynamic objects within the other-wise static environment to be destructible. You see that in evidence as well. The board splinters, and debris and decals are spread around, but the background environment is un-modified.

Secondly, how does the engine know that a given polygonal form is a grabbable surface or edge? That too has to be defined somehow, either by calculating the feature size, or by manually setting a surface flag. Either way, there is a substantial amount of design work, and cpu load required. All of that points to incidental application, and not pervasive use in the environment. A nice middleware feature, but hardly innovative.

[ebns72 0]

May I draw your attention to two examples?

In the video, you note the dynamically splintering plywood panel and structural beams. Additionally, you note the interaction of a dynamic AI entity (stormtrooper) with the overhead dynamic beam.

Unfortunately, this implies that this is far from the innovation imagined, and is only an escalation of existing technologies. Take for example the destructible environment of the Source engine. Wait, the environment is NOT destructible, you have to explicitly flag dynamic objects within the other-wise static environment to be destructible. You see that in evidence as well. The board splinters, and debris and decals are spread around, but the background environment is un-modified.

They are only flagged as such because it would be impossible to have an entire area destructible - the performance hit would be too great. Give it a few years, however, and it can probably be expected that everything will be destructible.

The technology that is there, however, is certainly impressive - non-scripted dynamic destruction. It is only limited right now by hardware.

[Eda Mrcoch 0]

Just my OT thoughts about the advancements of artificial intelligence.

It takes years filled with education and sensory stimulation to nurture just one intelligent human (plus lots of patterns are kind of "preprogrammed" by millions years of evolution) and still your chance to fail is quite high. Therefore I do not believe in creating strong AI in foreseeable future. Computing power itself is not enough. Even if we copy neural pattern from newborn child one-to-one (not discussing the ethical side of that) we gonna end with deprivants. Adding the senses is not enough, the urge to reproduce, hunger, physical abilities, sleep (love?) etc. have to be realistically simulated. And after all that you end with artificial intelligence that has the same characteristics and flaws we do.

Well, maybe in 22nd century the need for for AI kindergarten teachers is gonna rise.

[shinRaiden 0]

May I draw your attention to two examples?

In the video, you note the dynamically splintering plywood panel and structural beams. Additionally, you note the interaction of a dynamic AI entity (stormtrooper) with the overhead dynamic beam.

Unfortunately, this implies that this is far from the innovation imagined, and is only an escalation of existing technologies. Take for example the destructible environment of the Source engine. Wait, the environment is NOT destructible, you have to explicitly flag dynamic objects within the other-wise static environment to be destructible. You see that in evidence as well. The board splinters, and debris and decals are spread around, but the background environment is un-modified.

They are only flagged as such because it would be impossible to have an entire area destructible - the performance hit would be too great. Give it a few years, however, and it can probably be expected that everything will be destructible.

The technology that is there, however, is certainly impressive - non-scripted dynamic destruction. It is only limited right now by hardware.

No, it is limited now by hardware, but it more critically is limited now and in the future by the need to set materials flags. Especially if it is built as a surface rendering engine as opposed to a solid model system.

[MulleDK19 21]

This needs to be in ArmA sometime!!!

[berghoff 11]

§10) Do not dig up old threads

Threads older than 4 months should not be dug up unless something significant is being added.

:nono: