You are linking articles and referencing what company spoke people have said, but these people usually have very little knowledge of the technical side of game dev and they simplify things a lot. the people that do deal with the technical side of game dev are the people you are currently arguing with, it's really like EQ says, you are a customer arguing with a veteran baker on how to bake bread. Sometimes you just kinda have to take peoples word for it, I know I would and I probably have more experience then you.
Here is some cold, hard, honest advice for you: If you ever want to work in the industry, take a good look at yourself and drop the delusions of grandeur. You're not an expert. Industry vets see your posts as a running gag, a joke. People pass links around to your comments and threads for a laugh. Do you want to have a career some day, or would you rather be the clown people snigger about behind your back? I'm not telling you this for a sick burn, I'm telling you because in all likelihood nobody has had the decency to let you know. The industry is very small, and you're cultivating a very clear impression in the minds of many who just might be in a position to help or hinder your success/career.
My life can be summed as an encyclopedia. I try to have my worldviews challenged constantly, because I've had far worse experiences living a life of ignorance then constantly being told why I'm wrong and how I can learn from it.
I actually worry about the day when I can't be told why I'm wrong. It's for the exact reason why artists are always told to post their work and have people critique it. Without new ideas, there is no knowledge, and from there, there is no incentive to ever work hard.
I do want to work harder. I've had people from both inside and out of the industry tell me they were impressed by this. But this could never happen if I wasn't afraid to challenge ideas and status quos.
I actually want to see more people like this one day. They're not afraid to be laughed at, because they're in a cycle of constant improvement. Being wrong today is a better sacrifice if it means being a far more capable and intelligent person tomorrow.
Here is some cold, hard, honest advice for you: If you ever want to work in the industry, take a good look at yourself and drop the delusions of grandeur. You're not an expert. Industry vets see your posts as a running gag, a joke. People pass links around to your comments and threads for a laugh. Do you want to have a career some day, or would you rather be the clown people snigger about behind your back? I'm not telling you this for a sick burn, I'm telling you because in all likelihood nobody has had the decency to let you know. The industry is very small, and you're cultivating a very clear impression in the minds of many who just might be in a position to help or hinder your success/career.
My life can be summed as an encyclopedia. I try to have my worldviews challenged constantly, because I've had far worse experiences living a life of ignorance then constantly being told why I'm wrong and how I can learn from it.
I actually worry about the day when I can't be told why I'm wrong. It's for the exact reason why artists are always told to post their work and have people critique it. Without new ideas, there is no knowledge, and from there, there is no incentive to ever work hard.
I do want to work harder. I've had people from both inside and out of the industry tell me they were impressed by this. But this could never happen if I wasn't afraid to challenge ideas and status quos.
I actually want to see more people like this one day. They're not afraid to be laughed at, because they're in a cycle of constant improvement. Being wrong today is a better sacrifice if it means being a far more capable and intelligent person tomorrow.
Trying, failing and learning from your mistakes is a great practice if you're looking to improve. You need to actually put in the work though, not just talk about it. What you're doing in these threads is decidedly not failing and learning, it's just wasting the time of everyone involved.
If you want to get better, smarter, more experienced, you have to put in the work. Challenging the status quo by a means of incessant arguing with professions/experts who know infinitely more than you is not productive, it's just disrespectful and annoying. Show us with your actions that you're worth the effort. Less talk, more art, words are cheap as hell.
I just want to chime in on this specific point here, so that other people aren't misled by it. CPU performance is very relevant in all types of games, and being CPU bound in terms of general/total performance is common.
Trying, failing and learning from your mistakes is a great practice if you're looking to improve. You need to actually put in the work though, not just talk about it. What you're doing in these threads is decidedly not failing and learning, it's just wasting the time of everyone involved.
If you want to get better, smarter, more experienced, you have to put in the work. Challenging the status quo by a means of incessant arguing with professions/experts who know infinitely more than you is not productive, it's just disrespectful and annoying. Show us with your actions that you're worth the effort. Less talk, more art, words are cheap as hell.
I've been meaning to post a new art thread really soon. It's actually the opposite where I have far more art now, but I'm censoring parts of the story that might offend people.
For example, you mention the challenging status quo by arguing with professionals is annoying, but I've made cartoon characters who do exactly that. I've always wanted to make art that defies everything, but I also don't want to risk upsetting people who might not get the humor I'm going after.
Trying, failing and learning from your mistakes is a great practice if you're looking to improve. You need to actually put in the work though, not just talk about it. What you're doing in these threads is decidedly not failing and learning, it's just wasting the time of everyone involved.
If you want to get better, smarter, more experienced, you have to put in the work. Challenging the status quo by a means of incessant arguing with professions/experts who know infinitely more than you is not productive, it's just disrespectful and annoying. Show us with your actions that you're worth the effort. Less talk, more art, words are cheap as hell.
I've been meaning to post a new art thread really soon. It's actually the opposite where I have far more art now, but I'm censoring parts of the story that might offend people.
For example, you mention the challenging status quo by arguing with professionals is annoying, but I've made cartoon characters who do exactly that. I've always wanted to make art that defies everything, but I also don't want to risk upsetting people who might not get the humor I'm going after.
man, the ammount of bullshit and delusion you have will screw you up over time like Eq said.
I feel sorry for you since you genuinely dont seem to understand what a pompous ass that has nothing to show you look like. Your whole post history is 90% garbage. Stop polluting these boards with verbal diarreah and actually do some art. Otherwise keep beeing laughed at. People were more than honest and helpfull to you and you keep beeing a twat. Sorry im hungover.
I mean, another reason you probably won't get the jump to insane performance every time we switch console gens is they tend to be running on hardware that had to be locked in a year or so in advance before production. Not only that, but Sony and MS routinely sell their consoles at a loss, I remember reading that sony was losing about $50-60 per ps4 sold when it first came out, no idea if they are breaking even now. They can make that loss up with profits from games and online services etc, but still thats gotta be painful to justify for the first couple years.
What you are asking is for them to put in a shit ton of expensive components, and take an even bigger loss/gamble. I don't see madden and gta players (aka the mainstream) shelling out 1k for a console because they are gonna be getting technical rendering features they have no understanding of, or that make the game look 10% better to the untrained eye. Obviously 4k is where the focus is going and I think we are at the beginning of a hockey stick shaped graph in terms of adoption, so for sure that is going to be taken into account.
Games is first and foremost a business, and those console manufacturers have hundreds of people crunching the numbers in terms of production and marketing costs and how they can somehow hope to break even on a machine in a 5-10 year window. They know what they are doing. There is a big difference between what enthusiast gamers can hope and clamor for, and what a company can actually present to shareholders as a financially viable option. You could also probably thank the crypto market for less of a jump in a GPU next gen due to insane spikes in cost and demand.
I don't know if its just me, but personally I think the most visually impressive games of the last 4-5 years have been on these "underpowered" consoles. I remember having my jaw drop when I first played Ryse, Uncharted 4 is pretty much one of the most gorgeous games ever, and the new God of War looks like it's gonna blow my dick off on all fronts. I haven't seen much on PC these days that is hitting those visual targets even with all that extra power. Kinda proof that the technical side of things will only get you so far, the major gains are from insanely talented art teams nailing the execution based on years of experience squeezing every drop of perf outta those machines. And in todays market, those art teams are usually deployed on console projects, because that market tends to dwarf PC sales.
I mean, another reason you probably won't get the jump to insane performance every time we switch console gens is they tend to be running on hardware that had to be locked in a year or so in advance before production. Not only that, but Sony and MS routinely sell their consoles at a loss, I remember reading that sony was losing about $50-60 per ps4 sold when it first came out, no idea if they are breaking even now. They can make that loss up with profits from games and online services etc, but still thats gotta be painful to justify for the first couple years.
Really insightful post!
That's why you get a console generation like PS3/XB360. XB360 was out from 2005 until it was superseded in 2013 by the Xbox one. But those were incredibly informative years to not change up hardware. The difference in the hardware between a modern system and one 8 years behind is immense. They're more about squeezing everything out of them over time to make up the short term losses (which obviously was your point). Which I think in their own way helps push rendering tech a lot by doing more with less, which in turn helps them cram more and more under the hood when newer more graphically intensive rendering techs start to become the norm.
I'm really happy to see the push from GDC this year to see the stepping stones of real-time ray tracing. And even if that Star Wars presentation was done on a $60k US machine, it was still unbelievably beautiful, my wife couldn't believe it wasn't just film until one of the animations didn't look entirely lifelike. In 2-3 years hopefully we'll see enthusiast chips that can do it for significantly less.
My main worry is that I feel with real-time raytracing, it's going to take a long time for major developers to switch to it since it seems to be something that only the newest tech can take advantage of, whereas developers need to make stuff for the largest pool of players. I feel like we'll need a new Crysis type game, essentially a tech demo, to really push the boundaries and get people excited to upgrade their hardware. But in the short term I imagine it will be a big gamble for one of the bigger devs to push to using something like that for a project that could be 3-5 years down the road, which is what I feel they'll need to do to get out ahead of it.
Either way though, the future looks really bright.
For games I think real-time raytracing is still years off. It's more interesting for other fields, like look-dev, arch/product vis, draft rendering for film, it has a lot of potential.
I think seeing any real-time high fidelity stuff in the short term is going to be really cool. But I could see enthusiast hardware I think in 2-3 years at the ~$1000 mark. If you look back 5-ish years ago, which isn't an unheard of amount of time in game development, PBR was only just becoming a thing. Just a year or 2 later and it seems like it was in almost everything. Certainly a lot of companies scrambled last minute to change their engines to PBR or PBR-like. I know one of the projects I worked on, looking back at early bugs in the systems I could tell it was all done pre-PBR and then they changed over a lot of the rendering tech to it as industry trends began to change.
Plus, like the PBR shift, when people see it done right the first time (good 60fps performance) in a consumer game everyone else will scramble to do it too, because everything else just doesn't look as good anymore. I guess it's just a matter of when that will happen at this point, but certainly those planning for long term projects must have it somewhere in the back of their mind that trends will change and what is at the bleeding edge now may be very manageable when their project is released.
My reasoning for saying that it might be somewhat locked into just the new hardware is because I believe (and correct me if I'm wrong), on the Star Wars demo the main reason they were using the new Nvidia Volta chipset, which are made for AI/machine learning applications, is for the tensor cores alongside the Cuda cores. The tensor cores I think are the parts doing most of the machine learning stuff which is used to clean up the ray-tracing noise, and currently only the Volta series have those tensor cores. Maybe they'll find new ways to get it to work on existing graphics cards (which I think Otoy was doing with their Unity ray tracing engine(?) unless they also rely on newer cards as well). That was my thoughts on what could prevent it from being adopted right away by the big players in the industry, as soon as enthusiast cards are affordable enough to run it.
I'm just hoping someone will be at the cutting edge rather than waiting and playing it safe. But it could be like the VR rollout. Lots of small arcade games to start to show that the tech works, and slowly we see more full-fledged titles.
I'll also write the disclaimer that I only know at the most basic level of what's going on in graphics programming and the hardware that drives it (I'm an artist, and at the end of the day I just want to make pretty pictures). So if anything I've said is completely misinformed let me know, because I genuinely like to be better informed about this stuff, because if you aren't staying on it I feel you can get left behind really quick in this industry.
For games I think real-time raytracing is still years off. It's more interesting for other fields, like look-dev, arch/product vis, draft rendering for film, it has a lot of potential.
+1, though it's interesting raytracing is suddenly a huge topic - anyone else remember the raytracing tests done with Quake Wars back in 2008? And personally I'd be more interested to see improved AI and physics than graphics, but pretty lights sell themselves.
I've been meaning to post a new art thread really soon. It's actually the opposite where I have far more art now, but I'm censoring parts of the story that might offend people.
For example, you mention the challenging status quo by arguing with professionals is annoying, but I've made cartoon characters who do exactly that. I've always wanted to make art that defies everything, but I also don't want to risk upsetting people who might not get the humor I'm going after.
You dont need to defy anything. You need to start getting
8 hours Sleep / day Heavy workouts Homecooked Meals
For games I think real-time raytracing is still years off. It's more interesting for other fields, like look-dev, arch/product vis, draft rendering for film, it has a lot of potential.
Kinda depends to which degree and everyone defines real-time raytracing as something different. Be it the full image completely path traced like film, or these new hybrid systems where just certain global effects are traced, incremental lightmaps whatever.
Mainstream heavy use is away agreed, we will see how dxr influences the AAA engine makers/games. In the end the problem is more or less that a lot of investment is already done in the "hacks" solutions (various shadows, baking, probes, screenspace reflections etc.), and they are pretty efficient and well understood, also scale across various hw. So we would need mainstream performance so high, that it makes sense for engine makers to actually "simplify" engines and really rely on the new tech alone. The whole new filter magic is interesting independent of raytracing, in the end the filter doesn't care where the data came from. I would not be surprised if we see quality improvements of the "hack" solutions as well
As for AI/physics, it's often quoted why not more progress here. It's much harder to enforce minimums on this front. Games scale across various platforms today, gameplay remains pretty similar, visual effects is just the most easiest way to achieve scalability across hw, without changing the game.
Though I fully agree with you that we have very interesting times ahead (as usual hehe) in the tools and actual development front.
As for complexity in general, the last evolution is likely to just do everything in the cloud. It's the most energy/cost efficient setup. The professional market often uses this setup already, and the other mediums film/music are there already. It's also kinda a "must have" for that step to deliver a big jump compared to anything seen before to win people. I don't know how far we are away from latency levels to be that good, but until then, everything is gonna be relatively incremental, like we are used to.
For some stupid reason, the forum software can't embed Youtube unless it's formatted like so: https: //www.youtube.com/watch?v=tjf-1BxpR9c&t=2m30s Specifically the /watch?v= part.
When you go to teh "Share" thing, and set it to a specific start time, copy/paste that URL back into Youtube and go there. Then copy the new URL it makes.
I don't understand the Arguments against better AI. I am by no means an expert, but i imagine the lack of better Ai has not much to do with hardware limitations.
Better AI seems to be mostly a problem of necessity.
Dumb AI is part of the charm of some open world games like GTA.
Having smarter AI without having more dialog and voice actors is a bit pointless. Skyrim and Fallout NPCs would still feel like cardboard cutouts even if their behavior was more intelligent.
Most games where the combat is the focus are multiplayer, so better combat AI isn't high on the list
Story driven games with some branching choices seems to be the ideal place for better AI, but the hardest place to implement it.
Having smarter AI without having more dialog and voice actors is a bit pointless. Skyrim and Fallout NPCs would still feel like cardboard cutouts even if their behavior was more intelligent.
In the not immediate future I could see a text to speech AI that could dynamically change its lines as being very plausible, or rather given loose parameters of what this average NPC might say and the AI being able to ad-lib around those. But I see that as something pretty far out given the amount of work that needs to be done. AI responses so far are pretty bland unless you feed them something they're made for exactly, though games are like this already. Then they'll need to know the context of how their voice should be, quick, yelling, frantic, stressed, quiet etc. And so far the speech patterns themselves are still so robotic so they'll need to better learn to synthesize voice with all the irregularities of an actual person (but I suspect we could get that sooner than all the other things).
But I imagine game companies would love something like that. Open world games have tons of voice lines, imagine the convenience of the storyteller being able to just program a simple dialog tree of just simple ideas that the AI could go and figure out how to proceed on its own rather than having a writer spend hours honing very specific dialog (except when that is something that is desired). I know most open world games tend to go through a lot of changes in dialog so not having to get voice actors in every time something changes I'm sure is something that would be incredibly valuable.
Obviously not related to rendering techniques but still pretty interesting to think about.
The idea of going back to text based NPCs isn't a bad one. I could easily see a computer AI or even an advanced "Siri" being the core of a game and basing it around the AI intelligence. That way you avoid the voice acting problem, but could still have something really dynamic.
We already have games kinda based around that idea, but with scripted dialog.
Hell, you could make Halo where Cortana behaves like a real AI assistant. Shut up Cortana, or can you tell me more about this place? Would be really interesting to have one know when you need help, or when you are exploring or idling.
I just saw this from the google IO event and remembered back to this thread and the discussions of AI doing text to speech and found this really compelling. I'm curious how good it is in the real world vs this small use case. Still though, I'm ready for the days of being able to have deeper experiences in games and having a semi real conversation with a quest giver would be awesome.
Was this posted here? So basically, this is like the next-step from Photogrammetry, it's more like Videogrammetry..We will be able to capture not only real life shapes but motion too!
I can see how this can effectively be used in the CG/VFX industry but in thee games industry....Mmmm...
Remember when Matrix had to set up a whole line of cameras to have a similar effect ?
the blade runner VR demo/freebie use something similar, the main chick whos comes and talks to you looks like she was using something like that. The artifacts were similar, like between the legs and arms/armpits you get weird geo gapping and spiking but from a distance it was pretty impressive.
the biggest bottleneck for games I can imagine would be the memory. like how la noir had a huge amount of memory taken up by the video/face stuff and the disk size was massive in comparison to other games.
Replies
You are linking articles and referencing what company spoke people have said, but these people usually have very little knowledge of the technical side of game dev and they simplify things a lot. the people that do deal with the technical side of game dev are the people you are currently arguing with, it's really like EQ says, you are a customer arguing with a veteran baker on how to bake bread.
Sometimes you just kinda have to take peoples word for it, I know I would and I probably have more experience then you.
I actually worry about the day when I can't be told why I'm wrong. It's for the exact reason why artists are always told to post their work and have people critique it. Without new ideas, there is no knowledge, and from there, there is no incentive to ever work hard.
I do want to work harder. I've had people from both inside and out of the industry tell me they were impressed by this. But this could never happen if I wasn't afraid to challenge ideas and status quos.
I actually want to see more people like this one day. They're not afraid to be laughed at, because they're in a cycle of constant improvement. Being wrong today is a better sacrifice if it means being a far more capable and intelligent person tomorrow.
If you want to get better, smarter, more experienced, you have to put in the work. Challenging the status quo by a means of incessant arguing with professions/experts who know infinitely more than you is not productive, it's just disrespectful and annoying. Show us with your actions that you're worth the effort. Less talk, more art, words are cheap as hell.
I just want to chime in on this specific point here, so that other people aren't misled by it. CPU performance is very relevant in all types of games, and being CPU bound in terms of general/total performance is common.
I've been meaning to post a new art thread really soon. It's actually the opposite where I have far more art now, but I'm censoring parts of the story that might offend people.
For example, you mention the challenging status quo by arguing with professionals is annoying, but I've made cartoon characters who do exactly that.
I've always wanted to make art that defies everything, but I also don't want to risk upsetting people who might not get the humor I'm going after.
I feel sorry for you since you genuinely dont seem to understand what a pompous ass that has nothing to show you look like. Your whole post history is 90% garbage. Stop polluting these boards with verbal diarreah and actually do some art. Otherwise keep beeing laughed at. People were more than honest and helpfull to you and you keep beeing a twat. Sorry im hungover.
What you are asking is for them to put in a shit ton of expensive components, and take an even bigger loss/gamble. I don't see madden and gta players (aka the mainstream) shelling out 1k for a console because they are gonna be getting technical rendering features they have no understanding of, or that make the game look 10% better to the untrained eye. Obviously 4k is where the focus is going and I think we are at the beginning of a hockey stick shaped graph in terms of adoption, so for sure that is going to be taken into account.
Games is first and foremost a business, and those console manufacturers have hundreds of people crunching the numbers in terms of production and marketing costs and how they can somehow hope to break even on a machine in a 5-10 year window. They know what they are doing. There is a big difference between what enthusiast gamers can hope and clamor for, and what a company can actually present to shareholders as a financially viable option. You could also probably thank the crypto market for less of a jump in a GPU next gen due to insane spikes in cost and demand.
I don't know if its just me, but personally I think the most visually impressive games of the last 4-5 years have been on these "underpowered" consoles. I remember having my jaw drop when I first played Ryse, Uncharted 4 is pretty much one of the most gorgeous games ever, and the new God of War looks like it's gonna blow my dick off on all fronts. I haven't seen much on PC these days that is hitting those visual targets even with all that extra power. Kinda proof that the technical side of things will only get you so far, the major gains are from insanely talented art teams nailing the execution based on years of experience squeezing every drop of perf outta those machines. And in todays market, those art teams are usually deployed on console projects, because that market tends to dwarf PC sales.
@JordanN what's said has been said, take it on the chin and let's all move on eh?
As juicy as this drama is, I like reading the discussions of future advances in real time so I'd hate for this thread to be locked.
That's why you get a console generation like PS3/XB360. XB360 was out from 2005 until it was superseded in 2013 by the Xbox one. But those were incredibly informative years to not change up hardware. The difference in the hardware between a modern system and one 8 years behind is immense. They're more about squeezing everything out of them over time to make up the short term losses (which obviously was your point). Which I think in their own way helps push rendering tech a lot by doing more with less, which in turn helps them cram more and more under the hood when newer more graphically intensive rendering techs start to become the norm.
I'm really happy to see the push from GDC this year to see the stepping stones of real-time ray tracing. And even if that Star Wars presentation was done on a $60k US machine, it was still unbelievably beautiful, my wife couldn't believe it wasn't just film until one of the animations didn't look entirely lifelike. In 2-3 years hopefully we'll see enthusiast chips that can do it for significantly less.
My main worry is that I feel with real-time raytracing, it's going to take a long time for major developers to switch to it since it seems to be something that only the newest tech can take advantage of, whereas developers need to make stuff for the largest pool of players. I feel like we'll need a new Crysis type game, essentially a tech demo, to really push the boundaries and get people excited to upgrade their hardware. But in the short term I imagine it will be a big gamble for one of the bigger devs to push to using something like that for a project that could be 3-5 years down the road, which is what I feel they'll need to do to get out ahead of it.
Either way though, the future looks really bright.
Plus, like the PBR shift, when people see it done right the first time (good 60fps performance) in a consumer game everyone else will scramble to do it too, because everything else just doesn't look as good anymore. I guess it's just a matter of when that will happen at this point, but certainly those planning for long term projects must have it somewhere in the back of their mind that trends will change and what is at the bleeding edge now may be very manageable when their project is released.
My reasoning for saying that it might be somewhat locked into just the new hardware is because I believe (and correct me if I'm wrong), on the Star Wars demo the main reason they were using the new Nvidia Volta chipset, which are made for AI/machine learning applications, is for the tensor cores alongside the Cuda cores. The tensor cores I think are the parts doing most of the machine learning stuff which is used to clean up the ray-tracing noise, and currently only the Volta series have those tensor cores. Maybe they'll find new ways to get it to work on existing graphics cards (which I think Otoy was doing with their Unity ray tracing engine(?) unless they also rely on newer cards as well). That was my thoughts on what could prevent it from being adopted right away by the big players in the industry, as soon as enthusiast cards are affordable enough to run it.
I'm just hoping someone will be at the cutting edge rather than waiting and playing it safe. But it could be like the VR rollout. Lots of small arcade games to start to show that the tech works, and slowly we see more full-fledged titles.
I'll also write the disclaimer that I only know at the most basic level of what's going on in graphics programming and the hardware that drives it (I'm an artist, and at the end of the day I just want to make pretty pictures). So if anything I've said is completely misinformed let me know, because I genuinely like to be better informed about this stuff, because if you aren't staying on it I feel you can get left behind really quick in this industry.
Jules says the current requirement are 2 GPUs (Volta???), but by the end of the year, they will get it down to one GPU running at 30fps with no noise.
8 hours Sleep / day
Heavy workouts
Homecooked Meals
Then maybe, maybe mix in some game art
Just believe me. This is no troll post.
I wouldn't risk it.
Stick to Turmeric.
Mainstream heavy use is away agreed, we will see how dxr influences the AAA engine makers/games. In the end the problem is more or less that a lot of investment is already done in the "hacks" solutions (various shadows, baking, probes, screenspace reflections etc.), and they are pretty efficient and well understood, also scale across various hw. So we would need mainstream performance so high, that it makes sense for engine makers to actually "simplify" engines and really rely on the new tech alone.
The whole new filter magic is interesting independent of raytracing, in the end the filter doesn't care where the data came from. I would not be surprised if we see quality improvements of the "hack" solutions as well
As for AI/physics, it's often quoted why not more progress here. It's much harder to enforce minimums on this front. Games scale across various platforms today, gameplay remains pretty similar, visual effects is just the most easiest way to achieve scalability across hw, without changing the game.
Though I fully agree with you that we have very interesting times ahead (as usual hehe) in the tools and actual development front.
As for complexity in general, the last evolution is likely to just do everything in the cloud. It's the most energy/cost efficient setup. The professional market often uses this setup already, and the other mediums film/music are there already. It's also kinda a "must have" for that step to deliver a big jump compared to anything seen before to win people. I don't know how far we are away from latency levels to be that good, but until then, everything is gonna be relatively incremental, like we are used to.
https: //www.youtube.com/watch?v=tjf-1BxpR9c&t=2m30s
Specifically the /watch?v= part.
When you go to teh "Share" thing, and set it to a specific start time, copy/paste that URL back into Youtube and go there. Then copy the new URL it makes.
Hope this helps!
Dumb AI is part of the charm of some open world games like GTA.
Having smarter AI without having more dialog and voice actors is a bit pointless. Skyrim and Fallout NPCs would still feel like cardboard cutouts even if their behavior was more intelligent.
Most games where the combat is the focus are multiplayer, so better combat AI isn't high on the list
Story driven games with some branching choices seems to be the ideal place for better AI, but the hardest place to implement it.
In the not immediate future I could see a text to speech AI that could dynamically change its lines as being very plausible, or rather given loose parameters of what this average NPC might say and the AI being able to ad-lib around those. But I see that as something pretty far out given the amount of work that needs to be done. AI responses so far are pretty bland unless you feed them something they're made for exactly, though games are like this already. Then they'll need to know the context of how their voice should be, quick, yelling, frantic, stressed, quiet etc. And so far the speech patterns themselves are still so robotic so they'll need to better learn to synthesize voice with all the irregularities of an actual person (but I suspect we could get that sooner than all the other things).
But I imagine game companies would love something like that. Open world games have tons of voice lines, imagine the convenience of the storyteller being able to just program a simple dialog tree of just simple ideas that the AI could go and figure out how to proceed on its own rather than having a writer spend hours honing very specific dialog (except when that is something that is desired). I know most open world games tend to go through a lot of changes in dialog so not having to get voice actors in every time something changes I'm sure is something that would be incredibly valuable.
Obviously not related to rendering techniques but still pretty interesting to think about.
We already have games kinda based around that idea, but with scripted dialog.
Hell, you could make Halo where Cortana behaves like a real AI assistant. Shut up Cortana, or can you tell me more about this place? Would be really interesting to have one know when you need help, or when you are exploring or idling.
Was this posted here? So basically, this is like the next-step from Photogrammetry, it's more like Videogrammetry..We will be able to capture not only real life shapes but motion too!
I can see how this can effectively be used in the CG/VFX industry but in thee games industry....Mmmm...
Remember when Matrix had to set up a whole line of cameras to have a similar effect ?
the biggest bottleneck for games I can imagine would be the memory. like how la noir had a huge amount of memory taken up by the video/face stuff and the disk size was massive in comparison to other games.