Futurist Logo

3D Worlds: How Real-Time Rules


Sylvio Drouin

Real-time 3D computing is vast, but most people are not aware of it. RT3D tech powers online and mobile games, digital twins, industrial simulations and the much-ballyhooed Metaverse. And it has important implications for training machine learning and autonomous systems. This week’s guest expert, technology innovator Sylvio Drouin, has been working at the forefront of RT3D for decades. Join us for a tour of the emerging 3D future.

Analysis complete. No addtional information is required for context. Proceed with transcript display ...

View Transcript

document button

this week on the futurists I think that it's it's tooto understand how hard it's going to be to build as long as we don't understand what it is that we that we will need tobuild you know there is this representation of the metaverse of the special web as the Ready Player Oneexperience which is most probably what what it's not going to end up being okayso I think that the why and the the what have not been answered yet digital twinis really a way to represent the outside world to enhance the interaction of realuser with this real world so example if I do a digital twin of the of an entireairport I will be able to enhance the interaction that my Travelers goingthrough this Airport have with the airport facilities foreign[Music] welcome back to the futurists I'm Robturcic and my co-host Brett King here with me hi hey Brett you know one of thethings we keep hearing about on this show it keeps turning up is this topic of the metaverse this uh this idea thatsomeday we're going to be spending all of our time not on Zoom like we all have been forthe last couple of years during the pandemic but instead it's sort of a 3D version of Zoom uh where we'll be ableto be in the same place and look around and look up and look down you might think of it as like an immersive game world that we're going to spend all ourtime in now there's a lot of controversy about this because candidly that vision is pretty far out in the futurebut what a lot of people don't know is that real-time 3D has quietly grownto become an enormous force and I'd say real-time 3D is even big is way biggerthan the metaverse well of course this includes um you know the new tech thatwe're looking at with Apple's uh the the new rumor is they're going to becalled Apple reality glasses yeah yeah we heard about the trademark filings for Apple for realityit's not just apple right but Microsoft's been working on real-time 3D for ages companies like Nvidia powerit's uh there's a number of big companies in the space so this week what I thought would be fun to do is to bring on a good friend ofmine someone I've worked with in the past and someone I have great respect for who is an expert in this topic he's been working at the very Forefront ofreal-time 3D for more than a decade so let's give a big welcome to Silvio druanSilvio hi welcome to Lou thanks for having me we're so happy to have youhere uh now Sylvia your title is the SVP of innovation at unity and um Unity isone of those companies that flies under the radar for most people I would say a lot of people deal with unity or a lotof consumers touch Unity technology but they're completely unaware of it in a way it's sort of like you know BASF orwhatever that company is that that says you know we don't make the magnetic table we make it better what does unitydo and how does it make real-time 3D better but Unity is the unity is uh is what wecall a game engine and for a lot of people you know when I say this they say what is a game engine you know it is thepiece of software it's a piece of software that is used to do two things to create games by game developer aroundthe world and it's also the engine that power those games that make those game run when the consumer or the playerdownload and install and play that game so you know as much as people don't know about us uh at the Forefront as a brandthey certainly play there's certainly three billion people uh monthly that play Unity made games on iOS on Androidon PlayStation or Nintendo switch on Xbox you know on variousplatform so this engine which is called a 3D engine a 3D game engine is not justfor games it's to power other type of application in the in various uhIndustries in film in construction in architecture simulation digital twinlike you know there's like an unlimited number of application of tech we'll getinto that in a minute because we definitely want to dig into some of those You Dropped a bunch of keywords there but I think the big point that youjust made which I think is really important is that there are three billion people on Earth who are playinggames you know for most folks we tend to think of games as something that's secondary you know it tends to besomething you think that your teenage kid plays in his bedroom or something most people don't realize thatmore than half the population considers themselves Gamers and most people have at least a couple games on their mobilephone about 60 percent of people so three billion people are playing mobile games I I put that out there becausewhen we hear about the metaverse there's so much hype about the metaverse but if you took all the people in the world who are currently using the metaverse it'snot even a hundred thousand people it's a tiny number relative to the number that are using gamesbut 3D games are important because their conditioning half the users of theinternet every single time they play they're conditioning them to 3D interfaces yeah and 3D is how we dealwith the world 3D is what we're used to when we move around in the real world you know if you think about it it's a little bit unnatural that for 20 or 30years we've been sitting in front of digital screens and trying to understand the world through this rectangle ofglass you know a flat two-dimensional representation and so in some ways 3D might be the fullrealization of the potential of the web but that's a hard thing to do tell us a little bit about the the what makes ithard to create a 3D version of the web what are the challenges I think that it's it's hard to to understand how hardit's going to be to build as long as we don't understand what it is that we that we will need to build you know there isthis representation of the metaverse of the special web as the Ready Player One experience which is most probably whatwhat it's not going to end up being okay so I think that the why and the and andand and the what have not been answered yet you know building a game is alreadya very tedious process because if you want to build something that is rich in in content and narrative there is a lotof work and uh building an experience to uh to visualize a factory with iot withthousands and thousands of of iot device uh connected to the actual 3D simulationthat's a different set of problems right okay so okay so you're saying that we need to do proper problem definition andyes nobody's done a very good definition of what the metaverse is or what it should be or why we need it exactly butlet's talk a little bit about the challenges around 3D because this has been a long time coming and and to put it in perspective since games is thearea where there's a lot of traction and then later we'll talk about how we grow from games into other fields but gamesis an area where today there's a lot of use it's worth bearing in mind just to put this in a kind of historical perspective 20 years ago real-time 3Ddidn't exist and it was really hard to render things on the fly it was relatively primitive let's put it likethat you know so people might recall that back in the day console games that had the best power to render uh you knowrender sequences they had cut scenes little movies that would happen in between you know when the system wasmoving you to the next level you'd see a high resolution movie but that was fully rendered in advance it was basically just like playing a cartoon a 3Danimation and then you get into the game where the graphics were noticeably less Vivid and less sharp and less crispthat's the point uh bring us up to today what has changed in 3D technology tomake it possible now where you've got you know on a on an Xbox it's like a motion picture when you watch a you knowsports game it looks like you're watching the NFL on television what has changed in 3D technology in the last 20years but what's changes the evolution of the GPU you know of of the graphiccard graphical Processing Unit yeah that means like obviously the software to render evolve but the hardware is what'sbeen most uh interesting and it's going to carry on evolving you know and the company said what I wouldlike to I would like to but I would like Robert to to outline this two different real-time uh there's two differentmeaning to the real-time term okay real time these days is what we talk about when we can render a game in real timeit means that based on the input of a user the game can change in real timereal time applied to the metaverse or this like you know 3D layer that we'regoing to add to reality through various form of AR and is a different form of real time because it is a real timewhere yes things are going to be rendered in real time but things also the the data piece of the real-time uhaspect to the special web of the augmented reality set of experience thatwe'll see coming to us over the next decades that's a different set of that's a that's a different meaningokay what's the difference between the game responding to my input or let's sayan augmented reality experience responding to all the input of the world around it the the basic difference isthat the game answer to you using your you know the the source of uh data thatwill affect the game is either your your your screen that you touch on your phoneor your joystick that you're going to use on your Xbox so all the real-time input is coming from one user exactlyleads you to track whereas in a population you need all the input from the whole world so it's a lot moreinformation to change exactly exactly but I would assume that um game designers would make very gooddesigners of the web three three world for the the real world interfacesbecause they're just used to building in the 3D format they're using building in the 3D format for one player or formaybe multiplayer but that that that that that all play using the same modelwhile in an AR experience that will have a input from millions of data pointsinteracting the data point that is the one of the most difficult aspect of thinking about a a layer of realitypresented in real time for end users it's a different way of dealing I meandealing with millions of data points to actually make sure that the the proper content is triggered at the right timeis a problem that game developer and game designer have not faced at all sothis comes back to like how how do we train web 3 developersum you know like you know what are what are the skill sets they need is it a similar path to game designers or wellhang on a second it so you're conflating two concepts that are really important that we have to distinguish I think web3is a proposition that we're going to rebuild the web the highly centralized web 2 which is based on these biginternet platforms web 3 proposes to decentralize it using blockchain Technologies that's that's web threeyeah what we're talking about here is is the 3D web and I know the two terms arevery similar so I can see how it's confusing and it's easy to switch those two things but I think it's important to maintain a clear line between that butthe other point you just raised which is really important is that there are not enough 3D developers right now in the world to build these uh these theserealities these alternate realities whether it's the metaverse or some sort of industrial application like a model and that's a big topic for thisdiscussion we should get into that in a bit um but what Sylvia's doing right now is telling us about what the challenges areand I think this is the thing that people have been skeptical about when they hear about the 3D web and we'vebeen hearing this concept for more than a decade I've certainly been hearing it since you know for even longer than thatum people become a little bit skeptical about it frankly because it never seems to happen it sounds like a sunriseEternal Sunrise where the Sun never fully Rises on this industry but one of the reasons for that is that the problemset is incredibly difficult it's an incredibly Big Challenge I don't think people grasp just how big a challenge itis so what Sylvia was just describing is that within a um within the context of a gameyou can shrink the number of variables down the game world is known the game the level map is known the number of youknow uh adverse adversaries or bosses or hurdles you have to face those are all known and so the only variables then areuser input which is hard that's that's plenty difficult uh and then the Game responds in real time to that andrenders all that stuff for people who are not familiar with 3D rendering what's important to note is that this is very different from what you see in amotion picture where we've had 3D special effects for a long time for 30 years yeah the special effects in amovie you're talking about just one aspect one angle one View and it's locked off it's rarely movingum and so that can be rendered with a lot of precision but it takes months and months and months of rendering time togenerate that so typically with a movie uh they'll finish principal photography with actors on the stage in front ofgreen screen and then they'll spend nine months or a year in post-production rendering frame by frame those shots andthose sequences in the background so you can see the monsters and the special effects and the aliens and spaceships and all that stuff that doesn't existthat can't be photographed but that's a very different proposition from what we're talking here real time means thecamera can move the player can move and the whole world has to be re-rendered in real time to support that yeah I meanlet me see like like I think we can use a very simple example for the audience but a simple example if if I play a gameif I have a game I don't know like it could be like a shooting zombie game okay if I played this game immerse in aVR headset uh I know the game knows exactly where the Character Are I canmove the character around I can move my own character around it's easy theconstraints are well understood okay if I take this game and I play and and Iwant to make sure that it's it can be played against against the physical reality around me Ihave to cope with because I don't know in which environment my user is going tobe so I have to build a very different game okay the and a game that can thatthat will have to work everywhere I could be at the park I could be in my kitchen I could be in my living room Icould be on the bus I could be anywhere okay so if you think about this and youthink about the type of EX of of experience that will be produced the complexity is uh exponential that'sright that's right because you have to render it from every one of those perspectives right so the view from the bus is going to be very different fromthe view from the bus stop or something exactly the other factor is is other people right so multiplayer introducestremendous amounts of complexity because now the system has to render multiple points of view for multiple people at the same time andit has to do that with absolutely precise synchronization it can't be off by a millisecond because it would ruinthe Illusion for one of the other players not only that but it couldn't you know there's evidence that could make you sick for example that's exactlyright that's why people get seasick when they're wearing a VR headset if there's even a millisecond or two of of lag timeyou know so when you turn your head and it takes a second for the world to catch up you feel seasick because that's theexperience we have when we're on a ship at Sea so you're exactly right Brett and so the the um the more users you add themore complex that rendering problem becomes and the more processing power that you need have I got it right so farokay so this is one of the reasons why the metaverses that have been built so far are not very you know well populatedthey're sparsely populated because the system simply can't support thousands and thousands can you give us an exampleof what you you think about the what's been built so far in terms of of is like do you consider like Roblox asmy typos or it's I consider it Like A Primitive version of the metaversecertainly you know absolutely because a lot of the things you're talking about Rob we can do in Roblox but it's justvery low um low Fidelity that's right so let's break it down like this there are game worlds today that do supportmillions and millions of users those include Roblox and fortnite and they're excellent games but to the point we'remaking a minute ago those are worlds that are designed for a single purpose principally for one main purpose whichis to play a game so a lot of the variables are known now of course they're expanding into things like games and performance art and so forth sothose are like a Proto metaverse to Brett's point but they're all yeah but don't forget fortnite doesn't you can'tbe a million user in the same game instance that's exactly right yeah you can be like I think it's like a hundredplayer per game instance which means that it's like a multiplayer game that's exactly right and that's the limitationwe're just talking about the ability to render multiple points of view simultaneously is a very challenging problem yeah the second kind ofmetaverse though is this non-game metaverse Silvio's Point earlier thosemeta verses aren't very well defined we're not sure what they're for if you listen to people like Mark Zuckerberg share their Vision you know they tellyou that we'll do everything we do in the real world we'll have meetings we'll have friends we'll we'll have parties we'll you know we'll watch TV shows andlook at PowerPoints inside of virtual worlds I'm not sure that's such a compelling use caseum but you know certainly uh Facebook's um Horizon worlds is an example of a prototypical metaverse and there arenon-centralized versions companies like decentraland and sandbox have launched their own versions but these also runinto the exact same constraints that we just spoke about right so so what we're trying to do right now in this part of the show is Define what are the hardproblems to solve for a real-time 3D web to actually happen and one that we'vecome up against right very clearly in this very early part of the show is that we don't have a way to render that muchsimultaneously like we can't support a million simultaneous users do you think that's going to be solved in the futuredo you think there's a way to do that in the future that means uh just this is not just about rendering it there's like you knowthere is multiple problem uh associated with with a million user and it's alsolike do we want to be do you want to be in a world where there is a million user at the same time you know that's anotherissue you know like can we find narrative and game narrative that lead to a game where you would be a millionuser with a million user I think that you know when we talk about the real-time treat you have we're I thinkit's important to to outline that you know that they aregames games is not the real-time 3D web what I see as the real-time treaty where you know there is two things it's it'sfirst it is a way to present information using three-dimensional interface okaythe uh question here is is that useful okay yes I can present like a Amazonproduct or something I want to shop for as a 3D object that I can manipulate onmy screen or in my VR headset is that useful to complete my shopping or toenhance my shopping experience not proven yet okay so and then the the other aspect of thisis is yes will I serve the web going from an experience to the other experience to the other experience likeI do today when I use my web browser and I go from link to link to link to link do I need to do we see a future where uhFortune 500 companies retail shops uh anything will want to present theirinformation in 3D it's probably a be a mixture of what you have today and whatis going to be actually useful for user to see in 3d not clear yet again that'swhat are you talking about hey you're talking about AR applications right no I'm going to come to AAR now the otheraspect of this is a 3D layer on top of the physical reality that userexperience every day that is probably going to be more useful okay there's a lot of very useful application to add a3D layer when I'm visiting a city when I'm looking looking for the right roadto take to get to a place you know like I it's it's that's there's like multiple application fixing the the workerworking in the factory or a plant to do like a repair of like some Machinery sothat's very useful okay so the first one which is when we talk about the 3D webwhich is like the know the transition from 3D experience to 3D experience is a bit less defined in terms of how usefulit's going to be the second one which is which is the augmented reality one it'smuch clearer that it's going to be very useful now on top of these things you have games and you have entertainmentand that's that's that's not going to change entertainment in VR it has notbeen proven to be very appealing yet you know of the 10 million quests too that have been sold not many users use themthey're mostly in the closet okay twice a month on average we've tried it'sbecause one of the reasons that we've tried to jam linear contents you know games that were built for 2D screenseven if it if it's if they're rendering in 3D the actual narrative is like veryit's like a linear progression of scene in a VR even if it's if it's stillrender using the same 3D Tech you have more than 3D you have more than threedimension because you can move your head you can move your your hands you can do a lot of things that you cannot do witha 2d screen in front of you okay so so we have not understood the actualNarrative of how to create very compelling VR experiences and and gameswe're like you know we're like film wise like 125 years ago where it took 15 years to actually find the the to uh toto to actually the Cinematic language the language yeah yeah yeah yeah language itself we're at the same stagehere that's why people are not are not using this and so Roblox is like thesilent movies of the metaverse it's like the silent movie very well said all right let me see if I can round this upthen so so what we've covered so far is that real-time 3D is emerging and it's useful today in game worlds persistentgame worlds things like Roblox fortnite and other games um and by the way there's like a thousand of those game worlds isn't thatthis is not a small thing they're used by three billion people the big question and maybe the big opportunity is canthis be extended to other fields um for other kinds of entertaining or social applications business applicationsindustrial uses maybe City Planning or transportation planning simulations forthe environment and so forth that's a question we'll get into in the second half but what we've already heard inthis first part is a really important distinction that Brett brought up which is that there's really two kinds ofreal-time 3D applications or worlds if you will one is an immersive experience that you go into like a virtual realityor a game world that's a destination I'm going to immerse myself in this other worldmany of the metaverse concepts seem to be built on that premise but the second one is one where you kindof cover the world around you the real world you cover that world with data you have you you paint the world with pixelsif you will the real world and now that's visible minted augmented reality experience so that distinction I thinkit's important for our audience to get that that there's two kinds there's immersive worlds 3D worlds that you might go into or the real world thatmight be covered with data we're gonna go to a break now and after break we'll come back and we'll delve into some ofthe applications and use cases of this technology so you're listening to the futurists uh it's myself Robert turcic and Brett Kingmy co-host and we're interviewing Sylvia Joanne stick with us because in the second half we're going to get into thePractical use cases welcome to Breaking Banks the number oneGlobal fintech radio show and podcast I'm Brett King and I'm Jason henricks every week since2013 we explored the personalities startups innovators and Industry players driving disruption in financial servicesfrom incumbents to unicorns and from cutting-edge technology to the people using it to help create a moreInnovative inclusive and healthy financial future I'm JP Nichols and thisis breaking Banks foreignwelcome back to the futurist I'm Brett King your host with Robert tershek and our guest today is Silvio drawing fromUnity um Silvio um one of the things that sort of comes into this world of AR and and so forththat we're talking about is that the concept that when we need to representyou in the virtual world or place you in an augmented reality world we need to beat a you know understand some things about you and this extends into multiple areas now that we're thinking ofum around this concept called a digital twin or a digital equivalent of yourself we also have the emergence of digitalpersons which will be you know non non-player characters if you like that will will interact in this world but thedigital twin concept how does that fit into this emerging world of VR and ARthe digital swing concept is probably uh you know when we were talking earlier about the the the the why of themetaverse is probably one of the most uh well understood and useful concept youknow and they are real projects real project and real usage and the digitaltwin is really a way to represent the outside world to enhance the interactionof real user with this real world so example if I do a digital twin of the ofan entire airport I will be able to enhance the interaction that myTravelers going through this Airport have with the airport facilitiessomething that I would not be able to do if I did not have a digital representation a digital twinrepresentation of that airport so you can apply this to anything youknow buildings and mall and uh a block of a city or an entire one uh a plant uhyou know anything that is a physical system can be transformed into a digitaltwin maybe that's why that sort of makes it interesting for Amazon's acquisitionof Roomba yeah because Roomba gives you a A Primitive way of mapping out 3Dvirtual spaces at least a floor plan exactly that's right that's right and actually you know these phones that wehave here uh you know the reason there's three cameras on this phone is that that's how the phone can sense threedimensions right it's able to capture it's able to do that dimensional mapping people don't realize that but your phoneis constantly measuring the environment around you uh so we're starting to develop then sensors that's all thephone really is in that instance is a kind of sophisticated sensor that can sense three dimensions so we're startingto solve the problem problem of sensing in three dimensions that means that the network or you know the the cloud computing system can have anunderstanding of the 3D world around you that's one part of the problem that's not the whole problem the other problemthen is to render all that in real time that's what we were talking about in the previous section you know this conceptof digital twins not new it's been around for more than 30 years early use cases were uh cern's Large HadronCollider you know it's a gigantic multi-kilometer installation that's just too big which I had a chance to see withthe personal private uh so that's a that's an example of an industrial installation that's so bigthat no human can see the whole thing at one time therefore you need a digital simulation in order to understand itexactly the earliest users of of digital twins are NASA they were used in space because most of uh deep spaceexploration today isn't done by humans it's done by robotic systems it's all those Voyager uh all those Voyagersystems that are sent out into deep space or into the you know into the deeper into our solar system it's morelike the Rover that are that are basically not being Mars you know so those threw off Telemetry data that canbe captured and used to render kind of a digital version here on the earth and that's how you can control that remotesystem and understand where it is and how you know what its attitude is and position is and where it's at and soforth so those are like remote control systems and monitoring systemsum but I think what's important about the new generation of digital twins for industry is that they're no longer usedjust to monitor or render or show us a picture they're also used to control those systems in other words the modelis starting to be controlling the real world and that's where artificial intelligence and real-time data becomeextremely important for digital twins today you have to have high fidelity data and you have to have an AI that youreally trust because it's going to be making decisions not only about what you see but about what actually happens in the real world that might be a factoryor a plant or a business process can you comment a little bit about theuse of simulation in Industry this way 3D simulations simulation is really different than youknow it it's it's related to digital twin and not that means it's it's more used in in the world of in theindustrial world today to simulate and understand the problems in complexphysical system before they actually happen and uh and it's also obviously tosimulate those complex system before you actually build them so that you makesure that they're going to work and so the uh the amount of data Ai and machinelearning that we can feed into this basically enable those physical system builder to understand the problems andthe intricacies of what they're going to build and that's that's one way to generalize this is obviously likevarious case and and but in order to understand this as to why we would do 3Dsimulation that's the reason that's right yeah there's also the fact that ifyou have have built a simulation of the way a Machinery Machinery or system willwork that once that is built you could use that same simulated data or thatsame design data to Overlay for example 3D virtual understanding absolutelyproblem solving and fault solving and so forth the ultimate example of that not being necessarily just fixing equipmentin a factory but something in the case of like um augmented surgery where surgeons areusing that data you know a previous scans of you know a tumor or somethinglike that in a in an operating environment simulation also applies you know like if I'm building an airportagain let's let's come back to the example and I want to know with thearchitecture or the plan that I have how The Travelers will walk and flow throughthis airport I can actually simulate this to understand what I need to change in order to make sure that there's nobottleneck that's right and that's a really good point um because I think some people might be wondering how does a game engine fitinto all of this we've been able to use 3D you know a computer-aided design systems for for decades now so therehave been all sorts of ways to render architecture in 3D that's not new yeah that's not new simulation is reallybasically the merging of AI and game engine and that's it and the reason weuse a game engine is because the game engine has the motion the kinematics the physics so you can populate that 3D model theart the architecture if you will you can populate that model with people and dogs and you know kids playing basketball andballs exactly and all the physics of that will be uh will be will be quite accurate that's really important foranyone who's doing planning like urban planning or industrial planning or space robots in a factory you know you youreally want to actually understand well how they're going to interact so that's another example well and and thosesystems can be used to train an AI as well so an artificial intelligence can take really you know a virtual worldthat's rendered in 3D to us would look like a game world uh simulation butpopulated with simulated people simulated animals simulated trees and so forth all this stuff is really high highvarious similitude now it's high actually this is a very good point uh you just made simulation is a very goodway to actually produce the data that will be used to train the neural net Models All right so uh that's anotherway to uh you know there's like uh if you have complex neural network and youdon't usually you don't have the actual data that you need simulation can be used to generate that data yeah that'sright and that's useful for training purposes so with ML with machine learning more dataequals more training which means you get to a better result fast exactly and so we can accelerate the training uh youknow kind of like uh collapse the learning curve for the ml system by using a simulated worlduh we're getting deep into the weeds here um but the the the promise of all this technology from my viewpoint isthat something like industrial metaverse might emerge as a more compelling usecase sooner than consumer metaverse earlier you said there's not a really compelling consumer use case for themetaverse I think that's probably true most people are sort of like meh you know they're not sure why they would use it why would I put on a VR headset to goto the office on the other hand there's already compelling use cases in industry for simulation those are already widely usedand we can we can train astronauts in the 3D space we can train surgeons inthe 3D space we can we can yeah we can train robots and artificial Learning Systems uh so what is it the metaverseis it the metaverse because the metaverse is really about the connection of the different experience togetherif I have a single use case which is example a simulation of an airportis that the actual metaverse if it's not connected to anything outside I thinkthat there is a confusion in what the actual metaverse is you know if I have a single application of my AR glasses to Idon't know and there is like maybe I have like five apps on it and one is to find my myway around the city another one is to help me make food another one is to help me do this with this okay is this theactual metaverse if I go from uh I'm inNew York I'm at the Empire State Building and I put my glasses on so that I can understand the history of of theplace and I can press a button or I can tap a button and be transported into a game where the Empire State Building isthen it starts to be you know I think that over the years over the coming nextyears we're gonna we're gonna start to understand the actual the word metaverse as a set of linked experiences okay howdo we get there like how do we get what what do we need to solve between today and that vision of the future to getthere yeah what's stopping us from having a miniverse today but I think that you know I think thatthe initiative like the open metaverse of the standard uh is a very good is a very good effort into this direction youknow okay yeah uh okay so that that's about interoperability Between Worlds and forfolks who don't know what the term interoperability means it means um well for instance when uh this the apps youdownload on your iPhone won't work on your Android phone that's because they're not interoperable if you havedata that's trapped in one software program and you can't move it over the way things used to be with say your contacts in Microsoft Office uh you canexport those out easily in the past that was that was because it was not interoperable sometimes that's by Designin the tech World Tech companies often like to foist the proprietary standard that locks people into their system butalready now even in this early stage of the metaverse uh a lot of Industry leaders not all but a lot of thecompanies that are in the lead have come together to create an interoperability Forum to make sure that the worlds theybuild can link together and people can easily migrate from one world to the next uh so that's the point that Sylvia wasRaising I think that's a good point but there's other stuff that has to happen as well if we're really going to start to create simulations of the worldaren't we going to need a completely new generation of processors and and what's the load on the network these are BigData models yeah and the energy requirements yeah it's huge it's huge it's expensive uh because it's expensiveto use I know a thousand or ten thousand GPU in in the cloud so uh theinfrastructure the computational infrastructure will have to seriously uh evolve and uh and grow in order to andwhile you know keeping this through a affordable price yeah in order for us toactually you know go into this uh full speed so obviously the real-time 3D webis a great thing for the companies that make the graphical processing units those are companies like Nvidia and AMDfor instance silicon manufacturers but that's a considerable cost that someone's going to have to bearis that one of the reasons why this hasn't been broadly taken up outside of specific use cases and you know veryspecific micro worlds and simulations for particular Industries is that one of the reasons why it hasn't broadly caught on no I think that on from an industrialperspective it it has caught on that means it's it's clear that more and more and more companies as I you know as Ihad said like you know five years ago that there's going to be a need uh for fortune or there's going to be a want ora need for fortunate partner company to start going from a 2d world to a 3D world uh in all sorts of use of of usecases and that is happening that's clear okay what's not clear is the and it'sclear also that 3D that uh the the consumer is playing 3D games okay sothat's clear all right what's not clear is the browsing of the internet using a3D you know a set of of 3D experience or interfaces and then again what is thisyou know how will AR pan out and how will VR Banner so that's not clear yetokay so maybe it's a periphery around the periphery industrial use cases onone hand consumers playing video games on the other hand um we're starting you know we kind ofwork our way towards something new in the middle that spans do you think there's a possibility that the web thatwe understand is today you know this idea of basically a bunch of destinations they're usually rectanglesof information you know text-based worlds uh even if you think about YouTube is just a bunch of rectangleswith Tink text links yeah that's the web that we have today do you think that might just exist in parallel to some 3DWorld metaverse or some 3D experience that is very different from the web thatwe're familiar with the web we've had for the last 30 or 40 years your question completely clearly but Ithink that we are not going to one day wake up and suddenly everything is is going to be ofcourse not yeah yeah in front of us so there's going to be mixture of you know and co-existing of of 2D experienceswith 3D portals and you know we're not gonna this is gonna be and and we might be 20 years from now still readingarticle in a 2d web browser you know we're not necessarily want to read you know so it's always going to be thismixture and I think that over time the mixture will be very clearly definedas the use case become more clear and as the usage become more clear which isn't which it is not to date all right okaywell tell us what you envisioned for the future so one of the things we love to do in this show is to is to get aforecast yeah yeah take us out 10 20 30 yearswhen all these uh these kind of narrowly network problems have been solved so a long-term Vision what I see is obviouslythe evolution of augmented reality to a level where it enhanced our life and wedon't have to scroll on our phone all day because today we spend our entire day on the phone pulling information tous which is mean that we scroll and we scroll and we scroll don't stop all day and I think they will move away fromthat world to a world where uh where content will be contextualized triggeredbased on the context that I am in you know I will manifest intent and based onthe competition of millions of data point this content will come to me andthat will basically hopefully I would hope that it will allow us to have better human to human interaction andnot be always on our phone and so I believe that this will evolve a lot and that's a very interesting muchmore so to me than the VR aspect of what we're talking about and I believe so Ibelieve this in terms of evolution long term machine learning and uh and AI willobviously evolve you know a lot now if we can make Quantum Computingwork the mixture of quantum Computing and machine learning will be very veryimportant very very also dangerous but very very important yeah we're stillsome way away from that right now we'll still somewhere you know it could be 20years it could be 50 I don't know but once we have this once we have connected machine learning and Quantum Computingwe will I think that long term 50 like years 100 years we're going to connectnanotechnology systems to to that so then you're going to have a very powerful AI able to create physicalthings so uh that's one of the that's a bit of like you know long term long term if youreally ask me long term that's what I I believe is gonna happen I I just had aconversation with a technologist who I'm uh friendly with who said that he is now using um AI for game design and he saidlook it's a hack it doesn't work 100 but this is his technique he'll be with his daughter at the park and he'll starttalking to gpt3 and he'll start with a short description of a game and ask gpt3to write a game description and he said he gets a decent game description yeahthen he uses he takes that description and goes to Dolly two or it's amid-journey and asks it to simulate or knock out some uh AI generated images you know some to portray like what mightthat look like that game description now he gets some images generated um and then he goes over to anothersystem that allows him to simulate a rough model of the game he said look it's not perfect it's it's a Cheesy hackbut he said with zero effort while he's playing in the park with his daughter he can actually start to knock out screensof a potential game design and at least get to the point to say is this worth pursuing is this worth exploringwhat you're describing is that in this future scenario where these artificial intelligence systems are so good that they cangenerate not just words and text or images but they can actually start to generate entire experiences and maybeeven start to engineer physical products for us yeah that's that's your vision for the futurethat's a powerful you know I started to think about this like seven years ago where I'm like whatis the future of content creation right and like if I want to bring like you know 100 million and half a billionCreator to create content it's way too hard today so it was clear to me thatmachine learning driven content authoring will would be the actual future and we are you know if you lookat gdp3 and you look at Dali and you look at mine journey and you look at others uh we gonna get there we're gonnaget there within like you know the next decade we'll be able to create complete 3D scene and complete 3D games by justtalking to the game engine yeah and that's very powerful you know yeah infact yesterday on Twitter I saw a woman uh has developed a hack she's she's built some sort of system where she cantake all the single frames of a video so she just shoots a video of herself walking down the street yeah and she canget uh Dolly too to generate um and line up almost perfect with almost perfect Precision a virtualoutfit and so it renders a team by scene and then she found another program she can import all those individual shots intoand it recreates the video and what she ends up with is um basically simulation mapped onto thevideo stream that she started with and again it's a hack it's a workaround yeah but those sort of things that we'reclosing the gap on those capabilities pretty quickly now so fast exactly that's what's so astonishing I mean gpgthree I saw uh well I've heard talk about gpt4 that it's going to change theworld yeah right that um it is the closest thing to the touring test that we've ever seen secondly you know I'veseen demonstrations recently with these AI um you know artistic tools taking avideo frame and in the example I saw was a tennis player playing on a tennis court and completely changing theenvironment so that you take the tennis player and put him on the surface of Mars or the moon or whatever just bytelling the AI to change the uh the background it's evolving very very fast right now so that means means we'regoing to be giving superpowers to everybody very soon where people with just uh just by talking to a phone or bytyping a few commands on a keyboard they'll be able to generate video scenes and manipulate video scenes so I have areal concern about this because that's like giving people superpowers to generate a lot of deep fakes and they'llbe incredibly compelling deep fakes and you can just imagine where that's going to lead to so I think we're alreadythere you know like right now we could create convincing deep fakes that thisis where we need to have a different way of thinking aboutum not just digital assets but identity and those sort of things you know I likeyou think about the fact that today we carry around a physical driver's license and a passport to identify ourselves inthe in the physical world that's not going to work in a world of deep fakes and so forth because you're going to beable to simulate someone almost perfectly the only way you're going to be able to tell is whether it's asynthetic ID or a real person ID I think butI think the concept of identity which is more related to your to what you wereinitially talked about about web tree is key to uh to the future of uh ofentering this new era of a massive more amount of data that's going to come atus yeah and data that could be manipulated yeah the various usage of this Techdeflate being one is going to become more and more and more dangerous if we don't establish some form of realidentity which is you know too easy to to fake right now yeah you know we needto do something about this because well you know we we will be certainly gettinginto the topic of identity on a future episode of the futurist it's interesting that it parallels it right yeah it'sinteresting how we talk about 3D for the whole uh podcast but then we end upsaying oh you know what in order to evolve in this world of 3D we're goingto need to protect identity and really identify the source uh yeah yeah allthis ability to simulate different worlds leads us to question what the real world consists of and what's reallytrue so we end up questioning the truth and questioning realityyeah because you see imagine that in 20 years or 15 years you've got uhrendering of reality so a photo realistic that you cannotunderstand if what you're looking at is the real world or the digital worldwhere does that go it's the simulation man you know well on that note thank you so much Sylvia dran the senior vicepresident of innovation from Unity thank you so much for joining us this week on the futurist session it's a real greatpleasure talking to you thank you if you liked listening to the futurist this week you know what you can do you cangive us a five-star review you can give us a shout out on social media all that helps other people find uh the episodesand the content and that helps us support the program on an ongoing basis also my thanks go out to the productionteam and the team behind us supporting us from provoke including Elizabeth severins uh uh Kevin herscham silby anduh Carlo working on the social media side and of course out to you guysum but uh we will be back next week with more compelling future focused content until then we'll see you in the future[Music] well that's it for the futurists this week if you like the show we sure hopeyou did please subscribe and share it with people in your community and don't forget to leave us a five star reviewthat really helps other people find the show and you can ping us anytime on Instagram and Twitter at futuristpodcast for the folks that you'd like to see on the show or the questions you'd like us to askthanks for joining and as always we'll see you in the future [Music]

Related Episodes