1 00:00:00,020 --> 00:00:03,240 Josh: Ejaz, I've been obsessed with these things called splats for like six months. 2 00:00:03,380 --> 00:00:08,840 Josh: And just last week, we had a new breakthrough in this revolutionary weird world of 3D imagery. 3 00:00:09,040 --> 00:00:11,720 Josh: So I want to start with this image that we're looking at on screen right now, 4 00:00:11,760 --> 00:00:14,720 Josh: because this is, it's pretty unbelievable. 5 00:00:14,720 --> 00:00:18,260 Josh: And for the people who are just listening, what it is, is there's a dude with 6 00:00:18,260 --> 00:00:22,700 Josh: the normal backyard and he has a Vision Pro and he has a photo that he took 7 00:00:22,700 --> 00:00:25,020 Josh: in the past of the same backyard in a different season. 8 00:00:25,020 --> 00:00:29,500 Josh: And with this new Splat technology, he's able to overlay a previous picture 9 00:00:29,500 --> 00:00:34,560 Josh: onto a current place and then actually walk through it as if he's able to relive 10 00:00:34,560 --> 00:00:37,820 Josh: the memory for the first time using these Apple Vision Pro headsets. 11 00:00:37,960 --> 00:00:41,700 Josh: So it's this unbelievable technology that Apple released just last week that 12 00:00:41,700 --> 00:00:47,260 Josh: allows you to relive the past in the present in a way that is totally immersive, 13 00:00:47,700 --> 00:00:51,600 Josh: totally submersive using these things like the Vision Pro or any sort of virtual reality headset. 14 00:00:52,410 --> 00:00:55,150 Josh: The new technology has some pretty unbelievable examples. So that's what we're 15 00:00:55,150 --> 00:00:59,750 Josh: going to cover in this episode is the weird world of 3D splats and how you're 16 00:00:59,750 --> 00:01:03,290 Josh: actually able to turn any photo you've ever taken, any video you've ever taken 17 00:01:03,290 --> 00:01:07,130 Josh: into something you could actually step into and relive again like it's the first time. 18 00:01:07,370 --> 00:01:11,050 Ejaaz: Yeah, it's pretty crazy. It looks like he's taken a picture of his garden from 19 00:01:11,050 --> 00:01:14,050 Ejaaz: December 10th, 2017, when it 20 00:01:14,050 --> 00:01:17,650 Ejaaz: was snowing and he's kind of transposed it onto his garden in real time. 21 00:01:18,010 --> 00:01:20,730 Ejaaz: I think the thing that shocked me the most from this, Josh, I did a double 22 00:01:20,730 --> 00:01:23,470 Ejaaz: take was they have a doctor who prop in 23 00:01:23,470 --> 00:01:26,430 Ejaaz: here it's like a telephone box who has 24 00:01:26,430 --> 00:01:29,430 Ejaaz: this in their gun that this is pretty insane um but i 25 00:01:29,430 --> 00:01:32,370 Ejaaz: saw a more recent example actually of this technology 26 00:01:32,370 --> 00:01:35,710 Ejaaz: this week um in a slightly uh 27 00:01:35,710 --> 00:01:38,790 Ejaaz: uh crazy application so uh 28 00:01:38,790 --> 00:01:42,750 Ejaaz: you might have heard that uh some some epstein files got released uh and some 29 00:01:42,750 --> 00:01:47,030 Ejaaz: documented footage pictures videography got released and someone decided to 30 00:01:47,030 --> 00:01:51,890 Ejaaz: splat the entire thing so what you're looking at here isn't a real uh video 31 00:01:51,890 --> 00:01:55,870 Ejaaz: but rather a series of images which have been splattered to form this kind of 32 00:01:55,870 --> 00:01:58,090 Ejaaz: like 3D immersive experience. 33 00:01:58,470 --> 00:02:01,930 Ejaaz: And he was able to generate this in a couple of minutes, which is insane. 34 00:02:02,430 --> 00:02:06,550 Ejaaz: And to kind of prove to you that this is like a real thing, what you're seeing 35 00:02:06,550 --> 00:02:13,590 Ejaaz: on my screen right now is a Google Drive of basically all the leaked images from these files. 36 00:02:13,950 --> 00:02:17,250 Ejaaz: And you can literally click on any of these. 37 00:02:17,350 --> 00:02:23,030 Ejaaz: Let's go with blue guest room two. and if you press w to zoom in you can now 38 00:02:23,030 --> 00:02:28,430 Ejaaz: literally zoom in and peek around the entire thing like let's see how close i can go first can 39 00:02:28,430 --> 00:02:29,470 Josh: You look at what's on the mic oh. 40 00:02:29,470 --> 00:02:35,790 Ejaaz: My god wait what's under the bed okay we don't know we don't know but like i 41 00:02:35,790 --> 00:02:38,870 Ejaaz: can't read the title you can really get into it oh i'm under the bed now 42 00:02:38,870 --> 00:02:39,490 Josh: Oh you're under. 43 00:02:39,490 --> 00:02:45,610 Ejaaz: Officially under the bed anyway this stuff is is just insane josh and uh this 44 00:02:45,610 --> 00:02:48,170 Ejaaz: is due to apple shop's model right 45 00:02:48,640 --> 00:02:53,060 Josh: Yes, this is the coolest thing. So Apple Sharp just released this model last 46 00:02:53,060 --> 00:02:56,400 Josh: week, which coincided with the release of these files, which created this funny 47 00:02:56,400 --> 00:03:00,520 Josh: convergence of two technology, or I guess a technology and breaking news at once. 48 00:03:01,100 --> 00:03:04,280 Josh: But I do want to talk about what splatting actually is because we're saying 49 00:03:04,280 --> 00:03:06,560 Josh: this funny word a lot. No one actually has a clue what it means. 50 00:03:06,680 --> 00:03:10,860 Josh: It's basically a way to make a 3D scene by storing it as a cloud of tiny. 51 00:03:11,020 --> 00:03:13,820 Josh: Soft little blobs instead of building traditional 3D models. 52 00:03:13,820 --> 00:03:16,500 Josh: So in the past, each as we've all played video games before, 53 00:03:16,520 --> 00:03:21,660 Josh: it requires a large machine to run them because it's a lot of textures and polygons. 54 00:03:21,740 --> 00:03:26,700 Josh: And there's just a lot of detail required to render fidelity with what these 55 00:03:26,700 --> 00:03:29,280 Josh: blobs do, what these splats do is each blob. 56 00:03:29,440 --> 00:03:32,300 Josh: It's kind of like a puff of spray paint and it's floating in 3D space. 57 00:03:32,380 --> 00:03:36,920 Josh: And then every puff has a position, size, shape, color and transparency. 58 00:03:36,920 --> 00:03:39,860 Josh: So to render an image from any viewpoint the computer 59 00:03:39,860 --> 00:03:42,700 Josh: projects all those puffs onto your screen and then blends them together 60 00:03:42,700 --> 00:03:45,880 Josh: like layering this transparent paint so it's much more efficient 61 00:03:45,880 --> 00:03:48,700 Josh: than the previous way of doing this which is 62 00:03:48,700 --> 00:03:52,180 Josh: lots of crazy rendering lots of compute required 63 00:03:52,180 --> 00:03:55,000 Josh: and if you were to create something like we're showing on screen now which 64 00:03:55,000 --> 00:03:58,000 Josh: is an example of 3d skulls that are sitting with dynamic 65 00:03:58,000 --> 00:04:01,080 Josh: lighting and it looks really real it's something you'd see in a video game you 66 00:04:01,080 --> 00:04:04,260 Josh: would normally have to render that overnight it would take forever to 67 00:04:04,260 --> 00:04:07,140 Josh: do but what this new splatting technology does and what apple's ai 68 00:04:07,140 --> 00:04:10,400 Josh: model does is it allows you to take this high fidelity 69 00:04:10,400 --> 00:04:13,240 Josh: rendering and turn it into a splat by using 70 00:04:13,240 --> 00:04:16,440 Josh: these blobs and that way you could render it as a single asset 71 00:04:16,440 --> 00:04:21,300 Josh: on something as light as your smartphone so it turns these really detailed complicated 72 00:04:21,300 --> 00:04:26,640 Josh: compute 3d images into something very simple so simple in fact that you can 73 00:04:26,640 --> 00:04:30,020 Josh: actually do it yourself you can make images yourself you can do this for your 74 00:04:30,020 --> 00:04:33,920 Josh: own content and you can do it for free almost instantly. It's really cool. 75 00:04:34,080 --> 00:04:37,060 Josh: And this is thanks to Apple's new Sharp model that they released last week, 76 00:04:37,200 --> 00:04:40,400 Josh: which is open source, which allows people to go around and play with it themselves, 77 00:04:40,520 --> 00:04:43,160 Josh: which I think is a really cool new paradigm for this technology. 78 00:04:43,800 --> 00:04:47,180 Ejaaz: So the way I'm thinking about this, Josh, and correct me if I'm wrong, 79 00:04:47,340 --> 00:04:53,760 Ejaaz: is if I take a 2D image, right, it's 2D and it's composed of a bunch of pixels, right? 80 00:04:54,080 --> 00:04:58,760 Ejaaz: If I use this splat technology on that image, it turns every one of those pixels 81 00:04:58,760 --> 00:05:03,500 Ejaaz: into a kind of like 3D object, which is why like these skulls, 82 00:05:03,600 --> 00:05:07,440 Ejaaz: for example, isn't a video of like the skulls from all different orientations. 83 00:05:07,440 --> 00:05:12,840 Ejaaz: It's like a couple of pictures of these pile of skulls or even maybe an AI generated 84 00:05:12,840 --> 00:05:14,900 Ejaaz: image or whatever and it's 85 00:05:15,190 --> 00:05:19,790 Ejaaz: splattered into these like 3d blobs and i can now kind of like maneuver around 86 00:05:19,790 --> 00:05:22,750 Ejaaz: it and look around is that kind of roughly on the right track 87 00:05:22,750 --> 00:05:25,930 Josh: Yeah and it's funny you'll notice with like the epstein examples 88 00:05:25,930 --> 00:05:28,550 Josh: there actually is a lot of fidelity in the way that it's rendered so if you 89 00:05:28,550 --> 00:05:32,530 Josh: look at a video game in the past that you've played normally you'll look at 90 00:05:32,530 --> 00:05:35,510 Josh: the details and it's very fuzzy it doesn't look very real because the computer's 91 00:05:35,510 --> 00:05:38,970 Josh: trying to save rendering of compute for the things that you're actually looking 92 00:05:38,970 --> 00:05:43,470 Josh: at that matter but these scenes that use this new technology They look so photorealistic 93 00:05:43,470 --> 00:05:46,930 Josh: because you're essentially repainting the world from these millions of soft dots. 94 00:05:47,050 --> 00:05:50,170 Josh: And it's fast because rendering is mostly throw blobs onto a screen efficiency 95 00:05:50,170 --> 00:05:55,170 Josh: rather than heavy 3D geometry or this like slow ray tracing that you're seeing on a computer. 96 00:05:55,510 --> 00:05:59,930 Ejaaz: Got it. And so the number one question that pops into my head then would be, 97 00:06:00,110 --> 00:06:04,530 Ejaaz: well, what's the cost difference for like doing this in the traditional way 98 00:06:04,530 --> 00:06:06,470 Ejaaz: versus like in the current splatting way? 99 00:06:06,670 --> 00:06:12,110 Ejaaz: And how long does this take? We got this tweet by Brad Lynch who tried out this this Apple software. 100 00:06:12,390 --> 00:06:15,090 Ejaaz: And he said he generated what we're looking at on the screen right now, 101 00:06:15,190 --> 00:06:19,430 Ejaaz: which is, I think it's an image of him sitting by like an open ocean and like, 102 00:06:19,670 --> 00:06:23,690 Ejaaz: you know, he's got his Apple Vision Pro on and he's like kind of peering around 103 00:06:23,690 --> 00:06:26,250 Ejaaz: and he's in his living room. He's moving the images side by side. 104 00:06:26,310 --> 00:06:30,830 Ejaaz: And it made it, it took him 10 seconds to generate this on his MacBook Pro. 105 00:06:31,050 --> 00:06:34,250 Ejaaz: And I'm assuming because he used Apple Sharp, which you just said was open source, 106 00:06:34,510 --> 00:06:38,030 Ejaaz: it's the cost of downloading the software and just running that on your computer. 107 00:06:38,230 --> 00:06:39,430 Ejaaz: Is that, Do I have that roughly correct? 108 00:06:39,570 --> 00:06:45,670 Josh: It's totally free. It's totally open source. And it is totally available for users of any computer. 109 00:06:45,950 --> 00:06:49,170 Josh: I mean, you could render it on a laptop. You could render it on a phone. 110 00:06:49,330 --> 00:06:50,810 Josh: You could do this instantly, anytime. 111 00:06:51,070 --> 00:06:56,150 Ejaaz: Hold your horses, Josh. If you're saying that, that means we should do this ourself, right? 112 00:06:56,330 --> 00:06:58,030 Josh: Perhaps we should do a demo for the audience. 113 00:06:58,050 --> 00:07:00,310 Ejaaz: Let's not take his word for it. Let's do this ourselves. 114 00:07:00,470 --> 00:07:03,210 Josh: Let's do it ourselves. We're going to do it live here. 115 00:07:03,450 --> 00:07:07,710 Ejaaz: Yes, we have the Apple Shop kind of software here. We're going to upload an image. 116 00:07:08,210 --> 00:07:11,830 Ejaaz: Let me see if I can find a convenient image of Josh oh what do you know 117 00:07:12,220 --> 00:07:15,940 Ejaaz: what do you know i have it here i'm curious how long this is gonna take look 118 00:07:15,940 --> 00:07:19,280 Ejaaz: at us good looking gentlemen on the screen we look good in 2d i would love to 119 00:07:19,280 --> 00:07:21,580 Ejaaz: know how good we look in 3d and 120 00:07:21,580 --> 00:07:25,680 Josh: That's the fun thing too is if you have like loved ones or your children or 121 00:07:25,680 --> 00:07:29,920 Josh: people who are younger you could really freeze these moments in time special memories special trips. 122 00:07:29,920 --> 00:07:32,900 Ejaaz: You've gone we're done josh by the way sorry oh my god before i even finish 123 00:07:32,900 --> 00:07:34,540 Ejaaz: my explanation we're done okay here 124 00:07:34,540 --> 00:07:35,580 Josh: We are let's see this walk me through. 125 00:07:35,580 --> 00:07:39,720 Ejaaz: The picture here oh it's like a watercolor effect you can see it materializing 126 00:07:39,720 --> 00:07:46,700 Ejaaz: oh it's rendering oh this is weird this is weird i'm zooming in oh wait i need 127 00:07:46,700 --> 00:07:51,100 Ejaaz: to oh my god wait can we oh my god i can see you from like above yeah 128 00:07:51,100 --> 00:07:52,560 Josh: Wait try to go closer and get. 129 00:07:52,560 --> 00:07:55,280 Ejaaz: In deeper wait hang on let me get in deeper let me get into this face get 130 00:07:55,280 --> 00:07:56,180 Josh: Up in our faces. 131 00:07:56,180 --> 00:08:01,360 Ejaaz: Wait how do i how do i like zoom oh my god oh oh so we can like see us from 132 00:08:01,360 --> 00:08:06,760 Ejaaz: the side too that's wait that's pretty crazy david's face is a little warped 133 00:08:06,760 --> 00:08:11,860 Ejaaz: david's the guy the less handsome guy uh in the center um our arms are looking 134 00:08:11,860 --> 00:08:15,600 Ejaaz: pretty good we're going to the gym oh we could look at us from above as well that's pretty get in 135 00:08:15,600 --> 00:08:16,860 Josh: There scroll into the image i'm. 136 00:08:16,860 --> 00:08:19,620 Ejaaz: Trying i'm trying to sorry i don't i don't mean to zoom into 137 00:08:19,620 --> 00:08:23,140 Ejaaz: our crotches here but uh this is the this is the furthest i can go in which 138 00:08:23,140 --> 00:08:26,820 Ejaaz: i'm kind of upset about hang on maybe if i maximize oh wow okay that that does 139 00:08:26,820 --> 00:08:31,440 Ejaaz: look better oh my god well what i'm impressed with here josh is number one how 140 00:08:31,440 --> 00:08:36,960 Ejaaz: quick that took to take it's funny on the screen it said uh it gave us a countdown from 60 seconds. 141 00:08:37,240 --> 00:08:41,120 Ejaaz: And it ended up producing the entire rendering that we're looking at right now 142 00:08:41,120 --> 00:08:44,540 Ejaaz: with 47 seconds to go. So it took 15 seconds to make. 143 00:08:44,920 --> 00:08:48,500 Ejaaz: And I'm touching my laptop right now. It is not warm at all. 144 00:08:48,620 --> 00:08:52,120 Ejaaz: So I'm assuming it didn't cost anything energy-wise as well. 145 00:08:52,600 --> 00:08:56,340 Ejaaz: Also, the fidelity of this, Josh, is actually really, really good. 146 00:08:56,460 --> 00:08:57,680 Ejaaz: It's way better quality. 147 00:08:57,800 --> 00:09:02,300 Ejaaz: I mean, I think I look a bit kind of out of it. I look like I've had a few drinks. 148 00:09:02,420 --> 00:09:05,620 Ejaaz: Maybe I did on the night, actually. But it looks really good. 149 00:09:06,330 --> 00:09:10,450 Josh: Yeah, it's impressive how quickly it's able to render this and how low cost 150 00:09:10,450 --> 00:09:13,530 Josh: it is and how lightweight it is. I mean, you could just run this in a browser very simply. 151 00:09:13,790 --> 00:09:16,830 Josh: It's not a very compute intensive thing. And it's really cool. 152 00:09:16,930 --> 00:09:18,810 Josh: So this is an example of a photo. 153 00:09:19,570 --> 00:09:22,530 Josh: There are three kind of tiers to the splatting. There's the photo first. 154 00:09:22,670 --> 00:09:26,470 Josh: And then second is this in between before we get to videos, which is this example 155 00:09:26,470 --> 00:09:28,230 Josh: that you're seeing with Casey Neistat's studio. 156 00:09:28,410 --> 00:09:30,810 Josh: Now, a lot of people who watch YouTube, they know Casey Neistat. 157 00:09:30,930 --> 00:09:31,750 Josh: They love Casey Neistat. 158 00:09:31,870 --> 00:09:35,810 Josh: And this is the most iconic place in the world of Casey. 159 00:09:35,810 --> 00:09:38,770 Josh: What the meta team has done is actually go 160 00:09:38,770 --> 00:09:41,790 Josh: through and create a giant splat of the studio so that 161 00:09:41,790 --> 00:09:44,830 Josh: anybody with a vr headset can actually put the headset on and walk 162 00:09:44,830 --> 00:09:47,550 Josh: around it now what we're looking at on screen it looks like it's an 163 00:09:47,550 --> 00:09:50,350 Josh: actual video of the studio but the reality is is that it's one 164 00:09:50,350 --> 00:09:53,510 Josh: large splat and it is a full fidelity 165 00:09:53,510 --> 00:09:56,270 Josh: splat so if you put on goggles and you walk through 166 00:09:56,270 --> 00:09:59,730 Josh: the space you can actually go and read the bindings 167 00:09:59,730 --> 00:10:02,550 Josh: on the books you can like walk up to all 168 00:10:02,550 --> 00:10:05,490 Josh: the shelves and peer at all the little things all the little trinkets that 169 00:10:05,490 --> 00:10:08,430 Josh: are on them it is a full and total high fidelity 170 00:10:08,430 --> 00:10:13,230 Josh: scan of the studio but in a very lightweight way if you used to try to do this 171 00:10:13,230 --> 00:10:16,770 Josh: you would need a supercomputer to render this and you need a supercomputer to 172 00:10:16,770 --> 00:10:20,890 Josh: run it you couldn't even do this on goggles that would be like disconnected 173 00:10:20,890 --> 00:10:26,110 Josh: from a computer now thanks to this new technology you can scan real places into 174 00:10:26,110 --> 00:10:28,250 Josh: the cyberspace and it's kind of acting as that, 175 00:10:28,940 --> 00:10:32,020 Josh: almost like a preservation technique, where if there's a place that you love, 176 00:10:32,120 --> 00:10:35,720 Josh: if there's a place you want to remember, you can actually scan it and then relive in that forever. 177 00:10:35,920 --> 00:10:39,560 Josh: You can capture this place in its full fidelity, exactly how it is today. 178 00:10:39,720 --> 00:10:41,360 Josh: And I thought this was a really cool example. 179 00:10:41,640 --> 00:10:46,040 Ejaaz: I mean, what I find super cool about this is like in the traditional way, 180 00:10:46,140 --> 00:10:49,660 Ejaaz: you would have to take a million pictures and stitch them all together, 181 00:10:49,660 --> 00:10:53,460 Ejaaz: which would have taken you hours and hours and efforts and probably a bunch 182 00:10:53,460 --> 00:10:55,580 Ejaaz: of people to get involved to help you produce. 183 00:10:55,580 --> 00:10:58,420 Ejaaz: Also, I like that it's to scale as well, Josh. 184 00:10:58,560 --> 00:11:02,220 Ejaaz: Like a lot of these simulation kind of videos or games that I've tried with 185 00:11:02,220 --> 00:11:05,740 Ejaaz: Apple Vision Pro and stuff like that just seems kind of unrealistic. 186 00:11:05,960 --> 00:11:08,160 Ejaaz: Obviously, like maybe you're playing a fantasy game or something. 187 00:11:08,280 --> 00:11:11,780 Ejaaz: This is like to scale. It's like you're walking through, you're not going to bump into anything. 188 00:11:12,440 --> 00:11:15,700 Ejaaz: I just think it's awesome. But then the natural question that pops into my mind 189 00:11:15,700 --> 00:11:18,140 Ejaaz: is, well, can you do this with video? 190 00:11:18,400 --> 00:11:21,500 Ejaaz: And I think, you know, we had our answer a few months ago earlier this year 191 00:11:21,500 --> 00:11:27,980 Ejaaz: when this viral tweet went about of this guy who's kind of like directly speaking to a camera. 192 00:11:28,240 --> 00:11:31,700 Ejaaz: But you can see in this video that someone's navigating around him. 193 00:11:31,820 --> 00:11:37,320 Ejaaz: And this is just like, you know, a 2D video taken head on of this man sitting down in his chair. 194 00:11:37,720 --> 00:11:42,200 Ejaaz: And he is able to navigate around him in every single different type of direction. 195 00:11:42,680 --> 00:11:46,740 Ejaaz: You can peer at him at a kind of like angle perpendicular to him. 196 00:11:46,880 --> 00:11:50,940 Ejaaz: You can see kind of like the way that his jaw looks like. And this is all generated 197 00:11:50,940 --> 00:11:53,000 Ejaaz: through splats. None of that is real. 198 00:11:53,160 --> 00:11:56,020 Ejaaz: None of that was actually filmed with a camera to the side of him. 199 00:11:56,180 --> 00:11:59,200 Ejaaz: This is all generated via splats. Super cool. 200 00:11:59,500 --> 00:12:01,780 Josh: It's fun to think of... 201 00:12:02,190 --> 00:12:05,970 Josh: When you capture things to think of your camera as a paintbrush or 202 00:12:05,970 --> 00:12:08,850 Josh: maybe even a can of spray paint like we were talking about earlier 203 00:12:08,850 --> 00:12:11,710 Josh: where if you can just capture the smallest amount 204 00:12:11,710 --> 00:12:14,430 Josh: of detail of a specific part of an image you can then 205 00:12:14,430 --> 00:12:17,430 Josh: render it all fully in a 3d way like 206 00:12:17,430 --> 00:12:20,590 Josh: we're seeing another image here where you can zoom in on the video you can pan 207 00:12:20,590 --> 00:12:24,790 Josh: left and right and that's because it was kind of scanned like it was this can 208 00:12:24,790 --> 00:12:27,590 Josh: of spray paint you you want to kind of spray paint things and then you could 209 00:12:27,590 --> 00:12:30,610 Josh: relive them and capture them and i think it's such a cool new paradigm where 210 00:12:30,610 --> 00:12:32,810 Josh: they're driving through the city streets or they're watching someone dancing 211 00:12:32,810 --> 00:12:34,530 Josh: or whatever these examples are. 212 00:12:34,710 --> 00:12:38,850 Josh: If there's something in your life that you want to capture, you can just do it and then relive it. 213 00:12:39,050 --> 00:12:42,090 Josh: And this is particularly interesting if you're a user of iPhone, 214 00:12:42,090 --> 00:12:44,990 Josh: because Apple's really the company who has been leading the charge of this. 215 00:12:45,230 --> 00:12:48,930 Josh: And if you use an iPhone, you're aware of the camera sensor, 216 00:12:49,070 --> 00:12:50,150 Josh: right? How they're kind of lined up. 217 00:12:50,330 --> 00:12:56,730 Josh: And when you shoot a video using these top two, you actually give real 3D spatial depth to it. 218 00:12:56,790 --> 00:12:59,690 Josh: And that's also because there's a lidar scanner on the bottom of the camera as 219 00:12:59,690 --> 00:13:02,710 Josh: well so apple has all the tools here built in to 220 00:13:02,710 --> 00:13:07,150 Josh: create the highest fidelity splats possible and now they're rolling out the 221 00:13:07,150 --> 00:13:10,790 Josh: software to enable that to happen even more so on these handheld devices that 222 00:13:10,790 --> 00:13:14,670 Josh: we all use it's like i guess the last example is it's kind of like if you take 223 00:13:14,670 --> 00:13:18,110 Josh: a black and white photo you could add color to it this is taking a two-dimensional 224 00:13:18,110 --> 00:13:21,930 Josh: photo and adding a third dimension to it and that's a really cool unlock my. 225 00:13:21,930 --> 00:13:27,670 Ejaaz: Mind naturally goes on to um like applications like what can i use this technology for. 226 00:13:27,850 --> 00:13:30,290 Ejaaz: And I think through a bunch of the examples that we've shown so far, 227 00:13:30,470 --> 00:13:35,330 Ejaaz: it's kind of cinematic and maybe even veered towards gaming as well. 228 00:13:36,120 --> 00:13:39,680 Ejaaz: Hollywood is the instant industry that I think of that I'm like is going to 229 00:13:39,680 --> 00:13:41,960 Ejaaz: get completely run over by this, right? 230 00:13:42,120 --> 00:13:46,100 Ejaaz: I know for a fact that they spend months, in some cases years, 231 00:13:46,580 --> 00:13:53,740 Ejaaz: to render a single visual splat that we've been looking at throughout this entire episode. 232 00:13:54,100 --> 00:13:58,040 Ejaaz: And so I think that this is going to cut costs down by like tens of millions of dollars. 233 00:13:58,280 --> 00:14:02,520 Ejaaz: And it is going to cut time down and even jobs as well. I know they have teams 234 00:14:02,520 --> 00:14:05,540 Ejaaz: of different people with different skill sets to stitch all the images together, 235 00:14:05,660 --> 00:14:09,060 Ejaaz: to get the right grading, lighting, to like post-processing of a bunch of these 236 00:14:09,060 --> 00:14:10,980 Ejaaz: images and then kind of make these visuals. 237 00:14:11,320 --> 00:14:14,900 Ejaaz: There's no way that this doesn't get disrupted by it. It also got me thinking 238 00:14:14,900 --> 00:14:17,500 Ejaaz: of one of my favorite shows on Apple. 239 00:14:18,340 --> 00:14:21,020 Ejaaz: Finally, we're talking about Apple and now they have this. Like one of their 240 00:14:21,020 --> 00:14:22,400 Ejaaz: hit shows is Severance, right? 241 00:14:22,520 --> 00:14:26,240 Ejaaz: And I remember season two, they have this like crazy scene where like we've 242 00:14:26,240 --> 00:14:28,700 Ejaaz: got the camera panning around him in all different kinds of ways. 243 00:14:28,700 --> 00:14:34,860 Ejaaz: And Ben Stiller had an interview on this where he said each episode costs roughly $20 million to make. 244 00:14:34,980 --> 00:14:38,760 Ejaaz: And this particular scene alone costs $10 million. 245 00:14:39,440 --> 00:14:44,400 Ejaaz: Now, if he had something like a splat technology, right, he could generate this 246 00:14:44,400 --> 00:14:49,980 Ejaaz: in a couple of minutes or even under a minute, like we showed ourselves earlier on. And for a fraction. 247 00:14:50,610 --> 00:14:53,690 Ejaaz: Of the cost it just it just it's a no-brainer for me josh 248 00:14:53,690 --> 00:14:56,590 Josh: Yeah i don't want to say that hollywood is 249 00:14:56,590 --> 00:14:59,550 Josh: under attack but they are definitely in need 250 00:14:59,550 --> 00:15:02,290 Josh: of rapid innovation quick because this is 251 00:15:02,290 --> 00:15:05,110 Josh: a second front that there is being disruption on we 252 00:15:05,110 --> 00:15:08,110 Josh: had we talked about google's vo3 and all the video generation 253 00:15:08,110 --> 00:15:10,950 Josh: models how well they understand the world how good 254 00:15:10,950 --> 00:15:13,710 Josh: they're able to generate video now this merges that 255 00:15:13,710 --> 00:15:16,510 Josh: gap where you can actually take the real world 256 00:15:16,510 --> 00:15:19,370 Josh: but you could capture it much more efficiently than you ever have before and 257 00:15:19,370 --> 00:15:22,670 Josh: much more fully which creates a lot more dynamic optionality 258 00:15:22,670 --> 00:15:25,690 Josh: for these shots so if you can't create it 259 00:15:25,690 --> 00:15:28,610 Josh: in the real world using a splat well then you can create it in the digital world 260 00:15:28,610 --> 00:15:31,790 Josh: using ai and what i understand is that people 261 00:15:31,790 --> 00:15:34,670 Josh: in hollywood they're already starting to use stuff like this where they are 262 00:15:34,670 --> 00:15:39,650 Josh: capturing things once instead of 10 times and they're using ai they're using 263 00:15:39,650 --> 00:15:44,070 Josh: splats to just kind of massage the scene to get exactly what they want if it 264 00:15:44,070 --> 00:15:47,870 Josh: wasn't perfect on that first try and it just saves a huge amount of money But 265 00:15:47,870 --> 00:15:49,290 Josh: there are more use cases for this. 266 00:15:49,710 --> 00:15:55,730 Ejaaz: Yeah, so you put this one in the dock, Josh, of a Swiss glacier collapse in 4D. 267 00:15:56,070 --> 00:16:00,470 Ejaaz: Is this, like, can Splats be used as, like, a prevention model for these kinds of things? 268 00:16:00,650 --> 00:16:03,370 Josh: Yeah, so earlier this year, there was a big landslide in the Swiss Alps, 269 00:16:03,390 --> 00:16:06,550 Josh: and it took out an entire village, and it was very dangerous. 270 00:16:06,650 --> 00:16:07,690 Josh: It created a very uncertain situation. 271 00:16:08,300 --> 00:16:11,920 Josh: Times because it's hard to access that area and people didn't know what was affected. 272 00:16:12,120 --> 00:16:16,180 Josh: So a helicopter came through with a big camera array and it just swept the whole 273 00:16:16,180 --> 00:16:18,740 Josh: valley. And you could see on the video the before and after. 274 00:16:18,920 --> 00:16:22,900 Josh: And they captured this incredibly high fidelity splat up the valley. 275 00:16:23,040 --> 00:16:26,500 Josh: They could then diagnose immediately what areas were most in danger, 276 00:16:26,780 --> 00:16:29,300 Josh: what areas needed the most help, how much danger there actually was. 277 00:16:29,440 --> 00:16:34,100 Josh: And they were able to observe all of the things at any time without needing 278 00:16:34,100 --> 00:16:37,960 Josh: a specific video feed of a specific area. So let's say you were looking at a 279 00:16:37,960 --> 00:16:39,020 Josh: specific location on the mountain. 280 00:16:39,240 --> 00:16:41,940 Josh: Well, you could just pause the splat and you could zoom in on that one area, 281 00:16:42,080 --> 00:16:44,180 Josh: even if you didn't capture it with a video camera. 282 00:16:44,440 --> 00:16:47,560 Josh: So there's a lot of utility for these outside of just entertainment. 283 00:16:47,560 --> 00:16:50,560 Josh: There's also safety and other things like this. 284 00:16:50,620 --> 00:16:53,860 Josh: I just thought it was a fun example of a real world use case that actually happened earlier this year. 285 00:16:54,400 --> 00:16:57,300 Ejaaz: Love it. It's been a longtime dream 286 00:16:57,300 --> 00:17:00,820 Ejaaz: of mine to go to Japan, and I've been fortunate enough to go a few times. 287 00:17:00,940 --> 00:17:04,660 Ejaaz: And I'm going again next year. And I kind of thought about like, 288 00:17:04,720 --> 00:17:06,940 Ejaaz: how do I share this experience with different people? 289 00:17:07,140 --> 00:17:11,300 Ejaaz: And I spotted this one, Josh, where it's a tweet that goes, we've made it possible 290 00:17:11,300 --> 00:17:16,160 Ejaaz: to walk through the hot spring town of Yamagata, Ginzan's Onsen with an avatar. 291 00:17:16,380 --> 00:17:20,220 Ejaaz: And this is like a real life rendering. And it looks like a game because, 292 00:17:20,360 --> 00:17:24,240 Ejaaz: you know, it's been generated one of using a splat machine or splat model. 293 00:17:24,930 --> 00:17:30,910 Ejaaz: It is to scale. These are real life shop fronts and stores and homes and streams. 294 00:17:31,550 --> 00:17:34,410 Ejaaz: But obviously, it's like a simulated game environment. And it got me thinking 295 00:17:34,410 --> 00:17:37,210 Ejaaz: like, this looks like something out of GTA, Josh, right? 296 00:17:37,330 --> 00:17:41,750 Ejaaz: And I'm thinking like, this would change the way that you create simulated realities. 297 00:17:41,930 --> 00:17:47,990 Ejaaz: Like imagine the Sims game, but using real life worlds, and it can be generated 298 00:17:47,990 --> 00:17:51,670 Ejaaz: in real time to reflect different kinds of people, personalities and shops. 299 00:17:51,730 --> 00:17:54,850 Ejaaz: Like imagine if New york city was updated every 300 00:17:54,850 --> 00:17:57,610 Ejaaz: single second or day or every hour uh to 301 00:17:57,610 --> 00:18:01,530 Ejaaz: reflect accurate goings-ons in that city i just think this changes gaming forever 302 00:18:01,530 --> 00:18:05,130 Ejaaz: right because one of the things that i loved about gaming is that uh it had 303 00:18:05,130 --> 00:18:07,990 Ejaaz: kind of like a preset story but then when you got to the end of the story i 304 00:18:07,990 --> 00:18:11,590 Ejaaz: was like damn i can't i now have to wait like a year or two years until the 305 00:18:11,590 --> 00:18:16,010 Ejaaz: second one comes out gta what is it five or six which one have we been waiting on for 306 00:18:16,010 --> 00:18:17,270 Josh: Oh we're waiting on six. 307 00:18:17,270 --> 00:18:18,930 Ejaaz: Six right so we've been waiting 308 00:18:18,930 --> 00:18:19,430 Josh: For like 12. 309 00:18:19,430 --> 00:18:22,330 Ejaaz: Years 12 years over a decade for this game 310 00:18:22,330 --> 00:18:25,050 Ejaaz: now you can get the second game or the third game or the fourth 311 00:18:25,050 --> 00:18:27,970 Ejaaz: game or the fifth game immediately if we had these kind of 312 00:18:27,970 --> 00:18:31,590 Ejaaz: generated simulated realities but 313 00:18:31,590 --> 00:18:34,710 Ejaaz: it kind of like i played this out in my head josh um 314 00:18:34,710 --> 00:18:37,910 Ejaaz: the end game for these splats surely has 315 00:18:37,910 --> 00:18:43,030 Ejaaz: to be world models right world models um is supposedly going to be a big trend 316 00:18:43,030 --> 00:18:48,050 Ejaaz: next year in ai models where you create these simulated realities or environments 317 00:18:48,050 --> 00:18:52,950 Ejaaz: of the real world that we live in today and you stick in an AI agent or an AI 318 00:18:52,950 --> 00:18:54,950 Ejaaz: model to kind of generate synthetic data. 319 00:18:55,050 --> 00:18:58,850 Ejaaz: So it lives out its life and it kind of figures out how humans perceive things. 320 00:18:59,310 --> 00:19:00,890 Ejaaz: Aren't splats just world models? 321 00:19:01,070 --> 00:19:05,530 Josh: They're not actually. I think splats, you can think of a splat kind of like 322 00:19:05,530 --> 00:19:08,470 Josh: what a neural link is to the human brain and AI. 323 00:19:08,730 --> 00:19:13,170 Josh: A splat is kind of like to the physical world and the virtual world. 324 00:19:13,330 --> 00:19:15,450 Josh: It's the bridge that combines the two together. 325 00:19:15,710 --> 00:19:18,410 Josh: So what we just saw in that last example is, 326 00:19:19,040 --> 00:19:23,020 Josh: you're able to walk through Japan and capture it with the camera and then merge 327 00:19:23,020 --> 00:19:25,680 Josh: that real world data into cyberspace. 328 00:19:26,060 --> 00:19:30,040 Josh: And if it was a world builder, it would just kind of create these virtual worlds. 329 00:19:30,160 --> 00:19:35,020 Josh: So what I see is kind of the way this goes would ideally be a combination of 330 00:19:35,020 --> 00:19:40,460 Josh: the two where splats are unbelievably efficient and are easy to capture the real world with. 331 00:19:40,660 --> 00:19:43,400 Josh: And then you could take an AI model, a world model, and you 332 00:19:43,400 --> 00:19:46,380 Josh: could apply extra fidelity on top of it depending on how much compute 333 00:19:46,380 --> 00:19:49,420 Josh: you have so you could think of the splat as 334 00:19:49,420 --> 00:19:52,200 Josh: a way of scanning the real world into the digital world and then 335 00:19:52,200 --> 00:19:55,540 Josh: the ai world models are a way to increase the fidelity 336 00:19:55,540 --> 00:19:58,280 Josh: using neural nets to predict what should be there 337 00:19:58,280 --> 00:20:01,120 Josh: to fill in all the blank spaces and to make it feel 338 00:20:01,120 --> 00:20:04,440 Josh: like more of a a real world plus experience 339 00:20:04,440 --> 00:20:07,600 Josh: so if you were walking through your childhood house and you were scanning it you 340 00:20:07,600 --> 00:20:12,380 Josh: can take a low or a high res but not totally high res version with a splat and 341 00:20:12,380 --> 00:20:15,480 Josh: then use ai to enhance it so then you can actually capture this place that's 342 00:20:15,480 --> 00:20:18,820 Josh: special to you and relive it forever using this cool new technology so this 343 00:20:18,820 --> 00:20:22,940 Josh: very much feels like a bridge into this future hybrid between the real world 344 00:20:22,940 --> 00:20:25,000 Josh: and the digital space okay. 345 00:20:25,000 --> 00:20:32,040 Ejaaz: That makes sense so if the mission is to help ai understand humans in all forms 346 00:20:32,040 --> 00:20:36,520 Ejaaz: of the way that they sense things the way that they perceive things sight um 347 00:20:36,520 --> 00:20:41,660 Ejaaz: audio visual stuff um instead of like relying on them to kind of generate it 348 00:20:41,660 --> 00:20:43,320 Ejaaz: from a bunch of data that we feed them, 349 00:20:43,460 --> 00:20:47,480 Ejaaz: we can kind of take our reality and surroundings, compress it into this splat 350 00:20:47,480 --> 00:20:50,840 Ejaaz: model, and then feed that into an AI model, a world model, 351 00:20:51,060 --> 00:20:53,420 Ejaaz: a simulated reality that they're kind of operating in. 352 00:20:53,520 --> 00:20:57,380 Ejaaz: And they'll have a more accurate depiction of how humans perceive the world, 353 00:20:57,460 --> 00:21:01,000 Ejaaz: which will then accelerate us to whatever the hell AGI is going to end up being. 354 00:21:01,520 --> 00:21:04,240 Josh: Yeah, it's like if you played with like Nano Banana Pro, for example, 355 00:21:04,420 --> 00:21:07,760 Josh: and you added an image that was old, and maybe it was a black and white image 356 00:21:07,760 --> 00:21:11,660 Josh: that was very low quality it can add color and it can make it feel more high 357 00:21:11,660 --> 00:21:14,940 Josh: quality that's kind of what this does before for more virtual spaces. 358 00:21:15,690 --> 00:21:19,910 Ejaaz: Well, what I like about that is we're just going to end up with an abundance 359 00:21:19,910 --> 00:21:22,790 Ejaaz: of data and data has been lacking. 360 00:21:22,970 --> 00:21:26,450 Ejaaz: I think at this point, every single model has been trained on the same corpus 361 00:21:26,450 --> 00:21:31,270 Ejaaz: of data and we need to start tapping into private buckets of data to add kind 362 00:21:31,270 --> 00:21:33,250 Ejaaz: of value or intelligence to an AI model. 363 00:21:33,530 --> 00:21:35,330 Ejaaz: This kind of bypasses that entirely 364 00:21:35,330 --> 00:21:38,450 Ejaaz: and creates this kind of synthetic but really accurate environment. 365 00:21:39,050 --> 00:21:41,930 Ejaaz: That's super cool. And then like in terms of like the end game here, 366 00:21:42,050 --> 00:21:44,950 Ejaaz: Josh, like do you think Apple is going to forge the path here? 367 00:21:44,950 --> 00:21:49,130 Ejaaz: Has Apple somehow dug themselves out of the grave or rather dug up themselves 368 00:21:49,130 --> 00:21:52,770 Ejaaz: out of the grave that they've kind of left themselves? They haven't been involved in AI or anything. 369 00:21:53,410 --> 00:21:58,370 Josh: No, they haven't been. And this is not by any means a real attempt at AI. 370 00:21:58,610 --> 00:22:00,070 Josh: This is kind of a separate thing. 371 00:22:00,410 --> 00:22:04,710 Josh: This is in regards to their vision platform. This is kind of like what the future of compute is. 372 00:22:05,130 --> 00:22:08,590 Josh: Everyone's building a pair of glasses. We have Meadow. We have Google now. 373 00:22:08,730 --> 00:22:12,070 Josh: We have, I'm sure Microsoft is working on HoloLens. 374 00:22:12,290 --> 00:22:16,810 Josh: Apple is the Vision Pro. everyone is working on this new spatial compute platform. 375 00:22:17,190 --> 00:22:20,850 Josh: Apple is definitely furthest along this path and granted the Vision Pro gets 376 00:22:20,850 --> 00:22:23,190 Josh: a lot of hate because it's very expensive it doesn't have a lot of use cases 377 00:22:23,190 --> 00:22:26,670 Josh: but what you're seeing here is an early prototype for what the future of this 378 00:22:26,670 --> 00:22:30,250 Josh: compute will look like when applied to actual consumer products. 379 00:22:30,670 --> 00:22:33,910 Josh: So if you scroll down to one more beneath this there's a really fun example 380 00:22:33,910 --> 00:22:38,610 Josh: where we take the splats that were mentioned that we mentioned in the video 381 00:22:38,610 --> 00:22:41,610 Josh: and you can actually put them and pin them on your wall in your apartment. 382 00:22:41,870 --> 00:22:46,170 Josh: So as you walk through the real world, you're able to pin these photo frames 383 00:22:46,170 --> 00:22:49,650 Josh: and it has the splat built in so you could walk and actually look into the photo 384 00:22:49,650 --> 00:22:50,750 Josh: and relive that experience. 385 00:22:51,110 --> 00:22:53,830 Josh: There's some other examples that they have where they've pinned widgets on the 386 00:22:53,830 --> 00:22:57,670 Josh: wall and as you walk into your apartment, these widgets that are like a calendar 387 00:22:57,670 --> 00:22:59,230 Josh: will just be present on your wall. 388 00:22:59,550 --> 00:23:03,470 Josh: And again, it's this merging of the digital and physical worlds and it looks 389 00:23:03,470 --> 00:23:06,710 Josh: real. It looks like it's embedded inside the wall. It's kind of embossed in. 390 00:23:07,070 --> 00:23:11,350 Josh: But what this leads to is this merging. It's this... 391 00:23:12,330 --> 00:23:15,370 Josh: Combination of digital and physical through these augmented glasses 392 00:23:15,370 --> 00:23:18,090 Josh: that we're going to get and splats are a really important part 393 00:23:18,090 --> 00:23:21,330 Josh: of that so when apple open sourced their model earlier this week that 394 00:23:21,330 --> 00:23:24,070 Josh: was a really big deal because it allows other developers to also lean into 395 00:23:24,070 --> 00:23:27,050 Josh: this and you could see even in this example you could scan in people's faces and you 396 00:23:27,050 --> 00:23:29,870 Josh: could speak to them in real time as if they were sitting right in front of you so it's 397 00:23:29,870 --> 00:23:33,790 Josh: this this fun entry into the metaverse and this is almost what i wish i saw 398 00:23:33,790 --> 00:23:38,070 Josh: meta was doing because meta being their new name meta they should be leaning 399 00:23:38,070 --> 00:23:41,550 Josh: into building some sort of a metaverse which is the combination of these two 400 00:23:41,550 --> 00:23:44,350 Josh: worlds and it seems like apple's actually the furthest ahead and this is kind 401 00:23:44,350 --> 00:23:48,670 Josh: of what it'll look like when it's implemented across consumer products as we go it. 402 00:23:48,670 --> 00:23:52,410 Ejaaz: Looks like something out of star trek or star wars um like you know you got 403 00:23:52,410 --> 00:23:56,010 Ejaaz: this holographic simulations of people speaking to you i think um 404 00:23:57,120 --> 00:24:01,220 Ejaaz: A major trend that's helped us get to what we're looking at today in front of 405 00:24:01,220 --> 00:24:05,440 Ejaaz: us and make all of this feasible is just massively reduced costs of things. 406 00:24:05,640 --> 00:24:10,440 Ejaaz: Like we've just spoken about like the cost of producing a splat or like a Hollywood 407 00:24:10,440 --> 00:24:13,280 Ejaaz: traditional version of this would have cost tens of millions of dollars. 408 00:24:13,300 --> 00:24:16,300 Ejaaz: And now it takes a couple of seconds and download an open source software. 409 00:24:16,460 --> 00:24:19,420 Ejaaz: That is just massive. I think the next biggest trend 410 00:24:19,420 --> 00:24:22,600 Ejaaz: is going to be the form factor specifically like i 411 00:24:22,600 --> 00:24:25,480 Ejaaz: can't help but sorry hate on how 412 00:24:25,480 --> 00:24:28,620 Ejaaz: big and uh bulky these uh where's 413 00:24:28,620 --> 00:24:34,880 Ejaaz: the where is it um big and bulky the uh apple vision pros look on on people's 414 00:24:34,880 --> 00:24:39,400 Ejaaz: heads i'm like that just looks so dorky it also kind of reminds me of um google 415 00:24:39,400 --> 00:24:42,160 Ejaaz: glass which is obviously a completely different product but looked also really 416 00:24:42,160 --> 00:24:47,160 Ejaaz: dorky and crazy for people to wear uh it seems like the form factor is going to be glasses, Josh. 417 00:24:47,940 --> 00:24:53,840 Ejaaz: Meta's making them. It was leaked this week that Apple is also potentially working 418 00:24:53,840 --> 00:24:56,140 Ejaaz: on a glasses version. That's not going to be Apple Vision Pro. 419 00:24:56,280 --> 00:24:59,200 Ejaaz: It's going to be much more slimmer, sleeker, thinner. 420 00:24:59,660 --> 00:25:04,580 Ejaaz: And then you have Google that's releasing Google Glass 2 2.0 next year. 421 00:25:04,760 --> 00:25:08,300 Ejaaz: And then Amazon apparently is even releasing one as well. So it seems like glasses 422 00:25:08,300 --> 00:25:09,500 Ejaaz: are going to be the form factor. 423 00:25:09,820 --> 00:25:14,720 Ejaaz: I think it's now cheap enough to produce at a much larger scale so that anyone 424 00:25:14,720 --> 00:25:15,640 Ejaaz: and everyone can use them. 425 00:25:15,760 --> 00:25:18,760 Ejaaz: But also I think like the components of making these glasses, 426 00:25:18,920 --> 00:25:21,700 Ejaaz: like the transistor and stuff are also cheap enough to run this technology. 427 00:25:21,860 --> 00:25:25,180 Ejaaz: So we're kind of at a culmination of all these trends coming together where 428 00:25:25,180 --> 00:25:29,240 Ejaaz: it's going to make the spatial reality that you've just described happen in 429 00:25:29,240 --> 00:25:30,360 Ejaaz: real life, which is super cool. 430 00:25:30,360 --> 00:25:33,760 Josh: We're getting close. It's like apples are big, bulky, and expensive. 431 00:25:33,940 --> 00:25:38,260 Josh: They're $3,500. They weigh a couple of pounds. They work unbelievably well. That's what you want. 432 00:25:38,600 --> 00:25:41,620 Josh: Meta's glasses google's glasses they 433 00:25:41,620 --> 00:25:44,360 Josh: suck they're cheap they fit on your head but 434 00:25:44,360 --> 00:25:47,200 Josh: they're a terrible experience so as we converge to the 435 00:25:47,200 --> 00:25:50,280 Josh: middle of whatever that is as we reach apple's quality 436 00:25:50,280 --> 00:25:52,900 Josh: with meta's form factor that's when you're going 437 00:25:52,900 --> 00:25:56,360 Josh: to start to see this stuff everywhere because it will be cheaply accessible 438 00:25:56,360 --> 00:25:59,080 Josh: and a really phenomenal experience and like you 439 00:25:59,080 --> 00:26:01,800 Josh: said these costs are coming down quickly it is only a matter of 440 00:26:01,800 --> 00:26:04,820 Josh: time until we get that perfect middle ground and this 441 00:26:04,820 --> 00:26:07,860 Josh: hybrid product exists where we do start to get these experiences 442 00:26:07,860 --> 00:26:12,300 Josh: available to us in our day-to-day lives in a package that is reasonable to walk 443 00:26:12,300 --> 00:26:15,660 Josh: around in on a day-to-day basis so that's one of the things i'm most excited 444 00:26:15,660 --> 00:26:19,480 Josh: about is this new frontier of hardware that is paired and supercharged by ai 445 00:26:19,480 --> 00:26:23,500 Josh: and all these other cool pieces of software like splats that we're seeing um 446 00:26:23,500 --> 00:26:25,740 Josh: unveiled pretty much every week now so. 447 00:26:25,740 --> 00:26:28,700 Ejaaz: The metaverse is basically becoming a reality 448 00:26:28,700 --> 00:26:31,640 Ejaaz: and i i'm so glad that we've moved on 449 00:26:31,640 --> 00:26:34,660 Ejaaz: from it being a fad to it being a reality 450 00:26:34,660 --> 00:26:37,520 Ejaaz: maybe we were just kind of like five years 451 00:26:37,520 --> 00:26:40,200 Ejaaz: too early with all the nft stuff from the crypto sector that 452 00:26:40,200 --> 00:26:43,040 Ejaaz: we saw uh way back when um but this is this is 453 00:26:43,040 --> 00:26:46,180 Ejaaz: awesome and i'm excited to kind of see this scale to real life applications 454 00:26:46,180 --> 00:26:49,920 Ejaaz: josh like it's all and well seeing like these demos of things but i can't wait 455 00:26:49,920 --> 00:26:53,980 Ejaaz: to see the first uh splat movie so that i can kind of like see things from different 456 00:26:53,980 --> 00:26:57,540 Ejaaz: orientations get people's different perspectives i can't wait for it to hit 457 00:26:57,540 --> 00:27:01,640 Ejaaz: gaming and fast forward gta 6 7 and 8 so we don't going to have to wait another 458 00:27:01,640 --> 00:27:03,700 Ejaaz: decade for these releases. 459 00:27:04,200 --> 00:27:06,800 Ejaaz: And I'm excited about the costs and the form factors that are going to come 460 00:27:06,800 --> 00:27:08,220 Ejaaz: with this, like being able to wear glasses. 461 00:27:08,480 --> 00:27:12,060 Ejaaz: I'm curious, right? Because I was super skeptical when AirPods became a thing. 462 00:27:12,280 --> 00:27:16,480 Ejaaz: And then I'm like, oh, I wear them all the time now. And it's just kind of like embedded in culture. 463 00:27:16,580 --> 00:27:18,960 Ejaaz: It's going to change the way humans kind of interact with each other, 464 00:27:19,040 --> 00:27:20,540 Ejaaz: which is going to be super, super cool. 465 00:27:21,140 --> 00:27:24,240 Ejaaz: But that is it for the rest of this episode. 466 00:27:24,920 --> 00:27:28,220 Ejaaz: It has been quite a week and quite the year. 467 00:27:28,380 --> 00:27:33,420 Ejaaz: This is our Christmas day episode we hope you feel cheerful you know Josh and 468 00:27:33,420 --> 00:27:36,020 Ejaaz: I came in our best Christmas attire I came with uh 469 00:27:36,930 --> 00:27:41,090 Ejaaz: Let's call it the coal for being bad, I guess, the coal color. 470 00:27:41,230 --> 00:27:42,230 Ejaaz: And Josh, you're repping red. 471 00:27:42,630 --> 00:27:47,130 Ejaaz: That's awesome. If you are somehow so passionate about Limitless and you're 472 00:27:47,130 --> 00:27:50,250 Ejaaz: listening to this, I just want to say thank you. That's frigging awesome. 473 00:27:50,690 --> 00:27:54,390 Ejaaz: There have been thousands and thousands of you that have joined our community, 474 00:27:54,510 --> 00:27:58,250 Ejaaz: that have subscribed to us, that listened to us week in, week out. 475 00:27:58,390 --> 00:28:01,970 Ejaaz: And it means the world to us. It means the world to me, especially. 476 00:28:02,610 --> 00:28:07,570 Ejaaz: And it's just awesome to have you guys here. We know about like 80% of you aren't subscribed. 477 00:28:08,430 --> 00:28:13,390 Ejaaz: In the spirit of Christmas, we would love if you tap that subscription button, 478 00:28:13,630 --> 00:28:16,030 Ejaaz: if you tap the notification button, or if you're listening to this on Spotify 479 00:28:16,030 --> 00:28:19,770 Ejaaz: and you don't even want to see our faces, give us a rating. It would mean the world to us. 480 00:28:20,330 --> 00:28:23,710 Josh: Also, this is one fun fact is we just crossed 100 episodes for Christmas. 481 00:28:24,010 --> 00:28:26,850 Josh: So in terms of Christmas gifts, one, we have 100. 482 00:28:27,130 --> 00:28:29,910 Josh: Two, thank you. You guys listening is the Christmas gift. And three, 483 00:28:30,090 --> 00:28:33,390 Josh: we will continue to post all throughout the holiday season as a gift to you 484 00:28:33,390 --> 00:28:35,230 Josh: for supporting us. All year round. 485 00:28:35,570 --> 00:28:39,390 Josh: So thank you for the support. As always, you know, this is like the best part 486 00:28:39,390 --> 00:28:41,950 Josh: of it is just being able to just like see the success of it, 487 00:28:42,010 --> 00:28:45,110 Josh: see people listening, sharing it with your friends who would also enjoy it. 488 00:28:45,430 --> 00:28:49,470 Josh: Ejaz ran into someone the other day who randomly was talking about the show 489 00:28:49,470 --> 00:28:52,410 Josh: but didn't recognize Ejaz's face because he had never watched the videos, 490 00:28:52,410 --> 00:28:54,510 Josh: only recognized the voice, which I thought was so funny. 491 00:28:54,650 --> 00:28:58,070 Josh: So it's a really nice thing to see the message spreading. 492 00:28:58,290 --> 00:29:01,370 Josh: So thank you for that. I guess happy holidays to all who are listening. 493 00:29:01,970 --> 00:29:03,510 Josh: We are going to keep the episodes.