The AR Show: Karl Guttag (KGOnTech) on the Apple Rumors and Four-Corner Market Opportunities
11:32PM May 2, 2023
Welcome to the AR show where I dive deep into augmented reality with a focus on the technology, the use cases and the people behind them. I'm your host Jason McDowall.
Today's conversation is with Karl Guttag. Karl is an industry analysts speaker and the author of KGonTech, a technology blog at KGu ttag.com. Karl has 40 years of experience in graphics and image processors digital signal processing and memory architecture as well as micro displays for use in heads up displays and AR glasses. He's received 150 patents related to these technologies and many billions of dollars of revenue attributed to those inventions. Karl spent nearly 20 years at Texas Instruments and was named a TI fellow the youngest in the company's history. In the 25 years or so since he's been a CTO at three micro display system startups in two of which he was also a co founder. He was also recently the Chief Science Officer at Ravn, a company developing a hardware and software platform to deliver mission critical intelligence to military and first responders. Like my previous interviews with Karl, this was a long and wide ranging conversation that is split into multiple parts. In this first part, we touch on the rise and fall of high end digital photography, silicon based cameras sensors, the challenges that unique manufacturing processes have with volume and price, the downsides of the smartphone supply chain rumors about Apple's efforts in VR and AR the size of the VR market, the military and enterprise opportunity for AR for corner markets, quote unquote, for corner markets, and how they apply to VR and AR and the benefits and risks of video pass through VR. Part two will continue with a deeper dive into micro LED display and popular optics technologies. As a reminder, you can find the show notes for this and other episodes at our website thearshow.com. And please support the podcast at patreon.com/theARshow. Let's dive in.
Karl, on your blog, you take so much care in capturing a good image of the various display and optics technologies that you cover. How did you develop and feed that passion of yours for photography?
When I was young, I actually bought early on I took a lot of pictures. My mom was a picture taker. And then somewhere in the family that passed down to me where I took a lot of of the pictures of the family. And then in high school, I decided to get a film SLR film based SLR and started taking pictures, but you really couldn't get very good at it because it's very expensive back then, imagine you'd spend, I don't know several dollars, which would be like 10 or $20. Today on A roll of 24 pictures. And then you spent another several dollars, which would be my $20 worth today getting in the developed. And so you really couldn't develop the art or learn how to really use a camera well. So after a while I gave up on that I went back to using snapshot cameras. Then when I left AI back in 98 I bought another film SLR but I got a little more into it and started taking pictures. Still used it mostly like you would use a snapshot camera didn't really get to develop the art much. But then around tooth and I took him one time I was at Disneyland I was putting together I was working with I kind of gotten into Photoshop, which ties back to other things I did ti back in the 1980s. We made graphics accelerators, I was the technical lead of the 340 graphics family, which was the first moderately successful graphics processor using a bunch of high end graphics cards. And those graphics cards were largely used for either Photoshop or Autodesk AutoCAD. So that's that's where those my graphics chips tended to go into. So I knew about Photoshop and I got into as I left TI I got interested in doing a work with Photoshop. I like working with pixels. I was doing graphics chips at TI, then I got to work with the kind of the end result of it, manipulating images in Photoshop, which is something I enjoy. Well with that SLR I took a bunch of pictures and I was compositing these photos together. You know, I took like 1000s of pictures at Disneyland back then I had a little bit of money. Film was still expensive. And then I scanned like 2000 negatives to it so I could bring them into Photoshop so I can edit them. That was around 98.
By 2000, a company called canon came out with a digital SLR called the D 30. And I was one of the first in line to buy one of those after scanning 2000 photos because when you scan stuff it back then the software wasn't particularly great. You cannot imagine how much crap the smallest little dusts and the people who print stuff if you went to your local camera store to get them developed. They didn't take that much care. So there are bubbles, there's dust. There's all kinds of problems on the film. You never can clean it in the negatives you can't clean them properly. The scanner I had was not particularly high quality and it couldn't get rid of all the dust. So you end up doing a lot of editing just to clean the photo up after you scan it into Photoshop. Well if you
Do That 2000 times, you have a digital camera, which I believe back then cost over $1,000 may have been 2000, it was expensive back then it'd be like five or $10,000. Today, I did this math one time, and it was shocking how much that camera cost. And that was a three megapixel, three megapixel, APS C, which is like six tenths of the 60% reduced to full frame, digital camera. So that in turn got me into going to a photo site called DP review. So I was a very DP review was founded around 9819 98. And I was definitely a heavy user by 99. So I was one of the early adopters of a company of a website called DP review. And those of you who are in the photography community may know in 19, and 2007, DP review was bought out by a little company called Amazon company that barely existed at the time it was founded. And then just this last week, Amazon announced that after 25 DP review has been around for 25 years DP, there's shutting it down, Amazon decided to shut it down and lock the site. And that had this has the entire digital photography community and a roar, because it was the main site for information if you had a question about how your camera because it was well organized, it had broken down by camera type. And even within if there was popularity of a certain camera style like Canon, I can manufacture there would be breakdowns within the cameras you use. So I've been using that site really, for almost all of it's 20, about 23 or 24 of his 25 years I've been on that site. And to have that go away, is a big shock. And they're really and I've been looking around this i and a bunch of other people have been looking around, there is nothing equivalent to that site on the internet. And they say they're gonna lock it and cancel it. Because what they've got, it's got not only their main pot body, it's like reviews and stuff. But I was mostly on the on the discussion groups, where you get a lot of help you try to get help and give help, it's a very communal thing where you try to help people and people help you. Because when you have these complex cameras, so I kind of learned how to do photography, because once you have a digital camera, the virtue of it is you take a picture. And with a film, even if you could afford infinite amounts of film, your cycle time between taking a picture and seeing what that meant, or what that did, could be days or even a week. Whereas with a digital camera, you take the picture, it doesn't cost you anything to take the picture, you can see it instantly see now you can experiment out the heck out the wind Wang with it, because you can try different focal lengths, you can try different depths of field and you can play with all those technical things that only really high end photography pros could do prior to digital photography and digital cameras. So anyway, so that's so that became my resource for learning how to do digital photography, and doing really doing fairly good photography and and that as time goes on. This is like a 98 one of the big things that happened in the early 2000s, there was a company called fobian. I think there were around 2002 or three or there abouts. And they came up I got quoted, there's a book called by George Gilder called the silicon AI. I was quoted in that book, you can actually look it up under Google Books, that page 270 It's actually got me on it, I got into a fight, sort of a, a sparring match with people who said fobian is going to take over the world. Well, since most people don't really know who fobian is, you can see that didn't happen. And what fobian did, it's kind of neat. Technically, it's really neat scientifically, and this is going to tie into micro LEDs a little bit. What fobian did was they had, they figured out that if that silicon would act as its own filter. So what they did is they put photo diodes at different levels of depth in the silicon, they had basically one depth for blue, one depth for green and one depth for red. And what happens is, is if you put those diodes at different depths in the silicon, it will naturally filter the thing and it's not as good as using clay color filters. It doesn't give you the separation that you'd like. So what you then have to do is apply a bunch of mathematics and stuff and try to extract from the signals that you got from those three layers, what the color was. So what that is what it did was it gave you full color at a pixel whereas famously, almost all cameras today use a some form of color filter
are. So they have a red, a green and a blue filter, oftentimes what they call a bear pattern where they have two greens, a red and a blue. For every four pixels, that's the most common one. So they filter it out that way, but they get real good color discrimination that way. Plus, they don't they don't have because the layers are not stacked on top of each other. Each photo site gets to collect all the light. So yes, you do get the light filtered out. But you don't have the issue that light didn't progress real well through silicon silicones, fairly opaque. So it was absorbing a lot of the light. Plus, you didn't get to get as big a photo site because the photosites are also getting blocked by the layers on top of them. And as I said, this ties back a little bit into micro LEDs today. And one of my issues with with the we know that Extendo and we know that recently, there's been papers from MIT about doing stack micro LEDs, well have some of the same issues. When you stack micro LEDs, how do you get the light from the lower layers up? So it's kind of like going in reverse? In the case of the fobian sensor? How do you get the light from the top down. The other big, big problem though the fobian had was that discrimination because the silicon filtering wasn't, didn't give you a lot of discrimination from one color to the next, there's a ton of overlap in the spectrum from the three sensors. And because of that, it gets kind of difficult to really tell the colors, they get some really weird color a lot. Sometimes it worked out great. But a lot of times it didn't work out the bigger problem they had was it wasn't going to scale, they had this weird process for making it. And this is something I learned once again, from my days at TI I was helping people may know I was heavily involved in the thing called the video RAM. And all we did was add a shift register. And later there was a basically a strip of RAM along the edge of the die. And it added maybe 5% to the die area. It turned out that drove the cost up by 2x. And you might say How does 5% drive up the cost to x? Well, it's a lesson in how you manufacture stuff. Because what happens is, because it was different than the super high volume drams, they didn't have the what they call a bit banger testers, these massive arrays of testers, they didn't have the degree in which they would do the development and the reiterations. And the mock mass changes and everything in the infrastructure you did for this massive quality of DRAM. So the problem that the synchronous or the synchronous video, RAM or the video RAM had was that they we did not have the volumes to justify. And if you can't get the volumes, you can't get the price down. If you can't get the price down, you can't get the volumes up. And it's just a vicious cycle. So I learned from that, that even a small change in a device has a can have a big impact on price. And I applied that when I got into the fobian discussion, I said, you're now talking a very, very special unique process, you're not just adding a little bit on the side of a dye. You're trying to have a completely different semiconductor process with these three layers of stuff that you're trying to develop. And yes, it can be done. You can do it in a lab. But can you do it at scale? I was also very concerned about the discrimination to that people got it. It's kind of like this one factor of people talking about a single factor analysis where you just worry about one thing and don't worry about the whole problem. And that's what basically fobian did, they worried about this single factor, which was getting rid of the color filters and had a lot of bogus arguments like oh, the filters will die. Well, it turns out with photography, unless you're aiming at the sun, you're not going to get that kind of bright light on a filter that would damage it. And they had some spurious arguments that argue against that stuff. So I got existent baits online. Anyway, my debates with that ended up in this book by George Gilder called the silicon AI. And so I was calling out he was kind of pro he was very pro he knew Carver Mead was the one of the big backers behind the fobian. And he was kind of the book is very pro fobian. And I was one of the dissenters that he cited in the book for my discussions on DP review, bringing it back to dp review. So anyway, bring us up to about the 2011 era when I got back in my boat actually before that when I was Asendia. I was kind of the company photographer as well. And one of the things I learned the cheap cameras, you really can't control all the things you need, like we use field sequential color. And I applied this all the time when I'm taking pictures. People don't know people think I'm getting nosy on most of other technology that's like like with micro LEDs we know they do pulse width modulation. Well if you shoot a micro led at a high shutter speed, you're not going to get anything you may get one pulse you may not get all the pulses captured in the case of field sequential
hollered that we use, use an L cos and DLP uses. If you catch her too fast, if you have two highest shutter speed to faster shutter speed, you may only catch one color. So the colors are going to be really bad. So each of these technologies has their own kind of quirks and issues because the camera does not work like the eye. So a lot of what I had to do, and this was back when I was at Sandia, and then since is trying to figure out, okay, this is the temporal or the time based way this technology is working, how can I get the camera to replicate as close as I can to what the I see. So that's, that kind of happened through the 20, elevens. And all. And then, a few years ago, I kind of fell off, I was doing photography, most of my photography interests was just in filming, shooting things for the blog. And somewhere along the way, I heard that this mirrorless camera stuff was really taking off. Because we went from DSLRs, where you had a literally a fold mirror, you flip a mirror up and expose the sensor. Well now, essentially, all new cameras are mirrorless. And by mirrorless, it means like unlike a DSLR. By the way, single lens reflex basically means that there's a single lens and a reflex mirror mirror that flips out of the way that's where the SLR comes from is single lens reflex, that mirror flips up out of the way as you take the pictures and we hit the shutter button, that flop you hear is two things here, the mirror flopping up, and then you then the shutter goes, that's what happened on a DSLR. Well now on a mirrorless camera, the shutter is still there on a lot of them, and that's starting to go away. But you no longer have a flip up mirror. That means you can make the camera smaller because you can move the lens closer to the sensor, you don't have that mirror in the way. And so anyway, I decided I started to get back into photography again. And I started looking to start so I started participating pretty heavily again a few years ago and DP review and I bought a lot of fairly good high end, Canon based equipment. That's mirrorless. So I've got lenses and stuff. When I started photographing a lot of things. My big thing was to photograph world war two airplanes, I've really liked doing that. Unfortunately, I also was there for the accident that was here in Dallas last year, which is kind of sad, very sad. So I got back into photography, again, somewhat from the blog kind of like came up and down. I got into the technical aspects, particularly time Bayes dealing with all the various types of thing I shoot. So invariably, I had the same thing just at the latest shows I was at of asking people, well, what kind of what are you doing? Like even when I asked prickly with micro LEDs, most micro LEDs use some form of pulse modulation. To control the brightness, they use pulse width or multiple pulses. So you kind of got to know what that time period is. So you can average the image over in a frame. So I think that kind of covers that. But anyway, I'm really sad to see DP review go. It's also a classic lesson, in the problem of having a French effectively one company only entire thing, because we're going to see a massive wipe out if it goes through your some rumor that somebody might buy our DP review. But I don't know if Amazon will want to sell a lot of times big corporations, when they decide to cut cut their losses. They just say to me just throw it in the trash. They don't want anybody touch a lesson they want is somebody making success out of something they said wasn't worth it. So I'm really scared, it's going away. And it's really sad because there's a ton of information, a ton of history is going to get wiped out, we'll see if the Wayback Machine or anything captures that. And then there's a ton of just the knowledge. It's a real hurt on the whole digital photography industry. And as you may know, and this is a nother, I guess a little side story here. But
you know, digital photography, they were at the beginning of it really digital photography took off in the late 2000s and really took off after 2000. But then it kind of had the double edged sword of the cameras and phones coming in because what happened is they grew like crazy from about 2000 to 2010 12 that in there, it grew like crazy, but in 20 by 2012 but 2007 or eight the iPhone comes in and with the iPhone and other smartphones coming along, photography flipped and now most people already had a free camera built into their phone. And so cameras that are not kind of the higher end are going away. So really the cameras that are bought today are mostly slightly higher end they're either what we call blogging cameras, which are kind of all in ones but aimed at video. And then you have the professional and serious amateurs I guess I'd be in the serious amateur class. But that's that's us left doing camera photography. Now everybody else is using cell phones. our smartphones. So that's kind of the way thing that go, but it's an interesting kind of insurance, see our markets and stuff go up and down and left and right. And things you think will be there forever. Like I say, if you look digital cameras and 2010, you think they were going to the moon, but really, they topped off in about 2012. And then they slid for a while and they've been going down. And we may have hit another plateau where basically, you shouldn't lose, lose all the people that were that would have were captured by the cell phones. And now we're kind of seeing maybe a little flattening out with the serious internet. But there's not much. There's basically no point in shoe market left. You know, if you want a small little point and shoot camera, there's basically no market for it, you buy high end cell phone or a cell phone as good good photo features.
At the time, the phrase was the best camera, there is the camera you have with you something along those lines,
but you can still do things, you know, you can fake it, you can try to fake it with the cameras, there's no substitute for capturing more light with a bigger sensor like I went to full frame mirrorless that was my big thing is I never I never bothered to go full frame before I stayed with APSC for 20 years. And then I went to full frame, you can just do things were full frame and with the big lenses and expensive lenses. The problem is nowadays, I mean, if you want a decent kit with a mirrorless camera, I mean I have a fairly high end I have an AR five which is a almost a $4,000 camera, just the camera and then you're talking really high end lenses cost two and a half to 3000 apiece. So if you want a really your camera kit, you can be 10 or $20,000 into the camera. If you want to do it as an amateur, they're trying to come back and do kind of an entry level products that are APSC again mirrorless. But you still can be talking a few $1,000 to get there. Now that may seem crazy. But remember when I bought I spent that much I figured out that a Canon our five I bought which is a 45 megapixel camera with amazing capability had does up to 8k video 4k, really high quality video, all that stuff. With 45 megapixels, adjusted for inflation cost less than a D 33 megapixel D 30. Did in 2000. We are spoiled. And this is another problem. This is a problem that affects AR too. We are really spoiled by how cheap technology has gotten. And that what would have cost. You know, I always use the example of television set televisions used to costs hundreds of dollars and adjusted for inflation, a 23 inch which was big screen back then a 23 inch color television adjusted for inflation would be like 10 or $20,000 today, and you just people are so spoiled by it. And this will affect the car industry too. Because it's tough to get people to plunk down that money because they're so used to paying so little for the technology they have that it's a little hard it makes it actually becomes a barrier to entry for new entry new technologies to get in the market. When you say it's going to cost two or $3,000. Their expectations are very high and may not be fulfilled double with with new technology.
Let's jump on that one talking about this kind of expectation shifting over time. Apple in the smartphone industry has been a huge contributor to that right just the volume of sales of these devices. As you noted earlier, the volume drives down the prices that allows these things to be more accessible. And then, early in the smartphone era we had a lot of subsidization that was happening by the telcos themselves who were paying the cost of the phone effectively so that you would come and join their network so it was further suppressing the cost for us as a consumer and you know that $3,000 is a price point that today we have very high expectations for go out and buy a smartphone especially when switching carriers for very little and Apple thinks that that sort of $3,000 Price least that's the rumor is something that is palatable to consumers of their rumored forthcoming VR device. And beyond that we've heard and know from from insiders who work in these projects over the years that they've been working on AR is the next thing beyond the smartphone because today the smartphone is really all about iterating on the camera more than anything else. But anyway this this vision that Apple is professed that AR is gonna be the next thing after the smartphone. They've been working on these projects behind the scenes. They've organized VR projects that only mixed reality. Anyway, the latest rumor that you talked about in your blog is that Apple is abandoning the canceling their AR glasses project
and not canceling I think the words were delayed indefinitely
delayed indefinitely Okay, so they whatever rumor they're going to have about when it was coming is now gone is just that it's going to come whenever they feel like it's ready.
Yeah, and what gets me in and I've pulled this I've talked with Bradley Lynch on sadly it's probably about this and other videos I've done I talk about it. It's hard for me to under Stan, because I've always felt like both Apple and Mata were of the same mindset. At least this is the impression they give that VR was not big enough to send a thrill up there, like VR is a finite size market, I oftentimes say that the problem with the VR market is this real, but they know how big it is. The beauty of the AR market is it may not be real, but they don't know how big it is. So VR is seen by most people in the industry. And I think the powers that be at the big companies, that it's a subset of the gaming market. So while game market itself is kind of an OK market. And while VR has a very dedicated following, they're a little bit like, going back to our analogy with high end camera P. You know, I'm in the high end cameras. Well, you know, there's a market there. And I think that market is not going away. But we found out how big it is. It's not as it's not, it's a sub segment of what the digital camera market used to be. And it's a probably 30% of what the digital camera Mark was. Now it's profitable, because people pay a lot of money for very few cameras. Well, that's kind of the same analogy into the VR market. The people are really into VR, because there's a lot of talk that was what they call a closet where now I think it's called, you know, men have dropped the price on the metal lead to and dropped the price really low. And that caused a surge in activity. But since then the sales have actually dropped. I heard last year, VR sales dropped about two or 3%. And part of that they think is this called closet where where basically people are buying stuff. Oh, yeah, that was that was fun. I tried it. Yeah, one to go. So they put it in their closet. And then once I think sitting in the closet, the next time you want one guy says we already got one of those sitting in the closet. So I think Apple apples got to know that the market is only so big, at least that's the impression how now Apple can make it go, you know, Apple just by their sheer weight of numbers and influence can drive a higher price. And can can do more. But I don't know they fundamentally changed the market for who wants it. It's not what we call a four corner market. This comes from my days in Disney. But the ideal movie, there's nothing done better than say Harry Potter is a four corner market. It gets to kids, teenagers, adult and even old people my dad used to like Harry Potter when it came out. Okay, so that that covered everything from his grandkids. To him. We're into Harry Potter. So that's a forefront that covers everybody and covered male and female too. So it had had all age groups and and male and female. The problem with AR is reality is it's heavily slanted to kind of
you're talking about VR here,
VR, sorry, VR, VR slants heavily towards teens, to young male and young and male. So that's a very small, that's the problem that has that nobody has figured out how that category bronze out, there's probably a little more female in the market than there was. But it doesn't broaden extremely well. I mean, most of the people you see that are really into it. It's still weighed heavily to male, and basically teen teen to young, young adult market. That's a hard market to cut. Yeah, you can get there you can get your money. But there's that kind of Gi the kind of market that Apple enjoys. You know, if you're a company and you got that market, you got a big share of that market, you're doing well. But if you're an apple, it, it's a nuisance business. There's a look going back to Apple and DEP review. It's not clear the DEP review was really losing money, but it was a nuisance to Apple. So get rid of it to Amazon
to Amazon this case, Amazon sorry. So it's possible that Apple has a different set of use cases in mind, right? Today, VR is very much focused, as you noted a subset of the gaming market. There's also this kind of other enterprise angle to VR, and which is viewed as a training tool and even educational tool even in school systems. But from the from ride Consumer Perspective is very much kind of a subset of the gaming market, as you describe. Maybe it's the case that Apple is able to expand that that interest level by introducing different use cases around how people might engage with these things. What do you think?
I think it's tough. I think they I've said many times they have to live with the same physics as everyone else. And you know, even the best of the AR headsets, even the three MDR $300,000 To live first if you're Apple unless you're subsidizing it. You know if Apple wants their usual margins on it, they don't have any room at all, because they usually take higher margins than anyone else. But even if you you say they drop trial on margin and they just say I want to make this market happen. It's tough because the physics there's a lot of things with the AI the human vision side there think people that just get nauseous and sick with with VR and I doubt on Apple is going to fix all those things at a $3,000 price point. I haven't seen a $10,000 price point that solves all those things. And so I think there are issues there human factor issues with VR. And when you look at other applications, like training and whatnot, and this is the thing I say about a lot of so called Enterprise, as soon as you say, enterprise, they say, yeah, there's a real enterprise market, it's real. It's about 200,000 units a year. It goes back to the problem of VR, it's real, it exists. I believe there ar headsets are some good markets for AR headsets into the enterprise market. As you know, we've seen HoloLens sold some into that, and all but the problem is, is how do you get to the millions? How do you get to something that sends a thrill up a, a Tim Cook or a Zuckerberg flag? And that's where the problem gets to be is how do you get to the kind of ions that really move the needle?
Why do you think then Apple invest so much rumored rumored to invest so much time money effort into a $3,000 VR rig?
That's that? That's the billion multi billion dollar question. And I don't know if they were defending it sometimes, like, if you're an apple, sometimes you do things just that if I call it defending the fortress, you just say, You know what we cannot afford to lose. So we're gonna keep the motor running, we're gonna if if Matt is going to go into this with billions of dollars, because we know was, what's Metis spending, like about $12 billion a year on this, if Mattis spending 12 billion, you can spend some amount on it, and what I call cover the bat, and just say, Look, if you're apple at their size, you can spend billions of dollars a year on something even if it doesn't pay off. I'm surprised. And the rumor, the real Roemer that floored me, because I'm on I'm on the record on video saying, I could see Apple, I could see Tim Cook at some point saying, I could see him saying to his investors and whatnot, we had the absolute best, it was fantastic. But it just wasn't Apple enough. And I expected that to happen. Now what I'm hearing. I mean, what we're hearing, the rumor mill is that that actually the top management is pushing it to go out that the engineers are saying it's not ready. And that the management wants it to go out. And that's what I sometimes call was it. It's called escaping the lab, or basically management says, I put billions of dollars into this, go to market with it, for damn sake, either put up or shut up, go out there and prove to me that I should keep spending my money on it. And so I've seen that happen a few times. And almost always that's a colossal failure, because it's not ready. I mean, the engineers, if the engineers really don't think it's ready, it's very unusual that engineers think something is not ready. When it is, if anything, they're anxious to see their life's work, get to market you, you're you're more anxious to get it there. And sometimes you get things they're too early and they they die because they get to market too early. So I'm really kind of surprised. But I have seen that before i ti. The home computer was an absolute fiasco were famously are eventually brought down the CEO of the company, J Fred Busey, he put on in a home I got involved, I was involved a little bit with the home computer, I developed the graphics chip that went inside it. And if we hadn't canceled the program and had the CPU, I would have been the chief architect of the CPU inside the home computer at TI. But it was a terrible design. I knew that everybody in the company knew is a horrible design, except for maybe a few people who built it. And Busey thought this was the future it was going to be the next calculator chip, not realizing that it was a really bad design. It wasn't up what we had done. It was just a horrible design. I could go I could spend hours say, explaining what why the to home computer was at Amazon, as but it was best expressed by they took the best bag of parts and put it together in the worst way. If you had taken the same components that were in the Tiaan computer and built it differently, they would have been a pretty good product, but have been competitive with Apple. But the way it actually was been better than the Apple two. But at that stage, it was better everything. The problem is they the way they put together was absolute abortion in terms of how they held the but there were bottlenecks they built into the didn't have to there was just all kinds of mistakes they made. But he's really put together that the actual design of the computer itself was terrible. Anyway, so bringing it back to Apple, will they you know, I mean, they've got trillions, a trillion dollars in the bank every day. They can do anything they want. The question is, will they you know, are they really going to develop this way? I mean, I'm used to the Steve Jobs who, who holds the phone up and six months later, it's shipping, you know, the original iPhone up and And then it's shipping and then it's shipping in the millet. I'm not I don't know, this apple that that puts something up there and says, Well, we're gonna do a kind of a developer vehicle. And
it's a different different states in the market. Because when Apple came out with the iPhone, the mobile phone industry was already around for a long time. They're already two and a half billion mobile subscribers the moment Apple released a product, the two and a half billion people who could already who have already said yes to a similar device. And at that point, there's already 100 million smartphones being sold a year, that market existed in volume, the moment Apple put that product out. Whereas today, the VR market, as you noted, is not necessarily growing. Last year, the room was actually contracted a little bit there not hundreds of millions of users who are already buying an equivalent product that applicants Hey, look at ours is better.
Yeah. And really apple. I mean, Apple did a few things better. The main thing they realized is that web browsing was much more important. People forget that I had I had a blackjack, which was sort of a Samsung version of the phone, BlackBerry, Blackberry. Yeah. So um, you know, it had a camera in it. I could web browsers on it, not very well. But a web browser. The main thing that Apple did was they said, touchscreen is ready to go. And a lot of people actually it was barely ready. If you go back and look at how that happened. There's a very good video on what's it called Magic was magic. No magic pixie was the company, General Magic. General Magic was spun out more or less spun out of Apple, a bunch of Apple engineers left and developed General Magic, which, if you sit there's a video out there, that's great that talks about how they how many things they anticipated. And they've on record on video, in the day talking about these things that became that eventually became that Apple adopted into the iPhone, but but really what they did was they brought out the touchscreen. And they figured out that the touchscreen really wasn't ready. But they could figure out algorithmically to kind of just test keystrokes. So in other words, they kind of could infer from various keystrokes, what you really meant to type. And so they was kind of self correcting, even in its day, in early days, to be able to make a touchscreen work. But they figured out the touchscreen, and now all of a sudden, they tripled more than tripled for the same size phone, the display size. So all of a sudden, now when you have triple the screen area, you now can start getting a little better and serious about your browsing. You know, training that keyboard for the display or in one on be able to make that dynamic really opened up web browsing on the phone and really flip the phone from being a web browser device. But people forget, almost every decent phone in that day had a camera in it. They had web browsing in them, they had all that kind of basic stuff that wasn't invented by Apple, what they did do is build a much better design. And they kind of dumb and they also realized that they stepped off of there. The iPod many people forget the iPod, the iPod was not the first mp3 player, they were a very late entry in mp3. The big thing they did was they figured out how to skirt and deal with the with the licensing issue. Sometimes it's the secondary things. By the way, it's not the the obvious stuff that gets you but their big thing was to figure out how with iTunes to get around the licensing, okay, you got to pay people a little bit. You can't just steal it from everybody, you know, so they kind of figured out via the iTunes setup and how to make the software more workable, less of a, everybody has to engineer their own thing. That's also a good lesson, the AR market in the VR market. I think we're doing too many things that force everyone to go from scratch. I see this analogy. Even in I'm talking I'm actually writing an article now dealing with all the AR headsets I saw early this year, including the Argo. Well, the Argo, even though it has enough eye relief to support you wearing your glasses, still requires inserts. Well as soon as you require prescription inserts, you just cut your market by by huge fraction by huge amount. Because and the even though they had enough eye relief, they could have done it but because of other design considerations. They didn't so now you can't now you have to have inserts, so you have inserts, you made it problematic as an enterprise product. I mean, if you get down to one thing, HoloLens two did right, is that they allow you to wear ordinary glasses and because of that, you go to a you go to a museum, hey, you could put one on because you don't you can wear it over your glasses. You don't have to sit there and say well, what's your prescription and whatnot and prescriptions are really much more varied than people think they I mean the only thing that they provided the shows are diopter adjustments. Well I don't need that after adjustments. Matter of fact, one of my eyes is got a different doctor than the other quite a bit. But I mostly have a stigmatism well they don't fix this. They're not going to deal with this segment as a mom on a set. So there's too many. There are barriers like that you can build in and if you unless you make it where it's really easy to accept and easy to adopt. up, it's gonna be a barrier to your product, making it to market. So, but yeah, anyway, back to AR You know, it's really hard for me to understand what Tim Cook. I always thought that the AR market was the reason to exist. That was the reason that that that's a reason medicine I've always believed that meta bot bought Oculus with the idea was this will be our learning ground to get us into AR. Because no matter what you do, you can't get the VR market to be big enough. It's not four corners enough. The theory is that hey, with glasses, if we came up with the glasses form factor, then it was somehow replaced the cell phone. Now I don't think that's happening for a long time, if ever, but it's a really hard ass. Because there's people who don't wear glasses. And B, do people really want to build all that technology in the glasses that they buy, like I get new glasses about once a year, every year, am I going to spend all that money redoing it? How am I going to do all that? But that's the vision. That's the thing. But anyway, in the case of of Apple, if you've indefinitely suspended or definitely delayed your AR effort, then why are you doing VR at all? It doesn't, it doesn't seem to add up. What I always like people to do is I need I need to tell me two numbers, everyone will sit there and say, Oh, here's this application. And then I say okay, well how many units? Is it? Let's get a guesstimate of how many units that that represents. And that's where a lot of these things fail. You can sit there and say yes, there's a real training market. Yes, there's a market in museums, there's a market, on factory floors. Amazon delivery has a market amaz all the warehouses have a market, there's a lot of things, most of the AR in this world that I that I see as successful involves him free. That's a really big critical factor that has his and free thing. And I mean, you could justify AR, if you sit there really easily you just sit at the elevator speech. I always said for HoloLens two was look you okay, it's $3,000 with full loading, and everything else is four or $5,000 with software and all the other goodies we'll throw in. So what you do is you go and say yeah, that that person on the factory floor working on at Boeing working on an airplane or at Toyota working on a car that cost you over $100,000 a year for that factory worker, it salary benefits, equipment, space, liability, insurance, health care, the whole thing is way over $100,000, if I can make that guy 10% more effective, just 10% You've paid for this thing in a few months. And then it's paying dividends all the way. So the value case for AR in the business and industrial and enterprise environment is fairly compelling and fairly easy to understand. You can put a number on it, you can justify it. The problem you have is saying, Well how many of those are there. And that's where you get to this. I think the numbers two, maybe three or 400,000, that the outside it's not millions or billions. And when you talk in cell phones, you're talking multiplying by millions or billions, you know, by bit by literally hundreds of millions and billions. And when you're talking that it's hard unless you get a really high number for it, and you just don't have that kind of number. So you know, I say you know a lot of people in AR might be better off doing like, heads up display, Leisha, you have a potential for one per car
may even come back to heads up display. But this notion on this class of devices that is kind of emerging on the back of VR, and the thing that animals rumored to be working on, which is a video pass through device, this idea that you can mix in this idea that you're immersed in the video, you know, whatever the VR experience happens to be, alternatively, or Additionally, you can then mix in the camera feed from the real world. And so you have these mixed reality like experiences. And I've heard at some state that this type of this class of device could even replace the HoloLens, or the magic leaps of the world. What do you think is the case for video Pasto? And does it remove the need for for a higher end AR? Or high end? Seethrough sort of AR device?
Oh, I've heard that. And I'm fairly friendly with links and Sam Laroque. And he's made that contention very recently, I think maybe on your shop.
I think he did. He did. Dan made that connection on my show. So
yeah, I'm very familiar with that argument. I liked Stan at all. And he makes some good points. But I'm always saying and I've always said that there's that while AR and optical AR and AR passthru which is another forum mix are also called mixed reality. But mixed reality. For VR where you do camera pass through AR they have almost equal and opposite problems, almost everything that's easy with optical AR is next to impossible with pass through AR. And everything easy with pass through AR is next to impossible. With optical AR. They're almost a dissimilar set. And it's somewhat a grass is greener, you sit there and say, Well, I saw this, this and this. And I said, but the list of things you made worse are all over here. I think Stan, on your show was a little flippant or little under anticipated all the other difficulties. And examples are things like, I mean, the obvious one is delay. Yeah, you could work on the delay, you could get the delay down. That's one issue. One of the other things you have is is focus. On the case of the VR world, everything tends to be fixed focus. But you got to remember, it's not just the focus of the display of the display to your eye, but also the focus of the real world to the display. So how do you tell the camera where to focus? Or, or what do you do because when you look your eyes then change their focus, we've got to then see that with whatever your technology is respond to a quickly enough by the way, the ice jitters we call skating, or what's known as the Cades. So the eyes constantly jittering, the way you see is you actually arise or jittering all the time. So somehow you got to figure out where the eyes looking and make the camera that's sitting up in front their focus. The other problem in standard talks about this, I think what does optics, you know how you had to make us optics, then there's a really good paper I come along was 2013 2015, somewhere in there by Steve Mann on IEEE, it's available online. And he talks about the importance of centering the camera over the eye. Turns out that your brain is pretty sensitive to how your eyes work versus what you see. And so if you're going to put a camera out there, the real world, you really want the camera, absolutely centered on the eye, and you really kind of want it angling where the eye angles, you really want it, you have to kind of compensate and mimic that matter of fact, when you look at the medic quest Pro, you can see they're doing something really bizarre where they take one color camera, and then to kind of what are meant for be tracking cameras, but they're there their infrared cameras, and they map the color camera onto the two infrared cameras. And they're trying to create a synthetic image that's kind of in between all those cameras. It doesn't work too well. It's all real wobbly, and everything is terrible looking. But you can kind of see what they're worried about, which is what you really have to do is you really if you had enough cameras up there, you could then try to start to sense the size and image where the camera sits over your eye. In other words, imagine if I took a few cameras up there, put that all together. And now I tried to synthesize an image that looks that actually kind of responsible your eye. In other words, your eyes move it responds the same way. But now you got to process all that you got to take all that in, you got to compute it and not take too long doing it. So the problems kind of compound. When you do that. Besides that, even when you look at like the Lynx product, you still block out a lot of peripheral vision. So you've got a lot of things that don't work very well, it's just like, remember, you're putting us all up against your AI, and you're kind of fighting the human mind. And this turns into nauseous and headaches and whatnot. Because at some point, the brain says, Screw this, I'm out of here. It hurt, I'm gonna, I'm gonna send some pain to you to make you know that stop doing bad, I'm gonna send you some pain and make you stop. And that's kind of what goes on with this in a light hearted way. So and by the way, I'm gonna be talking about this at AWS, I'm presenting at the end of May at AWS, ie, this exact thing of the pros and cons of AR, and I tried to give you a fair set of you know, okay, here's all the good things and your all the bad things. And I don't think they're, they're there. And you could you could hypothesize that someday, I come up with this infinitely high resolution light field camera. And I have an infinitely high resolution display or high enough resolution for the eye display in both cases. And finally, I make all that work with all the processing to make it work and whatnot. But it's not anywhere in the next 20 years. So I'm a little skeptical, I think. And maybe I'm a little too down to earth with this. But I think there is a market for VR. I think what VR does well, it does. Well, I do see a market for pastoral AR, but I consider it augmented VR. If your goal is to find your mouse, you know you're doing VR, there's a lot of reasons to want that camera fate pass through for a little bit of safety and whatnot. Although, you know, as any of us who found who've had an effect, I think, I know, Bradley and Brad Lynch talks about what is it the VR to the ER, basically you forget about it, but you know I've had the headsets on and as I say, you can put a boundary around yeah, I'll put that little fence God that fence around you. And if you set the fence tight enough, there's no boundary that works. Because if you set the center tight enough, it's all the time giving you warnings and stop and up, up up and messing you up. And besides which, no matter how wide you set it, there's still a bookcase, you go like, you'll do this thing where you reach your hand real quick, and you bang into something, you smack your knuckles on something, because you're you reacted faster than the geofence good, or the fencing could get you. Or if you set the boundary real wide, well, then you just run into everything. So there's never really a good balance there. I don't see how any, I can't see a lawyer. Let's put it that let's bring it back to lawyers, I can't see a lawyer's putting, I could see a lawyer allowing an AR headset on a factory floor has to be very transparent, I don't believe you're going to get away with something that blocks 80% of the light and stuff, you got to be fairly transparent. Just give you a little bit information not getting your way much has to be very biased. I don't see that happening with a VR headset, ever, I think you have to be in a very safe environment, you have to be in an environment, the padded room. I mean, if you're doing if you're going to put a guy in a workplace and that environment, you're not putting him out on the factory floor with machinery that could kill them. You're not going to put a VR headset on that guy, there's too much liability. So I think the lawyers have to a bit. So anyway, but that gives you kind of feel. But yeah, I think that the there's there are real markets for each. The question is, is there a merge area? Between the two? Is there a gray zone between the two? There may be some like if you know a friend of a military training, you know, there are things you could do where you say okay, in the classroom, I might wear of a VR helmet out in the field, I might read an era helmet. I mean, I don't think HoloLens was decades. I know it was an insane program. The whole is HoloLens two program that was done with the army. But you know, their real use for that in training. But the thing that you're going to do Landa troops with that was insane. I think that was the height of it. That that was anywhere near ready to Atlanta troop to put somebody out there with spotlights shining. And the way that thing front projected with the fragility of it with the weight of it with the fact that you couldn't hold military gear properly. There were so many things that were fundamentally wrong and work near being weren't going to be fixed in 10 years. But if you wanted to do it for training, on the other hand, military have been using HoloLens one for years in training, and it makes sense, okay, you want to, you want to train with something that's a different game. And that's what you got to think you've got to do with these things. You got to sit there and say, Is this really realistic? That is a business you have to multiply the numbers out, you have to say, Okay, that's a real app. But is it is it a big enough volume that it's going to pay for itself, you know, that you're gonna afford to do it.
My belief is very similar on this notion of putting a camera on a VR rig, and doing video pass through. To me it ultimately creates a better VR product, a safer VR product. And one of the things that I really like about what Lynx is doing is that they copied one of the things he said HoloLens got right with HoloLens two, which is the flip up design, the flip up visor, like
the flip up design they did, I was asked this, I think by Brad or somebody who had the best, I mean, you could argue their image quality is among the worst of all the VR headsets. But they did do some things right, in terms of making a mix reality or try to make a mix reality. The flip up is very important because you've got to get it totally out of the way I still does. Don't think that let you go loose on a factory floor,
you move when it's flipped up. And then when you're stationary in front of some safe area though maybe it's a caution tape area zone safety zone, you can put it back down and then engaged with your overlays and I think you're
going to find that permit is going to get sloppy and and lose an arm and then that'll be the end of it. Sorry to spoil sport, but I just it's like I can see the end of it. It's kind of like, way back when when I was a TI I played I'm very naive. I'm very wholesome, you know, family type person and whatnot. So I wasn't brought up to do some of the things that our current politicians do. But But I was I was meeting with a big toy company I was trying to get my 340 into a toy and game this was back in that dead period between when the whole you know, we had clique, a vision and and all the toys that failed. All the video games that failed that I felt there was a video game market. I was trying to get the 340 family into a video game market. I was talking to a guy who worked with Hasbro and we had something kind of going down this is before Nintendo hit the market so we were trying to beat the the new Nintendo stuff out are actually the intent, the intent, the original intent of it out there for you had the 3d stuff and the better the newer, that kind of stuff. So we were trying to hit that market with a much better game console using the free 40 family. And I used to talk about how Oh, graphics is good, all this stuff is good for stuff and graphics is all good and whatnot. And he was kind of like, well, you can turn everything into pornography. So I was like, and these days, I'm even more a believer, because I see how see people's we know all the problems with social media today, and whether it's causing the youth. So anyway, let's go on.
Let's shift gears a little bit and kind of flip to the other end of this AR VR spectrum. And here we're talking a bit about this video pastor on the VR side, we're talking a bit about the the opportunity that HoloLens and Magic Leap had with a really fully functional see through AR devices. But the other end of that scale, are glasses truly wearable glasses that incorporate the sort of overlay that AR promises on the real world. But one of the big barriers to having smaller, lightweight AR glasses is small and efficient displays. And the technology that we've talked about. In the past. It's been around this industry, in the hopes column for a number of years and now beginning to be maybe on the commercialized column is the is micro led inorganic led micro LED technology. And maybe you can start a little bit about sharing your perspective on what are the hopes and the challenges of micro LED.
The conversation with Karl continues in the next episode. We talk about the current state of micro LED technology diffractive versus reflective waveguides historical comparables, if or when AR glasses will replace a smartphone, which technologies will win in the mid and long term, and many other topics. If you've enjoyed it this far, I think you'll continue to enjoy the conversation and please consider contributing to this podcast patreon.com/theARshow. Thanks for listening