I hope to be wrong, but secrecy like this usually doesn't bode well. Best case, I'm expecting something like Segway or Apple Watch -- overhyped from secrecy, but legitimate, but not all that amazing. Middle case is something like Theranos; not overt fraud from the beginning, but the result of a lot of pivots and coverup. But most often it is outright bad.
Short term practical or tactical secrecy, sure. And in some industries, probably not as bad a signal. But software-heavy hardware ecosystems aren't one of those.
Outside of wars, Ultra or Manhattan just aren't that good a way to work. Why cut yourself off from the global community?
I was, at one time very recently, very deep in the VR/AR space. It seems to me that there will be quite a bit of time before people accept projectors pointed at their eyeballs (or however you want to explain Magic Leap's "digital light field" technology).
$800M on $4.33B post- is a big chunk of a company, particularly one that does not have a working product demo yet. Maybe it'll all shake out positively, but I am of the opinion that this is not real.
Lastly, all of Magic Leap's demos are very clearly not using wearable hardware (and also not being demonstrated using any kind of light-accurate reproduction). For example, metallic effects usually require "black" as a shade and things like the Lumus Optical lenses cannot reproduce such an effect (despite being basically the cutting edge of lightfield reproduction).
In short, I am bearish on Magic Leap.
Also, I am super confused as to why Magic Leap is operating their own silicon fab... That seems crazy to me.
There is some kind of irony in the fact that an augmented reality company seems so successful in presenting investors with an image that nobody else can see.
> Maybe it'll all shake out positively, but I am of the opinion that this is not real.
But after all, isn't that the whole point of it? :) I wonder what kind of stuff they managed to project in the eyeballs of the VCs.
I see a bunch of negative votes so I'll expand as I was not trying to be facetious or sarcastic.
The first magic leap demo I ever heard of (and, because I'm not under an NDA with them this is all hearsay) was a cup of coffee on a table. Put $WEARABLE on, now there are two cups. Which is real?
I can easily imagine that, were this convincing visually, VCs would surrender their wallets with aplomb.
It'd be fantastic if it turned out that the Magic Leap demos were all just magic tricks set up to make it seem like it was really advanced VR technology.
Ex: in this case, when you put the glasses on there's a moment of not being able to see anything, and they just put another real coffee cup on the table.
So I saw this, and while I appreciate the conjecture, the narrative sounded unconvincing to me. I think that when we start talking about computational neurobiology, it's hard to sufficiently bound the problem such that computers can model it. Computers are very good at doing stuff that humans know how to do (and repeating it a bajillion times), but it's hard to get computers to do stuff humans don't know how to do (one of the recent interesting expressions of this that is counter to this point is neural networks, but I digress).
When he talks about "we made black with light" I feel like someone is playing "hide-the-ball" because I think of black as the absence of light. Perhaps I am just not modeling the problem correctly, but I did not take away anything useful from the comments in this video :(.
Benedict Evans works for one of their investors (A16Z) and has apparently had the demo. If you follow him you know he's not prone to exaggeration and guards his reputation carefully. Here's his take on the demo:
Michael Abrash wrote article about why hard AR is unlikely any time soon. It's not only that you need low enough latency for the human visual system, but also the ability to draw black (which is an unsolved problem).
I'd be surprised if Magic Leap was actually able to solve both of these problems in secret separated from the rest of the world.
Secrecy - for this long seems bad. It's also not the Apple secrecy where you say nothing and then have a big reveal. Magic Leap has put out teaser trailers with no information and talked about their 'revolutionary' technology without details. That doesn't inspire confidence.
I am surprised they're able to raise so much money though, but that's a mistake that's been made before.
I think it's funny how much science fiction inspires current technology, but our expectations for quality are so much higher for real products than their sci-fi counterparts. In fact, in sci-fi advanced technology is often made inferior to be more relatable to current viewers/readers.
For example, the holograms in Star Wars: They are made to look lower quality than the world around them, so it's visibly clear this is a digital projection (and not teleportation or a wormhole or something). But could you imagine if Apple released a holographic "FaceTime3D", only the holograms were monochromatic, filled with scanlines, visual noise and crappy audio quality? Reviewers would blast it, and a poor experience like that could even do more harm than good to the popularity of AR.
Star Wars, however, has a run-down, lived-in aesthetic. That was part of its appeal: it was futurey, but it looked old. Star Trek does not have that kind of aesthetic, and its holograms do not look the same.
I think you're underselling the technology. The hologram appeared to be a small attachment to a multi-purpose droid that was recorded off of a single camera. Being able to record and project holograms on your phone would be extremely impressive even today, regardless of the fidelity. And the Palpatine hologram was likely monochromatic because it was being received in outer space.
The hologram standard for a dedicated device without transmission issues was much higher, seen in the holographic game board that C3P0 plays with Chewbacca on the Falcon. But in general Star Wars was smart enough to keep the technology basically gimmicky - the real targeting computers and useful stuff was 2-dimensional.
Interesting blog post, but the conclusion seems hilariously optimistic to me:
>Eventually we’ll get to SF-quality hard AR, but it’ll take a while. I’d be surprised if it was sooner than five years, and it could easily be more than ten before it makes it into consumer products.
The post is already nearly 4 years old, and we've seen effectively no movement on the state of soft AR since it was written. I wonder if Michael still thinks we're going to see even very early prototype hard AR in the next 6 years.
As one of the commenters on the blog post mentions, AR in general is a whole order of magnitude beyond the next useful step. I don't need AR, just give me a wearable interface to my phone so that I don't have to pull it out of my pocket to interact with it. A simple heads up display has far fewer challenges to deal with than any AR system and would still be a huge leap forward in technology.
Smartwatches, sure. I have one and it's definitely an improvement over my phone alone when it comes to email and messaging. Still, I think there's potential for a Glass like device to be even more convenient.
I'm not too familiar with Cardboard, but I don't see how it's related.
Google Cardboard is a piece of cardboard (& software app) that you fold up into wearable - if quite dorky - glasses that hold your cell phone about 2 inches in front of your eyes. It's not exactly stylish, but it works - you do get a real VR experience - and it's completely free. It does augmented reality as well, using the cell phone's camera. One of my Googler friends showed me a cool demo where if you stared at a QR code in a magazine, you'd see elves and magicians battling it out on top of the pages, like a storybook come to life.
One of my wife's friends is doing a startup where you can print out the "gun" for a first-person shooter as well, so you get an immersive VR experience using nothing but your cell-phone and a couple pieces of cardboard:
I got to spend a day on the show floor at SIGGRAPH this year and AR/VR was the big trend. Most of the VR demos felt empty and lacked soul, but the Realiteer GermBuster VR demo was a blast! Definitely the highlight of the VR demos. I guess it demonstrates that (relatively) low-tech doesn't keep a game from being fun :-)
I think from the demos, it's pretty clear they won't be ale to 'draw black' - BUT - I think thats probably OK. Additive only AR could still be a really cool product, could still be widely adopted, be worth $5 billion, change the world, etc...
For a moment, I felt bad about having reservations about the prospects of ML - I've seen some neat public videos and announcements, but they all struck me as similar to a AAA videogame release from years ago, you know, "Not actual game footage" type stuff.
Then I come here and see you mention something that popped into my head as a related case - the Segway - and I don't feel like I'm being overly negative.
I grew up learning to deal with hype in part thanks to Saturday morning cartoon ads for action figures, so I'm trying to stay on a general "Wait-and-see" positive angle, but as time marches on, it's taxing on the patience.
Slightly off topic, but I don't see how this is similar at all to the Apple Watch. What secrecy was there about it, other than the completely normal "secrecy" that every product undergoes when it hasn't even been announced yet?
Apple was extremely forthcoming with details about it from the moment it was announced. They flooded the internet with concrete images, videos, and marketing material for months prior to its release. They had video tutorials on their site showing how to use mostly all of its features long before it was available in stores.
More importantly, the Apple Watch was just the latest product from a long-established product company that was basically printing money from other products. It went through the same development cycle that all their other products go through. This is tremendously different from a brand new, completely unproven company crystallizing out of the dust, given form from hundreds of millions or even billions of dollars of investor money, and then being extremely secretive about their only product on which the entire company will be based. Magic Leap and Theranos fit that mold, but Apple most certainly does not.
Furthermore, these promises of "mixed reality lightfields" make even Apple's marketing seem modest.
Just curious; recall the speculation that apple would be making a Tablet the day the original iphone came out? And apple denied it?
my question is "did they already know they were going to do a tablet, or was the tremendous hopes that consumers had that they would convince them to do so?"
surely their product roadmap now is likely much longer term than it was then, I'm just curious about the original iPad creation
From what I understand their original plan was a tablet and, if anything, the iPhone was a bit of a trojan horse to get people to buy a 'mini-tablet' and experience its benefits while directly having a good use case (who doesn't need a phone?!).
This blows my mind. Not because of the amounts raised without a single released product: investors have shown a propensity to throw spectacular amounts of money at baffling ventures.
Instead, my shock is the implied burn rate of these raises. LinkedIn reports Magic Leap is between 201-500 employees. They are no indications of them mass-producing a product. No sales, and their marketing seems to be exclusively articles about the enormity of their rounds.
How, then, is a company raising 790M 1.5 years after raising 500M? What possible expenditures, without mass product development, sales and marketing, consumes finances at a rate that warrants these raises?
Or, alternatively, maybe the founders and employees simply despise maintaining ownership of their company.
Aside from the massive research efforts they're undertaking I'm under the impression that they're also manufacturing. I'd guess that comes at a serious cost.
I really hope what they have is real, but my mind keeps going back to clinkle. A company that was able to raise a large sum of money with smoke and mirrors. Granted Magic Leap is raising an order of magnitude more than clinkle, but smart people have been fooled before.
I find it interesting that the top image (of the floating robot) has the text saying "no special effects" whereas the gif of the elephant in the hands (been available online for awhile) does not. This tells me that all the videos/gifs we've seen (the elephant, the whale splashing down thru the gym floor) are "aspirational" (AKA "bullshit").
Don't get me wrong, the floating robot is cool but this is doable on HoloLens (yet to ship but coming in Q1). Definitely seems like there's a real risk here of overhyping MagicLeap.
EDIT: Fixed the link to TechCrunch...If you saw the other link we're hiring :)
The floating robot is not doable on the hololens because metallic effects require black, which is not reproducible on the hololens display... Yes, I have seen the photos of hololens output that show black. I don't believe that's a reproducible effect on their wearable.
Disclaimer: I don't really know anything about this area, I'm just saying what I see in the floating robot. I don't think there is any black in that robot. It's more obvious if you look at the youtube video, but do you see how they have put a filter on real objects and distorted the colours? It seems to me that they are creating a black level with the background image. Then by using only bright colours in their foreground objects, the contrast makes it seem as though it is black. If you were to measure the colour (say by using a colour picker on the image), I'll pretty much guarantee that there is nothing in that image that is blacker than the background.
In the Youtube video they also seem to use a trick of defocussing the real objects in the background when their rendered foreground objects are in focus. You can see that most clearly with the planets. I think they have a macro lens on the camera to achieve this effect and I suspect that it would not be nearly so dramatic if you were to experience it with the naked eye.
If you don't have black you can't achieve metallic effects... If you disagree, please explain how it is possible to achieve said effect without black.
As far as I know, and I was pretty deep into this at one point, metallic requires black and without it you cannot achieve any sort of luster-like effects.
A subtle problem is where to put things that don't just add noise and clutter.
Look around your house or office, where would you like digital things to sit? Ok how about when you move 2 feet this way, whats the new answer for where you would like it so it's easy to see/interact with?
Someone else in this thread mentioned Theranos and while the two businesses are wildly different, I still think there's something to it: investors can be fooled by flashy presentations.
I'm going to continue to be suspicious of Magic Leap until I actually see something that delivers on the promises they are making. Luckily(!) for me, I don't have enough spare money to be an investor in these things anyway...
> I see AR being useful in verticals ie "pipe fitter" or "nurse".
I do think IT folks forget just how many nurses, doctors, pipe fitters, car mechanics, telephone engineers, etc etc etc are out there. Literally millions.
Hook these glasses up with google maps and other online databases will be neat.
I know someone who worked on a heads up display for paramedics. People have been trying to make these for a few years.
Just think how complicated it actually is for a nurse:
- it's got to supply actually useful information that isn't fairly trivial to get by picking up a chart
- it'll have to hook up with some sort of input device
- are they carrying around the AR device and the input device at all times?
- it's got to be wearable for 8 hours or it's completely pointless (you can't use it half the time if it's essential)
- it's got to be light and durable
- for almost every hospital it'll have to have specialized software written to integrate
- an amazing UI to use it without touching anything as they can't keep cleaning their hands to use it
- there's got to be total network coverage in every hospital room, it can't cut out in wi-fi deadspots (I have no idea how good this usually is)
And those are the easy problems I can think of without any particular domain experience. I imagine it will get a lot more complicated when you know what they actually do.
What's the first supposed application of any technology? I'm not really in favor of this idea, but here's a shot at what someone will do with AR: "what do people shopping with me and eating the free samples here at Costco look like when they're naked?"
Slap a depth-sensing thing (like Project Tango https://www.google.com/atap/project-tango/) onto the AR stuff, you can probably get a Kinect-like algorithm to sense peoples' limbs, bodies, etc., and come up with a hypothetical-location-in-3d-space. Then the hilarity will start with projecting cartoon body overlays. Successive versions of the app will do better recognition and fitting and come up with more realistic images. Eventually you can have AR "beer goggles" or whatever effect you want.
To me Magic Leap is an interesting test of the investment people at Google and Ali Baba. I'm giving them the benefit of the doubt, but if Magic Leap is a spectacular failure, I think it will be a strong signal that Google and Ali Baba's investment groups are similar to the groups at other big tech companies. i.e. full of strong political players without much industry acumen.
I'm particularly biased though. I've always been a bit skeptical of the amount of money corporate development people make ... I feel like they don't bring that much value to the table as individuals, even when the corporate development function is essential.
When you meet a good engineer, lawyer, or salesperson, you can usually pretty quickly understand what they are bringing to the table. With investment people I feel like they just make decisions, and then they aren't often measured against what the results of those decisions are.
The people that should probably actually be corporate development people ... the ones that are unsure of themselves without sufficient data. The ones that listen quietly and form measured opinions. Those awkward types ... the ones with valuable insight ... I never see that type coming to visit our offices as part of the steady stream of investors.
All I see are well put-together people, with good personalities, strong handshakes, excellently tailored clothes, good presentation skills ... and a fair understanding of the industry ... plus maybe slightly above average intelligence. They are usually warm and personable though. Its clear they have that skill.
I have no idea why the CEOs select these people for the job though. Any one of our engineers if tasked to evaluate companies like ours, would ask better questions, would be more intelligent about figuring out what is our marketing fluff, and what is actually ready to go.
Is this another Theranos? I hope not but all we are hearing are promises, NDAs and big rounds/valuations for a product not yet on the market. One kinda grows suspicious after so many howls.
It's a very shortsighted reasoning. If this turns out to be a completely lost investment, investors will become wary of investing in VR startups, which will have for consequences less R&D jobs in virtual reality (which itself will decrease negotiating leverage of all R&D engineers), and less VR products, sold at a more expensive price.
I don't know anybody who works at Magic Leap, but I'd rather prefer they don't waste all that money.
Whether it succeeds or fails, you might consider the opportunity cost of that financial investment and hours of work.
For example, that's a lot more than the development of the original Macintosh, roughly 0.5% of the Apollo program and roughly 3% of the Manhattan project (both in today's dollars).
This can't possibly just be a hardware play. This investment would buy 1.3 million Oculus Rifts at retail price. At component prices and figuring economies of scale lets say it could fund the construction of perhaps 5 million units. Are there really 5 million people waiting to buy an AR device? Google Glass sold perhaps 50,000 units--has the market somehow grown by a factor of 100x in the past couple of years? I just don't see enough people wanting to own this tech. Not now, not even in 10 years.
The more likely possibility in my mind is media (this becomes the thing that replaces 3D movies or 3D TV) where they collect a royalty on everyones content, or perhaps theme parks (they are in Florida after all). A Hogwarts where you could actually see the spells you cast, or an Imagineered experience where you could walk through Mary Poppins' animated countryside might be capable of generating enough revenue to justify this.
I agree with much of what you're saying. However I'm not sure your comparison with Google Glass is fitting. The market for a functional Google Glass might also have been a lot bigger than 50 000. Not a lot of people outside the tech-environment knew a whole lot about it, and there weren't really an eco-system around it since it was never a consumer-grade product.
So I think the market is bigger than previous numbers indicate, however I'm not sure how close to perfection it would need to be to show it's true colors. Hopefully the launch of Oculus and Vive might give us an indicator this year.
(As an anecdote,) personally I would want something like this if it actually had enough content to let me reuse the product without indefinitelly replaying tech-demos. :)
No, last time they did a real (aka not a fake PR YT clip) demo their form factor was a giant table with a Phoropter like strap on producing low res single color image.
It's going to work but it's not going to work amazingly so, probably just suffering from a low FOV. With that said, there's too much money and too many people being hired for this to fail incredibly hard; it just wouldn't make sense.
Is there anything stopping these guys from covering the whole view field ?
I've tried a number of VR headsets and they all feel like a stopgap solution, there has to be something better than wearing 2 x 4" screens mounted in a cone on your head....
The same thing which is stopping anyone from covering the whole visual field, the need for serious technological innovation. It's possible they've figured that out, but who knows.
This is AR, not VR, right? The two are quite different. VR does not aspire to be AR. It seeks to completely replace the world around you, not just add onto it.
From what I understand, upcoming VR headsets like the Rift and Vive have pretty solid FOVs, but Microsoft's HoloLens AR headset has a significantly more limited FOV.
Microsoft does. ;)
AFAIK that's because they're trying to keep the price-point low, since they're aiming for consumers.
Edit: AFAIK, the tech they're using can support a wider FOV, but then the price jumps pretty quickly. I definitely do NOT know for sure, however; that's just what we figured when they were first announcing.
Microsoft could slightly enlarge the Hololens FOV for a significant increase in cost, but it still wouldn't be good. They would have to change the display technology they use to have something immersive.
CastAR is pretty neat and damn simple, although you have to prepare the space. If I had to bet on a consumer AR technology, I'd probably bet on CastAR, although I haven't checked up on them in awhile.
CastAR is not AR in the same sense that ML or Hololens are supposed to be for any environment anywhere. Think of it like 3D glasses for specific surfaces.
Short term practical or tactical secrecy, sure. And in some industries, probably not as bad a signal. But software-heavy hardware ecosystems aren't one of those.
Outside of wars, Ultra or Manhattan just aren't that good a way to work. Why cut yourself off from the global community?