So here's my concern with Magic Leap. (And I say this with friends working there, so I WANT them to succeed more than anything, but I can't help but look askew at some of the secrecy)
If their technology is _so critically secret_ that everything needs to be under more NDA/IP protection than I've seen in almost any other consumer product, one would think that implies it's readily duplicatable if someone just had the secret sauce? But in that case, one would expect to see more results after how long they've been in dev. I've heard in many cases founders/VCs laugh at the concept of "keeping your million dollar idea secret" because of the concern that someone will steal it, let alone a concept so complex that a group of some of the better specialists in this area (AR/VR/robotics/wearables/vision) have taken this long to bring a product to market.
So is the concern just about entities like Google/MSR who would show up out of the blue willing to throw essentially endless money at the problem? Because off the top of my head they have not shown a great track record of even executing on _their own_ ideas with a compelling end product on a fast enough time-scale to be competitive against someone with a multi year lead. I'm just not sure I follow the benefit of this level of secrecy compared to the amount of skepticism it naturally generates in someone who would love to be an early adopter.
It's wholly possible that taking the idea from a concept to a working product is just so hard that they don't want to show it off before it's 'ready'.
Demonstrating an innovative product before it's in a state where people can really understand it is risky - getting written down by tech pundits would seriously increase the time it'd take to get the public to buy in. Arguably that's what happened with Google Glass; early adopters bought in but couldn't realise the potential of their investment, so there was something of a backlash. That limited Google's options for a bigger launch because people wouldn't have bought in. They apparently kept going but in secret. Magic Leap are doing the same thing but without the mistake of launching too early.
Google Glass failed not because it wasn't ready. It WAS ready product-wise AND technology-wise. It failed because it was just a bad product. Nobody likes to be video recorded without knowing, and nobody likes to make other people uncomfortable.
If Google Glass had instead launched WITHOUT the live streaming and all the features that violate people's privacy, it may still be around, no matter how bad it looked.
What excited people about Google Glass was NOT being able to record others without them knowing, but how it could potentially provide value by augmenting reality.
Likewise, I'm sure Magic Leap should be "ready" at this point. If they are not, it means they still haven't found their MVP, which is a bad sign. I have never seen a piece of technology that succeeded by launching 100% refined. All successful technology launch as something that's not quite perfect but even with the limited performance provides value to small set of audience. And this includes even Apple.
Not necessarily. One could imagine that with far superior voice interaction technology and display (rather than a tiny screen in the corner of your vision and cruddy voice interaction), that would open up Google Glass to substantially more compelling product uses.
Too many companies worry they don't have enough use cases covered and can't forget "key" functionality, leaving them with a bloated and directionless product. Better to keep it simple and work up. Just nail the first use case enough for the innovators to adopt it.
> Google Glass never attempted augmented reality though right?
I don't know what you're talking about here. Is this a rhetorical question? AR was exactly what excited the nerds about Google Glass. But Google lost its way along the way and decided they will also become GoPro.
There's a huge difference between having a camera and sharing the content you took from the camera.
I didn't see this reply for 3 days so the conversation is probably over, but I think they never actually showed AR promos for the Glass (not at all like Hololens for example). The nerds being excited about it seemed to be excited for AR on Glass was just them jumping ahead to the assumed conclusion rather than what they were showing with the tech (which was more "video chat and phone notifications anywhere" even in their ukulele-music-backed promo videos)
> It's wholly possible that taking the idea from a concept to a working product is just so hard that they don't want to show it off before it's 'ready'.
So if it's not 'ready', then what do GV investors know/see about the product that an average person couldn't see? In other words, if the product is in a half state, then what confidence do the investors have the team will get it to the a full state. It's like saying "invest in my rocket company that is going to take people to another solar system...and here's the demo of a rocket going to the stratosphere". No one has been able to send a rocket to another solar system but a plethora of organizations have done so to the stratosphere, so what additional insight does one have that makes me so confident that this is the team to get us to another solar system?
So if it's not 'ready', then what do GV investors know/see about the product that an average person couldn't see?
Google Ventures isn't Kickstarter. The people who work for GV do a lot of research before investing. They'll be able to extrapolate from the available data. Average people are very bad at extrapolation.
In other words, if the product is in a half state, then what confidence do the investors have the team will get it to the a full state.
They'll have talked to the founders, and trusted that they can do it, and, GV will understand that if they're wrong they will lose a lot of money. They don't have confidence. If Magic Leap was something you could be confident about then Magic Leap's founders would have raised debt rather than equity, and kept all the rewards for themselves. Confidence implies a lack of risk. GV has belief in the Magic Leap team, but that doesn't mean there isn't a huge risk in putting money in.
> The people who work for GV do a lot of research before investing.
Just out of curiosity, do you know this as a fact or is it speculation based on your assumption of the prestige of individuals there?
> Average people are very bad at extrapolation.
I do technology due diligence for a living. I work with firms like GV and am always surprised by the general lack of researching conducted (look at DJF and Theranos as a great example where seemingly un-average people didn't extrapolate correctly)
> They'll have talked to the founders, and trusted that they can do it, and, GV will understand that if they're wrong they will lose a lot of money.
> GV has belief in the Magic Leap team, but that doesn't mean there isn't a huge risk in putting money in.
That's precisely my point. What specific data points do they (GV) have, that, as the GP points out, leads them to believe this will work. Data which if publicly available would lead techies like you and I to lead the same conclusion. Further, there is a large assumption that this same data/info wouldn't appear to materially change the course of the company's success. Oculus launch a beta early..why does that need to be the case for them? People are right to feel skeptical - the publicly available data points don't add up.
I'm not sure Theranos is a great example for lack of due diligence. Most SV money stayed out of Theranos despite the hype because they couldn't get straight answers out of them on their tech. Obviously there were other investors jumping on board without sufficient due diligence but it wasn't the normal crop of SV VC.
Tim Draper of DFJ invested in Theranos very early on. His parents were neighbours with Elizabeth Holmes when she was a kid, which kind of explains why they invested in Theranos. No other reputable VC firm would touch Theranos, and Holmes had paid several visits to GV and was turned down each time. GVs bigger bets are on Uber, which turned out well, Magic Leap, and Carbon, a snazzy 3d printing startup which appears to be a solid company with some really innovative tech.
> The company raised $32.36 million of Series C...Draper Fisher Jurvetson...on November 15, 2006,
> The company raised $45 million of Series C1 venture funding out of a planned $100 million from Draper Fisher Jurvetson on July 8, 2010, putting the pre-money valuation at $1.1 billion.
> The company raised an estimated $573 million of Series C2 venture funding from The Lucas Venture Group (Menlo Park), Draper Associates (San Mateo), Partner Fund Management (SF) and Sandbox Industries (Chicago) in March 2015, putting the pre-money valuation at $8.4 billion. BlueCross BlueShield Venture Partners also participated in this round.
So anyone thinking of "stealing" their idea has to deal with the facts that 1. They're already way behind, 2. The monetary investment is massive, and 3. They may never catch up because they'll be in an endless employee poaching war
That's like saying it costs a billion dollars to develop a drug, so there's a big barrier to entry. Making the first pill (or Magic Leap) might cost $1B, but the second will only cost a couple hundred bucks. That's why there's a patent system.
You can just wait until the demand has been proven, and then jump in. That's what's happened to GoPro and to most PC/smartphone OEMs, and it's what happens to pharma companies when their drugs go off patent.
I don't think the money to be made in VR/AR will come from making the headsets themselves, but from providing some sort of software with lock-in. If Magic Leap has that, then they could become the Apple of AR, but so far it looks like they're just developing fancy hardware that the Chinese will quickly clone.
The Chinese "knock off" people have trouble making an equivalently good iPhone. A scanning fiber array based virtual retinal display can't be quickly cloned.
They've been building their own fabrication facilities in Ft. Lauderdale, they are trying their best to keep the chinese factories from stealing their IP.
The Chinese knockoffs (and Google, for that matter) have trouble making equivalently good software (iOS). It's a challenge in software development and hardware-software integration, not hardware development.
Put it the other way: what would be the benefit of publicly disclosing their secret sauce? Perhaps to help fuel their recruiting pipeline, but I can't think of any other good reasons to do so.
They don't have a product out, their secret sauce might change, Google/MSR might be working on something similar and just need a few hints in the right direction to get there, they get to keep a big first-mover advantage in whatever area their tech enables them to move into, and there's no opportunity for possible early adopters like yourself to get in on the action yet.
I'm guessing their tech isn't so much "critically secret" as there just isn't a benefit to their business case for publicizing what they are doing
The problem with VR/AR is latency. It needs to be 10ms or less to feel real. We aren't anywhere near it.
The same is true of touch screens: when latency gets below 10ms, the experience changes to an actual physical object. Current touchscreen latencies are terrible, but we don't notice it because they are still useful for input and we don't know any better.
If we can't even do it for touch screens, how can we do it for VR/AR? The glut of products is just a supply-side bubble.
Magic leap is way beyond designing low latency display drivers (hardware or software). They're designing and fabricating their own many-layer photonic crystals with complexity rivaling that of VLSI chips from a couple decades ago (except not on silicon, meaning they've had to reinvent a significant fraction of the process).
Latency may be an issue for them in terms of making seamless AR, due the limitations of existing camera hardware in terms of latency (existing camera hardware is not designed for low latency), but I suspect they're partnering (or will be partnering soon) with major players in the CMOS imaging realm. As far as display latency, I doubt it will be an issue for them. They're already having to design the entire display from the ground up---I suspect they have the forethought to design it as a low latency system.
Hololens runs at 60Hz and the leading VR glasses run at 90Hz. We're quite near it. Having used these products myself, it definitely can be enough for suspension of disbelief.
The reason touch screens have lag is because the render thread isn't as prioritized. An iOS device that isn't loaded with software has a basically instant response.
I would love it though if we prioritized the UI thread across all devices, it feels a bit nutty to have this much processing power these days, and still have a UI lock up.
Unfortunately, refresh rate is different from latency, it's the pipeline that matters.
I've tried iOS devices on display, with no extra software loaded (is that what you mean?). If you waggle your finger back ane forth, it's easy to see the lag.
If it is easily duplicatable this strategy allows them to build up tonnes of labour that would take a copier a lot of time to replicate. So there's a bit of time to capture up profits and evolve to the next iteration.
If you'd like an individual take on secrecy and inventions from a bootstrap / one-guy perspective, I wrote an informal essay discussing a lot of the questions and concepts you touch upon:
Ah, I'm just now noticing that this is your post. I enjoyed your writing in this post and others I read after clicking around. We seem to share interests, though I suppose that's not so surprising considering this is HN.
Why thank you for mentioning this, it's a nice thing to see on a Friday! This is a neat community in a lot of respects and I appreciate the challenges and questions and whatnots that I come across here. Cheers!
was totally thinking the same thing... 100% whatever they release unless it cures cancer and does AR and VR and uses displays that are like 1 million PPI, I think people are gonna be disappointed.. they are doing themselves a huge disservice by keeping it so hyped and so quiet.. same thing with the segway, people thought by the time it would launch that it would be like a legit hoverboard that shot beams of antigravity out the bottom.. then they launch and show people riding on two wheels with kids helmets and kneepads and elbow pads on.. I was like... um, no...
I fear magic leap is riding the hype train more than is healthy. They're setting their potential future customers up for disappointment. It seems unlikely to me that whatever they're working on is so revolutionary that we're all going to agree that the secrecy was necessary.
I'd be interested to know if there's any precedent for this kind of strategy working to increase profits. The only examples I can think of (e.g. Segway) are ones in which the hype actually worked against adoption of the product when it was released.
At the end of the day, I really do hope they succeed. It will be great when someone comes up with the analogue of the first iPhone except for AR.
Most people (including the technical) that I've mentioned the name to, don't recognise it or know what it does. Maybe its just us hyping it up amongst ourselves?
Suffice to say the tech that Magic Leap has is so much leaps and bounds better than Hololens that it will derail Microsoft's AR train completely once its launched. The hype is very real and so is the tech.
-Magic Leap is located in Hollywood/Plantation/Fort Lauderdale which is nowhere near Orlando
-For now, Magic Leap is not in the VR space, rather they are augmented/mixed reality
-Magic Leap is very much a hardware play, not a software one. From what I've heard (and this could be incorrect), their headset runs a heavily-modified version of Android, not a new operating system developed from scratch.
> While Magic Leap is clearly going after talent in locations that have established themselves in fields integral to VR/AR.
Florida – where the company is headquartered – is a burgeoning gaming and graphics hub with Orlando as one of the biggest video game development communities in the US.
Except they're in Ft. Lauderdale?
Still pretty skeptical of Magic Leap given time-to-market(public demo, etc). I do hope they do well, though - I grew up in the area, cool to see some interesting tech there.
Curious how this is #8 on the front page with 3 points[1].
Even more curious what the actual product of this company is going to be. A VR dev kit? Their website is devoid of anything but videos of people going "Ooohh!" at special effects.
Supposedly AR glasses with multiple focal planes and with localized occlusion (blacking out the world; don't know if this part is being done at the different focal planes, I'm more skeptical of it than any of the other parts, because all of their through-glasses views have been in very dim rooms).
> Supposedly AR glasses with multiple focal planes and with localized occlusion (blacking out the world;
I recommend being more specific when referring to occlusion with respect to AR systems. There are two primary forms: (1) "Real Occlusion" in which selected wavefronts are blocked (occluded) such that a transparent near-eye display is capable of displaying black, (2) "Virtual Occlusion" in which real objects are actively scanned/tracked to permit virtual objects to appear occluded being real objects.
ML system claims both capabilities, and I agree with you that Real Occlusion is several orders of magnitude more difficult. I think real occlusion is valuable, because in addition to displaying black it enables images to appear solid vice translucent. However, I don't feel it's an absolutely necessary feature for a quality AR systems. After all, ultra bright projection displays present perfectly fine images without displaying true black. They simply exploit the properties of human eye perception with respect to brightness contrast of adjacent pixels. An AR system can do the same.
Personally, I would prefer to have a quality AR system with unlimited FOV constraints and virtual occlusion capability sooner rather than wait for real occlusion to be perfected.
Microsoft has a series of tutorials using Unity for Hololens. Udacity just launched a new VR course using Unity as well.
However, performance can be a critical factor depending on the app, and C++ may be the only solution. For that you need low-level knowledge of shader+texture+render techniques, which for Hololens is generally all in DirectX.
There are very few business-oriented frameworks for any of the devices, almost everything is designed for gaming (short scripts and machine-oriented code organization). Implementing most of what we normally imagine being part of an 'app' has to be done from scratch.
I like the facet that they don't follow the hype and actually open offices where they can find the best talent for their company. they did the same by opening an office in Israel as well where they can find a lot of VR/AR experience.
If their technology is _so critically secret_ that everything needs to be under more NDA/IP protection than I've seen in almost any other consumer product, one would think that implies it's readily duplicatable if someone just had the secret sauce? But in that case, one would expect to see more results after how long they've been in dev. I've heard in many cases founders/VCs laugh at the concept of "keeping your million dollar idea secret" because of the concern that someone will steal it, let alone a concept so complex that a group of some of the better specialists in this area (AR/VR/robotics/wearables/vision) have taken this long to bring a product to market.
So is the concern just about entities like Google/MSR who would show up out of the blue willing to throw essentially endless money at the problem? Because off the top of my head they have not shown a great track record of even executing on _their own_ ideas with a compelling end product on a fast enough time-scale to be competitive against someone with a multi year lead. I'm just not sure I follow the benefit of this level of secrecy compared to the amount of skepticism it naturally generates in someone who would love to be an early adopter.