Slashdot Mirror
Magic Leap Used Fake Tech Demos and Is 'Years' Behind Schedule (ibtimes.co.uk)
New submitter drunkdrone writes: Magic Leap's coveted mixed reality technology has been the subject of intense speculation since it broke ground in 2014. Having secured billions of dollars in funding from some of the world's biggest tech giants, the secretive start-up has managed to stay at the centre of the VR/AR conversation despite showing little of the so-called revolutionary technology it has in the works. Now, the Magic Leap hype bubble may be about to burst in spectacularly disappointing fashion. According to reports, the Florida-based start-up is years behind on its plans and may have used deceptive product demos in order to keep interest in its tech alive. The Verge, which quotes an exclusive article from The Information, reports that Magic Leap's mixed reality technology has long since been overtaken by other products already on the market such as Microsoft's HoloLens, which Magic Leap's technology is said to most closely resemble. Allegedly, Magic Leap has struggled to scale-down a bulky piece of laser projection equipment used within the headset's display. "The crux of the problem appears to be Magic Leap's gamble on a so-called fibre scanning display, which shines a laser through a fibre optic cable that moves rapidly back and forth to draw images out of light," reports the Verge.
"Any sufficiently advanced technology is indistinguishable from a rigged demo".
Florida-based start-up
Wait, a scam in Florida? That's unpossible!
Note to VCs and other money-types.
When a candidate talks about 'revolutionary technology' make sure you see it actually working before you give them mountains of bucks. Oh, and make sure you get it independently tested, too.
Tech has already changed the meaning of 'innovative' to 'same as last year's model minus an interface port' now they're turning revolutionary into some ironic hipster term.
I hope they don't. When things fail like this I love it. This is capitalism at its finest. It's eating its own and leaving the stupid out to dry. I just wish it happened more. That's the only thing to be mad about here.
Y'all want this beast, you might as well fucking enjoy all of it. Don't cry when things go wrong and the market does what the market should.
What do you think they actually did with the money? Burn it with lasers? Pew, pew!
Pretty impressive to be "years" behind schedule, 2 years after you founded the company. They should declare "time bankruptcy" and start from scratch.
Until some physics breakthrough happens where we get cold fusion, not much will change.
Only the State obtains its revenue by coercion. - Murray Rothbard
The reality is that the tech industry has reached a dead end with the death of Moore's Law.
Are you implying that raw processing speed is the only possible avenue of innovation? Really? There are no other possible ways that technology can improve?
People who say "sheeple" have about as much sophistication as an AOL user, and in fact are probably actually AOL users.
Only private. And probably all faked well enough to fool investors.
There are two types of people in the world: Those who crave closure
The reality is that the tech industry has reached a dead end with the death of Moore's Law.
Is the problem really processing power, though? For a system like this, it seems like there are other problems bound to creep up:
* AFAIK, we still don't have good enough AI to figure out a spacial 3D world from visual input. I know it's still being worked on and there's been progress, but being able to place objects in the real world in this kind of augmented reality requires that the computer can figure out the layout of 3D objects within the real world.
* Even if you can render the graphics and place them appropriately in the world, there's still the problem of designing the UI. You need to create both the visual look of the interface, and figure out which gestures to use for different controls. The interface (input and feedback) needs to be easy and intuitive and provide clear feedback to user interaction.
* You also need to make the gestures such that they're read by the computer reliably-- that is, if I'm supposed to do a specific hand motion to activate a feature, the hand motion needs to be something that the computer will recognize almost every time it is performed, it needs to be distinct enough from other control gestures and natural gestures. Basically, people need to be able to control these systems without constantly activating various controls by accident.
These are fairly difficult problems for computers to figure out, and as far as I know, they're not really a problem of insufficient computing power. That is, as far as I know, it's not like we've developed code that can do these things and a UI that works well, but we need a computer 5x as powerful to run it in real-time. The problem is that we just don't have the design/code to do it.
It sounds like those investors had a very mixed reality already.
Perhaps not in terms of actual speed/# of transistors, but there's plenty of room for improvement elsewhere in the hardware. Additionally, I think we're only really just seeing the catching-up of software to the fact that things have multiple cores (and many specialised cores too).
Moreover, we're now in an age where computers are becoming totally ubiquitous and embedded in the very fabric of our day to day lives, so enhancements in software and connectivity will drive new advancements. 'Virtual assistants' will become more prevalent and gain more useful functionality, direct mind control interfaces (like the kind we're seeing developed by the prosthetic's industry) will continue to get better and the path of miniaturisation will continue. VR is obviously being pushed right now so expect to see that increase in prevalence (it seems to have gained more traction this time than the previous efforts with underpowered hardware and massive headsets) and actually seems to do a reasonable job. Something like Magic Leap is not too far off I would imagine, if not from them.
Finally, I'm pretty sure we've been hearing of the death of Moore's Law for some time now but something always crops up that extends it for a little longer. I'm sure there are many brilliant people working on the problem of faster/smaller/less power hungry all the time so I have no idea what the next few years are going to bring in terms of new discoveries there, but I wouldn't be so quick to write it off just yet.
The reality is that the tech industry has reached a dead end with the death of Moore's Law.
It's absolutely adorable that you think all progress in the tech industry is rooted in Moore's Law and that nothing more can be accomplished if we see a slowing in the rate at which we pack transistors into a given area on a chip.
You will see incremental progress from here on out, but no more large leaps like we have had for the previous 40 years.
All progress is incremental and Moore's Law is nothing if not incremental. If you didn't know that then you didn't understand what was going on. Moore's Law was just a observation of the fast but incremental development of semiconductor manufacturing. However it isn't the end-all-be-all of tech. It's not some fundamental law of nature, just an empirical observation of incremental change.
People will over-believe what they see. I learned this many years ago, preparing animations on a SGI workstation for use in courtrooms. If a lawyer showed up with an animation of an accident, the jury took it as real and would rule for his client.
It's a two edged sword for tech. Can't get any money unless I show a mockup or prototype. However, once a customer sees a mockup or prototype, they think it's mere inches away from production even though I tell them it's miles. I've even been told I just wanted extra money to "do science projects" instead of heading straight for production when I've tried to warn people how immature the tech was.
Just because we're running in to issues throwing transistors at the problem doesn't mean there isn't room for advancement.
We're coming along quite well making those transistor faster, and learning to better use them. More importantly we're getting VERY good at making data move in and out faster - Which for decades has been the real bottleneck.
Since the late 90s desktops and consumer devices have had fast CPUs starved for data.
The march has always been slow and incremental. Revolutionary products are usually outright fraud. The ones that aren't just employ existing technologies in a novel manner.
If you see a new 'revolutionary' product and any one piece of it seems to massively outpace any existing peer (Say crazy graphics performance, extremely goo motion tracking, insane battery life), you know it's bullshit.
If for nothing else, they'd be focusing on that one piece that's leaps ahead and marketing that - Rather than tying it to some complicated pie-in-the-sky venture.
There is a good argument for better software design being more important than Moore's Law when it comes to complex breakthroughs in computing. It can be hard to quantify how algorithm improvements compare to hardware improvements, but the field of numerical algorithms gives some insight.
In the field of numerical algorithms, however, the improvement can be quantified. Here is just one example, provided by Professor Martin Grötschel of Konrad-Zuse-Zentrum für Informationstechnik Berlin. Grötschel, an expert in optimization, observes that a benchmark production planning model solved using linear programming would have taken 82 years to solve in 1988, using the computers and the linear programming algorithms of the day. Fifteen years later – in 2003 – this same model could be solved in roughly 1 minute, an improvement by a factor of roughly 43 million. Of this, a factor of roughly 1,000 was due to increased processor speed, whereas a factor of roughly 43,000 was due to improvements in algorithms! Grötschel also cites an algorithmic improvement of roughly 30,000 for mixed integer programming between 1991 and 2008.
My guess is we have plenty of room for improvement as we find ways to live within the confines of physics. Even if we don't find a better alternative to silicon based computing, advances in computer science has the potential to improve our computational ability by a factor of millions without needing Moore's law.
-- All that is necessary for the triumph of evil is that good men do nothing. -- Edmund Burke
As far as software goes, isn't processing speed the only thing that matters? Adding more RAM, doing parallel computing, etc are just other ways of increasing the processing speed too.
On the contrary, software could go a long way to utilize the parallelism better. Heck, most consumer software if re-written carefully would achieve speedups well beyond a couple Moore's Law generations.
Doing that re-writing properly would probably also cost more, which is why people have been happy to follow the hardware route for so long.
Remember back in the late nineties, when people (mostly stupid liberals) convinced Congress and Clinton to pass a "Luxury Tax" on things like ... yachts? Remember what happened to all those industries that were building those things? Do you remember the fanfare when they repealed that tax?
Yeah, nobody remembers shit like that.
Agent K: A *person* is smart. People are dumb, stupid, panicky animals, and you know it.
As far as software goes, isn't processing speed the only thing that matters? Adding more RAM, doing parallel computing, etc are just other ways of increasing the processing speed too.
For the lazy programmers, yes. Outside of embedded systems nobody has cared about optimization for decades now because they can always just up the hardware specs and throw faster hardware at their inefficient code.
Getting off Spinning disks and onto SSDs has increased computing performance much more than processor speeds has. And it isn't even close. We've just been stuck on Spinning drives for so long, that we thought increasing processor speed would solve our problems, when that wasn't the issue at all.
Agent K: A *person* is smart. People are dumb, stupid, panicky animals, and you know it.
Sounds as bad as Coleco Chameleon
Twinstiq, game news
Do you? This happened in the early 90s, actually. HW Bush implemented it, Clinton repealed it his first year in office.
Slashdot: providing anti-social weirdos a soapbox, since 1997.
You haven't used a Hololens obviously. All of your points are adequately addressed - real time room mapping over multiple rooms, instantly understandable UI all controlled by (essentially) two gestures and voice. My mind was blown when someone walked between me and the 'hologram' I was playing with and the image was properly occluded to mask the person.Works incredibly well now and will get better. We have the flops, the hardware and the code to do it all now - you are out of touch!
Pension funds are in the VC business, so you're wrong.
I saw this first hand when I purchased an HP 49G calculator.
Many operations that would hang the 48G for seconds were instant.
If memory serves, they had the same processor, but the 49G had been optimized. When reading that it seemed like BS, but when using it, it was a shocking increase in speed.
Wow, sent an e-mail as suggested when clicking on "use classic" banner, and got a fast response that addressed my msg
You haven't used a Hololens obviously. All of your points are adequately addressed
Pure unadulterated bullshit. I've used the Hololens and it is nowhere near as advanced as you claim.
This is capitalism at its finest. It's eating its own
I think you mean cannibalism...
My eyes reflect the stars and a smile lights up my face.
Years ago my boss worked for a software company that sold layout and design software (desktop publishing essentially) to newspapers. They were quite successful and had a lot of clients across the world. They had a few ideas of some cool new features that they would like to build into their next release, and the boss thought it would be neat to demonstrate these future features at a major trade show, to get the clients excited. So they mocked up a convincing demo of how the product *would* work, complete with scripted mistakes (undo) and everything. They did this all live with a guy pretending to interact with the software. But it was all faked.
Well, they were right about the clients and potential clients. They were pretty excited. Very excited as a matter of fact. So excited that all of the companies that had signed on to buy their current version of the software immediately canceled their orders in anticipation of this new version. The problem was of course that it didn't exist and wouldn't for years if ever. Unfortunately that little demo completely killed the company. Their real product just couldn't compete with the hype of their imaginary product. Had they been honest about it up front, they would probably done fine and eventually implement many of those cool features.
All of it didn't expire in the 90's. The automobile luxury tax didn't go away until 2002 courtesy of George W. Bush. Get off your high horse.
How would that affect American companies, anyway? They still don't know how to build a luxury car that doesn't feel like the inside of a plastic lunchbox.
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
I heard they alos solved the battery problem by using an energy catalyzer
https://en.wikipedia.org/wiki/...
and they get haptic feed back from an EM drive thrust.
Some drink at the fountain of knowledge. Others just gargle.
Presidents don't make tax law. They give a suggestion to Congress, and Congress does whatever the hell it wants. The luxury tax was passed by a Democrat-controlled House and Senate. Bush was forced to sign it because the Democrats refused to pass a budget which didn't include a tax increase (leaving Bush no choice but to go back on his "read my lips - no new taxes" pledge).
Clinton repealed it because by then it was obvious it was killing the industries it effected (e.g. yacht sales plummeted and boat manufacturers were laying off workers - yacht construction is one of the few industries where it's relatively expensive to outsource the labor because you can't just put the end product on a bulk cargo ship for cheap transport from Asia). And the Republicans managed to gain control of the Senate thus preventing Democrat legislators from unilaterally dictating tax policy.
by Theranos
Some drink at the fountain of knowledge. Others just gargle.
Real news about a post-truth demo.
Confucius say, "Find worm in apple - bad. Find half a worm - worse."
Failing because you bet on the wrong tech? Fine. But faking tech demos to dupe investors? Go to jail. One of the stupidities of blind worship of capitalism is that capitalism only works if everybody has access to accurate information.
This space intentionally left blank
--sigh ----
So reporter Kevin Kelly - went to magic leap, put on the prototype and says:
Magic Leap’s solution is an optical system that creates the illusion of depth in such a way that your eyes focus far for far things, and near for near, and will converge or diverge at the correct distances. In trying out Magic Leap’s prototype, I found that it worked amazingly well close up, within arm’s reach, which was not true of many of the other mixed- and virtual-reality systems I used. I also found that the transition back to the real world while removing the Magic Leap’s optics was effortless, as comfortable as slipping off sunglasses, which I also did not experience in other systems. It felt natural.
Is he a shill? Like folks have said here. . it's really hard to deliver. . if it wasn't, every nerd with an idea would be making a billion bucks selling us our dream come true. But this article is painfully missing facts, and sloppy with f, u, and d.
"Presidents don't make tax law. They give a suggestion to Congress, and Congress does whatever the hell it wants." CITATION NEEDED-Don't look in the Constitution, or the news, neither will be helpful. The Republicans regularly threw hissy fits and shut down the government because Obama (and Clinton) wouldn't do whatever the hell they want.
Here's an interesting rebuttal to the idea that it is "another Theranos".
Some key points:
- the product that was the source of this report "was not the Magic Leap's latest prototype"
- the investors that bought into Theranos were "rich individuals whose life sciences experience began and ended with high school biology", but the Magic Leap ones were âoesending their brilliant professors from all the top schools to try and shoot us down.â
I want Magic Leap to be real because it sounds cool. I'm disappointed that they (apparently) faked one of their demos but there are several really positive reports about it from fairly reputable individuals who have actually tried it - so I live in hope.
Magic Leap's mixed reality technology has long since been overtaken by other products already on the market such as Microsoft's HoloLens
HoloLens has almost exclusively used fake demos from the very beginning. People who have actually used HoloLens report poor field of view and semi-transparent graphics, yet the demos all show perfect wide-angle non-transparent graphics that have clearly just been composited over the video signal. Magic leap tried a similar trick for their first demo (with the steam punk ray guns) but all the subsequent videos did appear to be shot directly through their device. Of course, we never actually saw the device, so it could have just been their unwieldy and unwearable prototype. The only new information here seems to be that Magic Leap are struggling with miniaturising their scanned fibre display, but that is quite a serious issue.
The site blurs the text and says you need to disable your adblocker to read articles. To get around it just remove the "color: transparent;" property from the style attribute of the "v_main" div.
Read. 110010001000 is a troll, but this statement:
no more large leaps like we have had for the previous 40 years.
Does not mean:
there will be no more large leaps in 40 years
as you bitch about.
The statement means that we will have no more (ever, within your lifetime, whatever) large leaps like the large leaps we've had throughout the PAST 40 years.
if re-written carefully would achieve speedups well beyond a couple Moore's Law generations
Yes, immediately offset by bugs and other wonders that result from thinking you have a better way.
You are right, software has a lot of room for further optimisation. But as much as we like to heap crap onto lazy software engineers who do nothing but copy and paste or implement standard high level libraries without advanced use of languages, it's also how bugs are avoided from re-implementation.
MS bought their tech, and did the ol' embrace, extend, rig their demo, extinguish.
Hmmm, there should be a Profit! in there, somewhere...
WARNING: Smartphones have side effects--most of them undocumented.
Oh wait, it was