Why do people keep giving Magic Leap money? | Hacker News Reader

Why do people keep giving Magic Leap money?

theverge.com

62 pointsrdoherty59 comments

I just can’t convince myself that AR is the next smartphone revolution. Most of the applications I have heard about feel like the iPhone X demo. Kind of cool they can do that but will I use that more than once to try it?

In addition to the challenges mentioned in the article, you have to convince people who are not wearing glasses that they should start to. And people who are that they should swap their super-light glasses for something heavier, that may be painful for their nose or neck. And that they need one more high tech device in their life to budget for, recharge and replace.

Then what are you really going to use it for? Giving you some meta data on people you see? Name would be nice but how many times in a week do I come across someone I sort of know but can’t remember the name and I would need it? Not many. Identities of complete strangers in the street? Outright creepy. Getting their birthday or the name of their dog? I don’t want my life to look like my linkedin feed where I am flooded with useless notifications about my 1000 contacts work anniversaries.

I am sure there will be some applications, mostly industrial / professional. Certainly the sort of thing you will want to hand over to people at a conference or networking event, where this metadata is super useful to make the right contacts. Or for shopkeepers to recognise their customers in real time and having a full history of their previous purchases. But as a mainstream consumer product, I am still waiting to hear about a convincing use case.

Google Glass was/is AR (information overlays). Magic Leap, HoloLens do Mixed Reality. The difference being that the synthesized objects behave as if 'attached' or otherwise interact with reality.

But isn’t it even more of a corner case then, basically certain games? Or the ikea “how does that sofa fits in the room” example (which I might use once every 5 years...).

More go for a run in the rainforest, walk down the street and see people in the latest virtual fashion, sit in your empty room and see it as opulent luxury, etc. Im inclined to think we are over a decade or two away from that (at least to the extent that the lines of reality would be truly blurred), but I understand thats what Magic Leap is hyped to deliver.

None of which will really work without direct neural interfacing. And we're probably three or four decades away from that.

Glasses and screens are always going to feel clumsy for AR/MR. The technology would need to shrink/improve by at least one and preferably two orders of magnitude to become unobtrusive, and that's going to be extremely difficult to do.

> sit in your empty room and see it as opulent luxury

That sounds incredibly sad.

Looking at how so many people use social media, it also sounds like it may be incredibly accurate. And that makes things even sadder.

Once the resolution, tracking, energy usage, form factor (...etc etc etc) is good enough these devices will be able to replace every screen you use.

> The difference being that the synthesized objects behave as if 'attached' or otherwise interact with reality.

That's also AR. MR is an umbrella term intended to cover both AR and Augmented Virtuality (which is basically injecting real-world objects scanned by a Kinect 2 or whatever into the right place in the world as viewed through a VR headset, that kind of thing). http://etclab.mie.utoronto.ca/people/paul_dir/IEICE94/ieice....

Are the terms being redefined now? AR used to mean all you described here - mixing synthesized object with reality in a context-dependent manner.

I suppose I could welcome this change, if it keeps people from putting Pokémon Go in the same bucket with Google Glass and HoloLens...

The most convincing use case that I've seen was the 'Find stuff in Home Depot' example that was maybe shown at Google I/O 2017.

Pretty much everything else about AR seems like hot air.

How much money is that one convincing use case worth?

I work in civil engineering and lots of people here are really interested in the potential of AR. Just something as simple being able to stand in an empty field and look around and 'see' where all the existing pipes are underground and overlay that with where the new water pipes are planned to go makes lots of people here very excited. If you can then 'move' a pipe in real time while in the field and have the CAD drawings back at the office update then that could be a real game-changer.

Basically don't look to Google for the 'hot' applications of AR. Look to companies like Autodesk, ESRI, Tekla, Bently etc.

As a specific example, best Microsoft's Hololens AR use case demos have been about construction and logistics. Those are not gimmicks invented by marketing but actual use cases that professionals in the field feel would be value adding.

The 'seeing the hidden pipes' is a good example.

Another is facility maintenance. The savings are not insifnificant if you could just guide maintenance teams to the problem spot using HUD navigation, instead of finding the broken gizmo after some search.

Completely agreed. I could easily spend an hour talking about all the potential real world uses for AR I see at work and non of them would have anything to do with consumer smart phones or Google/Apple's target audiences.

I've always felt that out of all the AR/VR companies that exist today, Microsoft will be the ones who make it ubiquitous. The average consumer is not the group who is able to bankroll mass production. The money Microsoft makes from selling expensive Gen1 devices to industry and professionals is how they will be able to R&D into cheaper units

>Name would be nice but how many times in a week do I come across someone I sort of know but can’t remember the name and I would need it? Not many.

I have a hard time remembering names, so this would be incredibly useful to me.

Same here.

Not to mention that even a half-assed facial recognition tech would probably be better at the task than my brain is currently.

I guess there is the possibility of putting porn everywhere...

I strongly recommend reading "Rainbows End" by Vernor Vinge.

It's a science fiction novel about a near future where everybody has AR contact lenses and can see the world the way he/she chooses (among other consequences of this technology).

I love the book but I don't think it makes the case for AR at all.

IDK much about magic leap. That said, they do no appear to subscribe to the "ship something and let users drive" school of tech companies.

So first, the financing environment is different than than Generation Google's. By this size, Google had (arguably) the proven-best online ad business, the (inarguably) most impoirtant web page, and a big deal IPO. Magic have that kind of money in stealth mode, with no real reason to speak to the public. Regardless of the "how we build things" philosophy, Google's generation simply could not have got to such a scale before interacting with consumers. There just isn't a precedent because this environment is new.

I think what you're asking is questions that Google's generation would have got (based on previous generations' schools of thought): what is AR's "killer app," the lotus notes to it's personal computer. Before those existed in users' hands, home computers were pitched as ways of organizing personal finances and recipes.

Anyway, I think you're right. The hows and whys of AR becoming important are all very vague. I suppose that most of Magic Leap's pitch (to themselves and investors) is centred on the technology itself, not its application.

The article suggests investors think of them as an intel, not an apple.

Whether or not that's a mistake, it does leave the most interesting (to us) questions still open. What is AR for?

> The article suggests investors think of them as an intel, not an apple.

So like Intel in the sense that they are a chip fabrication/mass manufacturing company that happens to design chips? Or the new Intel that wants to "add more value" with things like face detection in Windows hello?

The first, kind of.

I'd say that it's the intel that's knows what is valuable and defensible (not RAM) in a market that will be big. All they need to know is that computers need the chips intel makes, what moats they have (eg x86) and what measure they need to continuously improve on (Moore's law). This intel doesnt care if you want to use computers for recipes, ml or world of Warcraft. Doesn't care if you like Mac or PC. Doesn't need a killer app.

That argument would hold more water if they shipped something.

They still might

If (a big 'if', ofc) you handwave all the display, compute and cost issues better, AR is a no-brainer even simply as a better alternative to tablet and smartphone displays. Screen-on-a-handheld-box displays are trapped in inescapable trade-offs—screen too small, device too big, screen too close to the eyes, choose one; screen at a low angle, device hanging from a fastener contraption, auxiliary display in use, gorilla arm, choose one—that HMDs are not trapped in.

Going beyond that to 'real' AR applications, maps applications alone are enough to justify AR. Turn-by-turn on-foot navigation and local points of interest are use-cases which cry out for head-up display instead of forcing the user to divine information from a small handheld map (and audio cues etc.) Alternatively the user could do flatscreen AR by gorilla-arming their phone, but it's obvious why using a sufficiently good AR HMD is preferable to this. The big extra stumbling block here is once again technology—the coverage, accuracy, and speed of tracking with GPS etc.—not inherent uselessness.

UI killer, never take your phone out of your pocket again. Compatible devices in your home are inspectable and controllable. Even dumb objects could have a UI if they can be recognized. Your toothbrush could show usage and expiry info. Houseplants can remind you when to water them. You could have a large marker indicate where your keys are. Being able to annotate anything with dates, reminders, instructions, etc.

There's a lot of information and calculations we carry around in our heads that could be offloaded or supplemented.

I think these are great examples of technological aids that people really do not want in their lives. Our brains evolved long ago to a point where we can do many simple things concurrently, and there is a power/meaning in interacting directly with your environment - not through some smart aid that tells you that your plant is dying.

When you share your life with other people who knows when the plants were last watered, was it me 2 weeks ago or my wife yesterday? Watering plants is a silly example, meds, health, pupils in a school, patients in a hospital. All much more interesting to have meta data on.

Yeah, sorry the plant issue is a bit of a silly point. I just meant that you can tell by looking at the plant, feeling the soil that it needs watering or not. So many of the use cases I see for smart home devices seem contrived. "What can we do with this amazing technology?".

I agree voice-controlled devices are a great innovation - particularly for accessibility. But as a mass-market appeal, I've met a lot of people that love their smartphones, new technologies, but aren't sold on home automation. The big enduring technology successes in recent decades captured people's imaginations instantly. They didn't need much convincing beyond the initial idea.

A separate objection is that by not using your brain for daily routines, memory and organisational skills will suffer. Purely speculative, but I think people downplay the importance of menial tasks in giving people a daily rhythm and discipline. The argument I sometimes see is that it frees you up to do higher-order tasks. I think it's unlikely.

> Yeah, sorry the plant issue is a bit of a silly point. I just meant that you can tell by looking at the plant, feeling the soil that it needs watering or not.

A recent dicussion on HN about that guy who crowdsourced watering of his plant generally suggests that a) no, for many, many plants it's not that easy (especially those not in their natural climate), and b) people do routinely screw this up.

I for one vote the metadata-on-plants up; that would provide me more value than 99% of things the startup economy is producing... combined.

> So many of the use cases I see for smart home devices seem contrived. "What can we do with this amazing technology?".

I agree. But I feel this is because we mostly see the marketed ideas, which are optimized for selling products, without any regard for whether they're useful or not.

Some things I'd love technology would help me with, which are hard engineering problems, include:

- tracking expiration dates of foodstuffs without requiring me to do additional manual work, and displaying them without requiring me to click much (or preferably anything)

- tracking items around the house, so that I can find any misplaced thing by a simple query

> A separate objection is that by not using your brain for daily routines, memory and organisational skills will suffer. Purely speculative, but I think people downplay the importance of menial tasks in giving people a daily rhythm and discipline. The argument I sometimes see is that it frees you up to do higher-order tasks. I think it's unlikely.

I don't know. Myself, I can't stick to pretty much any daily routine, so that part of my brain is already broken (and always has been), and I appreciate any crutch technology can provide (the best, still, is living with a person whose "routine" part of brain is working and that will remind me about menial things to be done).

And - maybe that's because of my broken routine-brain - I find "daily rhythm" to be a soul-destroying concept. If you want to put your life on autopilot, instead of outsourcing repetitive tasks somewhere, why live at all?

Good points there. The autopilot thing - I don't see it like that at all. You can be doing a repetitive task and thinking at the same time. It can help you focus and relax. Also when you finish the menial task, you get a little smug satisfaction. "I've earned that beer". Autopilot for me is sitting on the couch asking your electronic friend if the fridge has cleaned itself yet.

Many favorite hobbies are repetitive and in theory could be outsourced. Some people enjoy knitting, raking the yard, organizing their record collection, finding Waldo etc..

Hey, it's all down to personal preference and where you get your kicks. I won't labour the point as I've gone on too much. Just putting forward another perspective.

Fair enough :). I do find myself oddly relaxed when doing some of the menial stuff, especially when that stuff can't be easily automated away (if it can be, then tend to think about how to automate it instead).

I suppose I blended two distinct topics too much in my response. The whole dislike for going on autopilot through life came from various attempts at enforcing a structure on my day. At some point I always start to feel a strong discomfort for such routines. These days, when I plan days out, I treat those plans as kind of sane defaults, things to do if I don't have anything more important, and not as absolute commitments.

That same reasoning would probably tell you that no one would want a "box in your home that you can talk to", yet Alexa and its competitors are doing quite well.

AR re-skinnable pets.

I want my chihuahua to have horse hooves.

And a dragon mask.

I want my manager to be a poop emoji.

Big opportunities in the social cyber bullying space.

That's only if you add a "Share" button to the personalized entity replacements...

Is this a thing you seriously want in your life? Because to me that feels like the definiton of "Kind of cool they can do that but will I use that more than once to try it?".

It's interesting and fun the first ten minutes, after that it'd be annoying.

Nothing stops this stuff from happening now on a different UI e.g. A phone. So if most of that info/control is not important/wanted on a phone, then why AR?

I would argue that the bad UX of current phones and apps is a big obstacle to this happening now. Having to take your phone out of your pocket, unlock it, open an app, point it at the object, wait for recognition, wait for data/processing, and only then see the result is a huge negative that demotivates most people from wanting this sort of experience on their phone.

With an always on AR system with specialised hardware and software for object recognition and good depth mapping and hopefully a design that allows for multiple AR services to run and provide overlays concurrently - you should approach the level of usability that will allow this to happen.

It’s not the future. Most people just don’t want this on a daily basis.

I imagine this is going to go like VR did. Last year we had a nearly violent vocal minority saying we would be fully immersed by now and anyone who disagreed got pounded on. AR will be hot for a few months and then basically find a niche or two and live there forever.

> What are you really going to use it for?

Names, as you suggest. I don't know how old you are, but when I was in my 20s I could easily remember the name of everyone I'd met. Now, in my 50s, I know more people, and meet more people, than I can keep up with. I'd love a popup for not just friends, but friends of friends.

Addresses. I can't tell you how many times I've walked past a poorly labeled building I was looking for. Street addresses, street numbers, and names of buildings.

Construction. This morning I drove past a construction project by my house. The city is doing something to a little park in my neighborhood. What? No idea. Demolishing it? Improving it? Carving out an easement for an extra lane of traffic? It'd be so awesome to point my phone, or goggles, at the construction and get more info.

Metadata of all sorts, much of which is available now but inconvenient--Yelp reviews of the restaurant I'm standing in front of, hours of the government building that's closed at the moment, historical information about a landmark, etc.

Basically, a context menu for the world. I think there's no question that that would be valuable. The pain point, for now, is the interface.

"It’s producing a regular stream of patent applications for augmented reality glasses technology, which everyone seems to agree is going to be important at some point".

If they own the patents that ar companies must use and ar is "the next big thing", maybe that's why people keep giving them money. They don't really have to produce anything more.

If true, that's incredibly depressing.

Magic Leap would become the complete rent seeking lead weight to the next platform revolution.

Exactly. Let's not forget that "Google bought Motorola for $12.5B, sold it for $2.9B, and called the deal ‘a success’", purely for Motorola's patent portfolio.

I haven't been keeping up with Magic Leap (ML) or recent Mixed Reality developments, but skimming the currently available information, my best guess is that Magic Leap is making a completely new system from hardware, to software and even content. Sort of like back in the Apple Lisa days, pre-1984 when the Macintosh was unveiled.

I've only heard of Microsoft and Avegant (that I just googled) working on Mixed Reality. The problem with these two and anyone other than Magic Leap is that they're only working on making the tech. They're not making a complete platform. Of course they could leverage existing platforms (PCs or Consoles) but then the interfaces would mostly be translations of them. Not that ML has shown anything better, but I get the sense that what's in development is that better platform. From history we can see how a platform arises. Imagine planning the iPhone/AppStore and building iOS and starter content for it. It's going to replace current movies, video games, and computer systems.

Oculus is making a complete platform but their bet is on VR not MR. It could likely be adapted to MR if that gets better traction, but without blatent copying of the human interface it would be a sub-par experience.

Anyone know if Microsoft or anyone else is actively developing an MR environment for market?

> Anyone know if Microsoft or anyone else is actively developing an MR environment for market?

You mean like the Hololens?

https://www.microsoft.com/en-us/hololens

Yes, but I was asking specifically about things other than the hardware and individual apps, like is there a flavour of Windows specifically for the HoloLens and how significanly different is that UI?

https://en.wikipedia.org/wiki/Windows_Mixed_Reality

I think you are using the word "platform" when you mean "user interface". User interfaces can be changed to suit the task at hand - e.g. MS Word vs Doom. vs Job Simulator vs Star chart

What is still missing is a good interface. Not this sticks with buttons, something that allows comfortable control and feels like a keyboard without requiring constant tension to hold.

It should be like a glove, with the fingertips as selector and the other fingersegments as keys. Thumb to tip is a-e, thumb to indexfinger segments up is systemkeys (sorted by need to avoid difficult to reach segments as being quertyfied. Thumb to third indexfinger segment (with indexfinger pointed) is click at a object.

The Palm could work as a sort of gesture-touchpad, for more complicated (cross-hand) keycombos.

How to implement? Well as a prototype- take a latex glove, glue short range rfid chips to the segments and rfid readers to the fingertips.

The rest is history as they say.

Thought about it- the select (leftclick on the mouse) should be the most natural point to something gesture, so its indexfinger pointing and thumb-tip touching the midfinger-tip. Also RFID is not bad for precise measurements. Problem is haptics, selection safety (visual display in AR, and user-comfort. Any glovelike device (even if its just a fishing net of cables and sensor) will feel uncomfortable after a while)..

The good thing is there are really a lot of comfortable combos possible with just the fingertips touching one another.

I previously posted a link to a guy's blog post where he did an examination of the patent filings and what the technology was likely to be able to do:

Discussion: https://news.ycombinator.com/item?id=13653537

Blog Post: http://www.kguttag.com/2016/12/06/magic-leap-when-reality-hi...

Sunk cost fallacy

Same like bitcoin. Even though it has no realworld use in its current form, it might have in the future so people speculate on that.

Not sure if it's a right analogy. Eventually bitcoin could be traded as an instrument but Magic Leap's case could be never be like that since if they're bankrupt no one would come forward to buy their bonds. People bet on bitcoin just like they bet on Gold or any other entity. People bet on Magic Leap with the hope that Magic Leap could be provide some magic - that eventually could be galloped by AR-hungry companies like Snap, Google, Apple - since all of them always want to beat FB - which already got Oculus.

Bitcoin has limited use, but I wouldn't say no realworld use. There are some very real places you can spend it, and some provide anonymous advantages (like VPN access). It's also seeing uptake in use for remittance and combatting hyperinflated markets.

Bitcoin has a use now, people use to store their wealth away from prying eyes, I think this is why China recently tried to ban it....

Magic leap could be the future, it could be complete vapour ware, though I like to be positive about these things as they could be doing things the Apple way and will only announce finished products, not technologies....

Reply on news.ycombinator.com