Magic Leap Review Part I | Hacker News Reader

Magic Leap Review Part I

kguttag.com

166 pointsUpshotKnothole70 comments

Open articleView on HN

I think Fernando Serrano captured it perfectly: https://twitter.com/fernandojsg/status/1017411969169555457

Make sure the audio is on; Absolutely brilliant.

I don't get the audio reference; what's that music from?

Amazingly hilarious. It's Jurassic Park OST. Starts off as an OST and rapidly degenerates along with the graphics.

The video is a reference to this No Man's Sky parody video: https://www.youtube.com/watch?v=RvAwB7ogkik

It's a genre of video where a shitty harmonica version is dubbed over the original theme.

I believe the jurrasic park one is the original;

https://www.youtube.com/watch?v=-w-58hQ9dLk

"They do move in herds"

Wow, I had no idea it was actually that terrible!

The fact that 85% of the normal room lighting is blocked by the lens means that wearing it would essentially be like wearing sunglasses.

On top of that, the diffraction grating also looks like it introduces a bunch of visual artifacts like smearing, ghosting, and rainbows.

I kind of assumed all along. I was crying vaporware the whole time though - I'm surprised they released something. The signs were there in the marketing - they never showed anything other than pretty cg mockups the entire time, the only videos they showed of functionality were clearly fabricated mockups - even < 1 year to release date. The closest they ever came to actually demoing the thing was showing some celebrity using it.

If they actually had anything nearly as impressive as they were claiming they wouldn't acted so obviously shady.

Despite this, I think they still deserve some credit for actually releasing something. I was on the Magic Leap hate train, but now that they actually released their gimpy headset I'm actually kinda rooting for them.

Theranos also released 'something'...

Not really- "releasing" implies a regular person could buy it... No human being outside of Theranos was ever allowed to be in the same room as an Edison machine.

Keep in mind that this author is a long-time Magic Leap critic. He's not exactly looking hard for silver linings.

Are you seeing a “silver lining” he missed? If not, don’t you think it’s possible that he’s been a long time critic because it was crap and exception from the get-go? If not, I think it’s worth remembering that this critic correctly predicted all of the major problems with the device months (and longer) before it was released. He called the “lightfield” and other marketing dishonesty out quite a long time ago, and seems to have been batting nearly .1000.

So... what do you see that he missed?

I understand that you’ve been very supportive of this for months, but unless you’re invested in or affiliated with Magic Leap it might be time to let it go in the face of the reality.

Yes, I've watched a few long form video reviews of the ML One, and I've generally been impressed with what I've seen in the videos. The reviewers have generally seemed impressed as well. They usually say it could use a bit of work before being unleashed to consumers, and that it's a long way from fulfilling all of AR's potential, but being that it's explictly a dev kit of a V1 of potentially a whole new industry, I think that's all acceptable.

This reviewer has clearly had an axe to grind with ML's claims for years, and this is a continuation of that.

EDIT: that doesn't seem like the same comment I replied to, did you significantly expand it?

I think you can quibble about the author's bias, but Magic Leap has so far failed to deliver on the promise that its early videos promised: the ability to seamlessly blend the real world with a computer-generated one.

Other systems in the same space (like Hololens, the Rift and the Vive) didn’t put such simulated imagery front-and-centre in their marketing materials.

They're not quite there yet on their miniaturized consumer hardware, but have you watched the longer form videos? What they've pulled off looks like a very strong start toward that.

The Rift and Vive are not in the same space as Hololens/ML at all, they shouldn't be lumped together.

You either didn't notice or forgot the bad press Hololens endured for misleading advertising:

https://mashable.com/2015/10/06/microsoft-hololens-misleadin...

https://www.theverge.com/2015/10/6/9465839/microsoft-hololen...

Going back further, Microsoft has a history of this with things like Project Milo on the Kinect. These things really don't matter in the end though. Its the technology itself that matters, and how good or bad it is relative to the market.

“batting nearly .1000” would mean he’s right almost 10% of the time. I think you meant “batting a thousand”, which would be written “1.000”

You’re right, thanks for the correction.

The silver lining is that devs actually really like the device. Despite the hype/reality factor that tainted the initial reviews the consensus is that the device itself is actually quite good. If all you read are the negatives, especially hit pieces from Guttag then I could see why you have the point of view you do. In fact you'll be hard pressed to find any reviews outside of Guttag and Palmer Luckey who state the device is bad:

Hololens developers:

https://brekel.com/category/magic-leap/

https://medium.com/futurepi/magic-leap-one-in-depth-review-b...

AR/VR reporters:

https://twitter.com/kentbye/status/1038927438279888896?s=21

https://uploadvr.com/heres-what-actual-developers-think-of-t...

New content is getting praise:

https://www.theverge.com/2018/9/19/17876488/angry-birds-magi...

And this is just the start. Their dev conference is in a week. Lots more to see and I don't think they're going anywhere for a very long time.

I'm sure you think that shows bias, but you could also say that his track record of being right means he should be considered extra credible. :)

>The fact that 85% of the normal room lighting is blocked by the lens means that wearing it would essentially be like wearing sunglasses.

As part of a costume I bought some $4 sunglasses at the thrift store and wore them in a dimly-lit bar the other night. Ended up bumping into people because I couldn't see.

I imagine this is what Magic Leap is like.

now u just need Magic Leap Autopilot :)

Each one comes with a free smart contract.

Well, of course it has to block most of the room lighting. They can't draw dark. The display has to be brighter than the ambient light. I'm surprised these things don't have an auto dimming shutter, like a welding mask, to adjust how much outside light gets through.

Well, there is this rear view mirror in cars which essentially draws dark by blocking out the area where the sun would reflect into your eye.

Expanding on that, maybe with some kind of holographic shenanigans, we can have something that not exactly draws dark, but occludes light very cleverly.

Like an LCD of some sorts.

Exactly my point. :)

The update rate has to be very high and the alignment perfect or it will drive users nuts. Which means it will probably be done soon, although not in this round. Doing it with lightweight short path optics is also difficult. All this stuff with diffraction gratings and Fresnel lenses makes for a crappy image. That can sometimes be overcome; the Polaroid SX-70 camera had a very compressed optical system with Fresnel lenses without compromising image quality too much.[1]

The other option is simply to show the real world via a camera feeding into the system. That's been tried many times, and it works OK, although lag will make users faceplant or ill.

[1] https://www.technologizer.com/2011/06/08/polaroid/sx70-light...

OMG I so much want an SX-70. I will make it happen sooner or later now that there is film yet again for it.

Well actually you can, it's called an LCD

The problem is focal conjugates; you can't darken something close to the eye and have it look like a dark region at a distance without focus blurring ruining the effect.

To be honest it doesn't look that much worse than the Hololens and the Hololens was perfectly fine. Not perfect, sure, but technology has limitations. People still buy LCDs even though they reflect ambient light, don't work outdoors, etc. I don't see what the point is of complaining about having to look through a diffraction grating if we have no other way of doing it.

No, hololens just isn't good enough. Won't see market traction, because it's too unrealistic, the immersion factor too low.

There's a minimum bar of realism we need to hit to have reasonable AR and VR. The original Vive hit that bar for VR, but nothing has hit that bar yet for AR.

Sure but that's because pretty much only because of the terrible field of view. Not because you're looking through a diffraction grating.

The worst part is that Hololens is now almost 3 years old and is going to ship a new model next year (rumored)...so Magic Leap is also so so so far behind. Microsoft is smartly pushing Hololens for industrial uses, where Magic Leap plans on military contracts? There is about a 0% chance the military will be using AR in any official capacity in the field for a while.

This kind of reminds me of Leap Motion, which is a very cool tech, but I could never find any practical use for it....at least Leap Motions are cheap.

Some of the frustrations I've seen of people using ML1 remind me a lot of LM, specifically the lack of uniform interface between applications.

This was reported just the other day:

"The original HoloLens has been out for a long time now, so a new (hopefully cheaper) version is overdue, but it now looks like we won’t see it before 2020."

https://www.techradar.com/news/hololens-20-could-be-delayed-...

Mirror: https://webcache.googleusercontent.com/search?q=cache:WNHDj6...

Theranos also talked a lot about military contracts.

Endless pool of money with little oversight and nowhere near private-sector level expectations on results

> No amount of money or effort is going to change the physics of diffraction gratings.

On the other hand there is an amount of money and effort that will persuade people to pretend the physics has changed.

Is there a chance of proper text rendering with this technology, usable browser windows and such?

I think overcoming the laptop screen and enabling "portable workstations" (portable like a laptop, not phone) could replace laptops and improve desktop -> laptop workflows.

This is very impressive that someone would take the time to experiment an document so thoroughly this product. I am thankful because I find this very interesting and I always thought of Magic Leap as vaporware.

Discussion of part 2: https://news.ycombinator.com/item?id=18110225

The last time I compared Magic Leap to Theranos I was downvoted to -4 (https://news.ycombinator.com/item?id=16713546). It's interesting to see people's reasoning back then vs the product now, 6 months later.

There's a huge difference between a highly funded startup that is trying to build something completely revolutionary and hasn't quite gotten there yet, and a company actively and knowingly engaging in fraud in the medical space.

Nobody is going to die because of Magic Leap.

> highly funded startup that is trying to build something completely revolutionary and hasn't quite gotten there yet

That's exactly what Theranos was, until it wasn't.

Sure, but so far Magic Leap hasn't done anything illegal or unethical or immoral... they've just been a bit behind schedule.

They also aren't in the medical space. Nobody is going to die because of them.

With all of their secrecy, you have no idea if they're being unethical or not. Their marketing materials are certainly nothing near reality.

You're being completely disingenuous.

Magic Leap is a glorified toy. It is not going to affect your life.

I’d agree otherwise, but the article mentions that Magic Leap is bidding for a $500M military contract.

Putting these toy lenses on soldiers can get people killed — both the fighters themselves as well as civilians are at risk from these digital goggles that block 86% of incoming light unless the AR graphics can do something really, really useful in the battlefield.

I doubt it's for the battlefield. More likely for planning, where you'd have a briefing with several people around a tabletop projection of the ground scene and AR tanks and characters moving around.

FTA: “for use with troops in the field (as opposed to training).”

Not quite clear either way. You could still be right, but the implication seems to be otherwise.

Until they start faking results to the DoD, the analogy falls completely flat

I wonder what legal repercussions one would be open to for faking results to DoD.

If the F35 is an example, I’d say a trillion dollar, ten year contract?

The Theranos board is looking for work and just the right people to move such a deal forward :)

Seems like you're moving the goalposts though: the original question was whether Magic Leap is "unethical," not how much damage they could do by being unethical. Releasing much-hyped, highly-produced, and very compelling marketing content while knowing your company can't possibly deliver on it certainly meets my bar for unethical.

It's certainly not as bad as Theranos, sure - but bilking consumers (or investors) can still be unethical too. If your bar for a company acting unethical requires literally killing people, you'll be overlooking a lot of bad behavior...

> That's exactly what Theranos was, until it wasn't.

You conveniently skipped the part where Magic Leap isn't actively engaging in medical fraud.

You deserved to be downvoted for that because (a) it's a baseless accusation and (b) it was intellectually lazy.

Theranos committed crimes and their fraudulent behaviour directly impacted on people's lives. There is no evidence Magic Leap has done either and there is nothing illegal or immoral about pivoting on your product.

I agree with what you said, except, is making big promises then not delivering really "pivoting"?

Yes.

I got downvotes,which is fair because my comment was stupid because I didn't read the comments I was replying to carefully enough. I missed the "delete" deadline so sorry about the baseless noise.

Of course delivering shitty results isn't pivoting.

Of topic: who cares if you get upvotes or downvotes? Do people actually care what others think of their comments? Please, state the unpopular opinion however brashly you choose.

I see karma as a bank for paying for incredibly trashy comments: for every upvote I vow to write at least one thoughtless hot take and defend it to the death

Here take my upvotes :P take them!

Actually I don't think you're that wrong to be honest :)

Imagine comparing Theranos to a toy.

I don't think Theranos is a good comparison... but I don't think it's fair to characterise AR as a toy either, some people probbaly said the same about early incarnations of the technology that drove high resolution colour displays, lots of technology looks like shiny toys to begin with, but that's only because you cannot forsee it's applicaiton... AR is definately general enough to have very useful applications.

Reply on news.ycombinator.com