[aresant]

If there's a bull case on FB to me it starts with the release of this Facebook Spaces beta.

Spaces - or something like it - along with an eventual great set of AR glasses or contact lenses - is the end game.

There's no question the long goal for corporate AR is to digitize the entire physical world, catalog it, and then sell outrageously effective ad inventory / flair against everything that we see.

Using a fairly controlled VR environment as the beta case for this to get us all hooked is a huge step in the right direction and they already own the entire social graph to execute in this direction.

I don't know whether I should be excited or terrified that it's FB leading this effort.

They have the scale to execute, they have the technology to support the crucial relationships but they are SO FUCKING INVASIVE into our lives as a company.

Raph Koster's lecture at GDC got some fairly broad attention on this concept, although his was geared more to the potential negative consequences, but it's still 100% worth a watch for anybody interested in the space.

https://www.raphkoster.com/2017/03/05/slides-for-still-logge...

[RangerScience]

>> There's no question the long goal for corporate AR is to digitize the entire physical world, catalog it, and then sell outrageously effective ad inventory against everything that we see.

From working in industrial AR for awhile, this is a tiny, tiny tip of the iceberg. The (actually very reasonable) buzzword bingo is smart contracts + blockchain + reality capture + HMI = rather a lot more than better advertising.

[str33t_punk]

what do smart contracts and block chain have to do with AR?

[RangerScience]

AR is a human interface (transparent head mounted display) plus reality capture.

HMD lets humans interface with the world of the computer. Reality capture lets the computer interface with the world of humans.

Reality capture feeds smart contracts; if the AR cameras see you're low on some stock, "the computer" can order more, and direct the people to pick them up. Smart contracts are a reasonable mechanism for doing that.

Block chain is a shared ledger that's suitable for multiple entities tracking anything (say, reality) together, and a communications medium that's good for things that need to be auditable (say, tasks).

[throwaway29292]

I agree with all of your points, except the last one. I don't think a shared ledger is required for interacting between multiple entities. We basically want a database that is beyond the control of any corporation - seems to me that a government-controlled database would fit the need.

The associated complexities and limitations with blockchain ledgers are a lot. The nodes must be propped up by large amounts of computing power, which only big companies/groups can provide, not individuals. And it's not as if the government can't be trusted, there are so many stock market indices & financial markets that aren't under anybody's control.

[DonHopkins]

Do block chains necessarily require wasting as much electricity as bitcoin mining?

[RangerScience]

No. Think of blockchains as formalizing consensus reality. Each block, some "lucky" processing node (miner) somewhere gets to determine what that consensus actually is. The various techniques (proof-of-work, -stake, -etc) more or less make it hard for a bad actor to be that "lucky" node.

BC does this by letting everyone submit answers, but making it really hard (computationally intensive, aka, burn electricity) for your answer to the be the one that's accepted.

PeerCoin (to pick one) does this by making your likelihood of being The One correlate with your investment in the worth of the coin (literally how many of the coins you control). If you're the One, either your answer is accepted, or you're selected to give the answer, and then it's accepted - I'm not sure which.

You can also limit participation in your blockchain in combination with these; for example, a closed (invite-only) bitcoin would still burn electricity, but not nearly as much.

[mathewsanders]

He mentions early in the talk about how some people working in VR/AR make a comparison that they're building the tech from 'Snowcrash' which he calls out as wrong model for thinking about VR.

A while later he describes a companies responses to building features after a player virtually groping another player in a virtual archery space.

The model for AR/VR that seems most plausible to me is what's described in the book 'Lady of Mazes'[1] - really recommend people check that out if they find this talk interesting.

[1]: http://strangehorizons.com/non-fiction/reviews/lady-of-mazes...

[cableshaft]

Thank you for sharing Raph Koster's talk. It gave me a lot to think about that I never really thought about before. I have a feeling this Spaces tech doesn't include a whole lot of what he's advocating for, and it probably should.

[stagbeetle]

> There's no question the long goal for corporate AR is to digitize the entire physical world, catalog it, and then sell outrageously effective ad inventory against everything that we see.

That's too far away to be able to predict accurately and it all rests on the lynchpin of the masses being interested in augmented reality.

Although slightly different, this was the case when computers were brought to the mass market. People didn't drop their real lives to become fully immersed in the digital world. Sure there is the minority of societal outcasts that spend their entire waking lives immersed, but that's all they are: a minority.

Of course, most things are on the uptrend to becoming "digitized" in the sense that everything you could ever need is online, but it's been a very slow course.

The average Joe barely even comes close to using the internet to its fullest potential. What makes you think something as complex and convoluted as AR (not to mention expensive) will be any different?

[aresant]

"The average Joe barely even comes close to using the internet to its fullest potential. What makes you think something as complex and convoluted as AR (not to mention expensive) will be any different?"

I think that AR will become as ubiquitous as our phones because all of the core functionality on a phone will be better on AR.

Instant alerts in my peripheral vision (again a non-obtrusive display is crucial)?

Call up directions while driving with cortana / siri and have lay over unobtrusively?

Watch TV on a 100" screen anywhere?

All things that I've done / played with in Hololens demos, all amazing, all "no question" will be massively adopted when they get the form factor correct.

Heck just look at a current app like SNAP - already on 50%+ of phones of 18-24 year olds - is a perfect app use case.

Camera, stories, sharing - all better and more accessible in AR.

And they're already demonstrating the desire / utility of AR via "filters" and their new "World Lenses" release today.

Want that bunny rabbit filter on all day as your "look"? Done.

Want it only accessible to your co-workers and not your boss, cool set it and forget it as long as the device knows your social graph.

[stagbeetle]

You've only proven my case.

That's a lot of functionality, and there is a market for this functionality, however this market isn't the mass market.

Like computers, phones are also barely used to their full capacity. There is a quote about how we have the most powerful tools and the largest repository of information all at our fingertips, yet we concede to our whims and spend it on frivolities (see: usage of social media and mobile games vs. non-entertainment apps).

The regular consumer does not give a shit about any of this supposed functionality of AR. You can tell them "oh it'll have this, and instant alerts will be unobtrusive, and the UI, it'll be optiimized to hell and back!" But, these are weightless promises. Consumer phones haven't even gotten to the point of seamlessness and great UX (though Apple is coming close).

Snapchat is simple, AR is most definitely not.

This is an adoption problem. You can't keep throwing features at it and believe the problem will fix itself.

[wlesieutre]

You don't need customers to "use AR to its full capacity" though. You need them to use one or two pieces of it enough to justify the device.

Sure, the early iterations might be more like Blackberry popularizing email-on-the-go for a small segment of the market. Not everybody needed that. But by now pretty much everyone has found at least one aspect of smartphones that they've come to depend on.

Does my mom use her smartphone to its full capacity? Certainly not! There are thousands of apps she's never even seen, some of which can do pretty great stuff! But she checks her email, organizes recipes, and takes pictures of her dog, and that's what she has the phone for.

Heck, I don't either. If anything I'm trying to use my phone for less stuff, but I'm still glad I have it.

[stagbeetle]

Then this goes back to the "what does AR offer that a smartphone doesn't already?"

"Are these features enough to get people to switch to AR?"

From the sound of it, AR is just a mobile OS port.

[wlesieutre]

And mobile OS is just a Linux port. The most fundamental difference is how available the device is. Computers are something you have to sit down and use, smartphones are something you have to take out and look at, and AR could be something that's immediately present whenever you need it.

Beyond being just a more available HUD, the AR aspect should make for more direct interactions with the real world. Lighting controls come to mind since that's the space I work in. What if instead of having light switches built into your house, they became a purely virtual construct that each user places into the world?

Right now, IoT lighting feels clunky because you walk into a room, take out your phone, open an app, pick a light, and move some sliders. What if you could just drop an imaginary button on the wall by the door for each preset, or a virtual knob for dimming? It's the exact same functionality, but putting it into the relevant physical space makes it a lot more useful without requiring additional hardware for every room. And if your guests have AR glasses too, then you have a default "guest layout" for your house that's automatically available to anyone on your wifi. They get access to the light switches, but they can't unlock the front door.

Plus you don't have to worry about replacing switches in every room if you replace the lighting system. Just the one HomeKit hub (or whatever it is) and the switches are completely imaginary.

I haven't spent a lot of time imagining where AR will go, that's just one thought off the top of my head.

[Applejinx]

>Call up directions while driving with cortana / siri and have lay over unobtrusively?

It's the future. Why would you be driving? You'll be in a self-driving car, bored. That's AR time. Isn't it?

[ReverseCold]

Open source distributed federated VR spaces? Yes.

[j_s]

Video of the talk: http://www.gdcvault.com/play/1024060/Still-Logged-In-What-AR