Michael E. Gruen

Augmented Reality is today’s Virtual Boy: it’s expensive hype no one will buy.

Technologists these days have been hard at work building 3D visual overlays, augmenting how you see the world. As it were, fanboys and fangirls have been hard at work telling us about our future in homemade videos; but, as technology advances, the real world will only get more real.

For the luddites and techies among my readership, please find the following video below and watch it as a point of discussion.

For the same reasons humanoid robots never seem to make the shelves of Walmart, this future vision (double-meaning intended) will never happen for the mass market—it’s too costly.

Let’s start with the cost of providing vision modification technology, enumerating by scenario:

Case 1: Camera Hacks. Some iPhone apps, like the Yelp application, have basic augmented reality features that overlay information over a video panel of whatever it is you’re looking at. The hardware cost is low, but the fact remains your augmented vision requires that you hold a piece of technology out in front of you like a goober. Offloading vision augmentation into a handheld device is clumsy and usually inconvenient; it’s a neat trick, but not much more.

Case 2: Super Glasses. Science fiction (e.g. Snow Crash, Accelerando, Caprica, Iron Man) often feature HUD-enhanced glasses that identify other people, overlay environmental information, or display text or video messages from others. Yet, fiction forgets that mobile embedded devices have (and will continue to have) issues trading off performance for reliable power. Modifying a scene in believable real-time 3D is difficult enough for an array of 3D rendering machines at Pixar, much less a pair of Ray Bans. The power and heat requirements would simply be too taxing to prove usable, and vision-augmenting would be limited to short bursts, not useful for regular wear.

Not to mention, glasses move around on faces throughout the day. The display would have to constantly correct for the minor, but highly sensitive differences as the glasses move around ever so slightly on the wearer’s moving head. And, like watching Avatar in 3D, you’ll develop a slight headache unless the optics are almost near perfect and consistent.

If—big if—you manage to mitigate these issues, how much is it going to cost you?

Case 3: Tiny Projectors. Imagine a micro-projector outfitted somewhere on where an image can be projected on your retina, fooling your eye into seeing things that aren’t actually there. Can’t imagine it? Neither can I—mostly for the reasons mentioned in Case 2.

Case 4: Optical Nerve Hacks. Imagine a device that could intercept the signal relayed from the retina to the optic nerve as it his the vision cells on the neocortex and offloads visual rendering and modification to a nearby machine, you still have to deal with the matter of bandwidth in rendering an enhanced vision for your neocortex so that it can make sense of it. But, if that technology were possible, why would you waste time, effort, and cost on only making things look more real or understandable. Why not make things simply more real or understandable at the fundamental level of understanding?…

…which brings us to

My Hunch: As technology moves forward, there’s little doubt that we’ll eventually find a way to make visual image enhancement commonplace. (Naysayers: thirty years ago, what if I told you that people would, en masse, elect to have lasers reshape their corneas, circumventing the need for glasses?)

If we’re at the point, as in Case 4, that we would elect to enhance vision directly to the neocortex, why not enhance the neocortex itself?

Strange as it may sound, the neurons in the neocortex that handle and make sense of your taste, your touch, your smell, and your sight, are identical. Instead, depending on what input they’re connected to, the neurons arrange themselves and adapt themselves to make sense of the signals coming into them.

Neurons are pretty neat in this respect. Watch how sensors on the tongue can help the blind to see.

Imagine, if you could, connecting a sensor to a portion of your neocortex (presumptively an area that was of very little use to you) and training your brain to make sense of that information coming in. What if it were a digital source, like the entire contents of Wikipedia?

As of January 2010, Wikipedia, including all of its images and all of its text, totals 2.8 Terabytes, or 2867 Gigabytes. If memory density increases 20% a year (as it has been) for the next 21 years, you’ll be able to fit Wikipedia into memory the size of the fingernail on your pinky. You could certainly fit a pinky nail underneath your skull.

So, if you could implant information directly on your brain and your neocortex could make sense of it, why would you need augmented reality? Your brain would do the work automatically. Say for instance that you, in 2010, wanted to look up “portmanteau”, you’d have to pick up a dictionary or type that word into the Internet, read the definition, understand the definition, and then apply it contextually. With a chip on your neocortex, you’d just know it. You would know it just like you can read this sentence without thinking too much about the character-by-character construction of its words. You would just know.

By the same token, when you looked at someone, you would just know their name. Or, when you looked at the Eiffel Tower, you would know when it was built, who designed it, who installed the elevators, and it’s mass in kilograms (or pounds) as easily as you see that it’s colored dark brown.

With deep vision into everything you were looking at, why in the world would you need something as crude as a live-drawn diagram to tell you how to make a pot of tea?

You wouldn’t— it’s too costly. And, as discussed above, you would just know the motions and the recipe by heart.

By the time technology capable of feeding modifications to your vision arrives, we should be able to augment your neocortex. This can, in turn, create real knowledge inside your head based on linked data pools. It’d be the end of visual infographics and the start of just data.

Linking data in your head, live, is cheaper, faster, more reliable, no matter how you slice it. And, until we can connect to the data inside your head, always-on Augmented Reality is too expensive—socially, technologically, economically—to become a reality.

Don’t fix problems—unwind them.

Humanity’s problem-solving process is demonstrably flawed. When we attempt to solve problems, our first instinct is correct conundrums with fixes. However, fixes tend to ignore the causes and roots of problems. In this manner we treat symptoms, not problems.

I am not talking in abstract: people’s flawed approach to problem-solving shows up everywhere in modern life, ranging from mandatory drug cocktails to automobile design.

Example 1: Drugging the Youth

In recent years there’s been a marked increase of diagnosed personality disorders among students. Students with short attention spans are labeled ADD; those with gobs more energy, ADHD. In response, we medicate—sometimes without option—to “fix” chemically-imbalanced kids.

It’s an easy solution: medicate kids and they’ll behave “normally”. As a society, this is how we addressing the ADHD “problem”. But, we’re actually just treating a symptom.

Consider why a kid might score as ADHD: their energy levels might be out of whack due to a high-sugar and largely high-fructose corn syrup diet, or their physical inactivity might be leading to heightened stress levels, leading to social anxiety and poor behavior. Or, their environment at home might not be ideal.

…or, he might just be an 8-year old who’s extremely curious about the world and explores it on a bike and by poking inquisitively at the dirt. Maybe our idea of normal “order” as it relates to Attention Deficit Disorder is flawed, or that we’re doing things to our kids that cause ADD. Instead of examining this, we seek to medicate. We treat the symptom.

Without looking at the secondary effects of medicating kids (or looking in the mirror, for that matter) let’s look at another example where we treat the symptom instead of the problem—and what probably most people don’t think about as a problem at all.

Example 2: Starting and Stopping Cars

First, let’s reframe a car trip as a set of problems:

As designed, cars transport a certain number of people from one point to another. Since you can’t just put people in a box and expect magic to take you there, there are a number of problems that need solving:

1. The car won’t move! The problem: it needs to go.
2. The car is now moving! The problem: it needs to stop.

This is how we fix these problems:

1. The car won’t move! Let’s spend energy: The engine will get it to move!
2. The car is now moving! Let’s spend energy: The brakes will get it to stop!

In both instances, we treat each part of the process as a problem needing a unique solution. But, what if we were to reframe it? Let’s think of braking a “un-going”:

1. The car won’t move! Let’s spend energy: The engine will get it to move.
2. The car’s is now moving! Let’s give the energy back to the engine and we’ll stop!

You’ve heard this called regenerative braking. While the technology is primitive, the concept is not. Ideally, the transaction cost of moving a person from point A to point B should only require energy to overcome wind resistance and friction. (And even there—futurists will tell you—there’s room for improvement.) We, however, currently spend energy at every step of the process: we treat each stage as a discrete problem to solve, not a discrete problem to undo.

In our minds, starting and stopping are both problems. But, opposed to identifying and treating them like opposing problems where one is the solution to the other, we treat them individually like symptoms. Where the cost could have been simply the cost of initiating and managing the start, stop, and travel friction, our universal solution is pricey and costly at every stage.

In terms of energy output, it seems our cars were primarily designed to expend energy, and secondarily to transport us from one place to another.* Inasmuch, when we “fix” problems by treating its symptoms, we generate new problems to fix. Over time, these problems layer over each other into never-ending abstraction.

Feedback loops

Unfortunately, we’re addicting to solving problems, not undoing them. Our economy is fundamentally based on solving problems and pain points, and common perception is that it’s cheaper to address problems by patchwork than to rethink how problems arose in the first place and undoing the situation.

And, unfortunately, it is. Treating problems symptomatically is easier. It’s easier to describe and easier to act upon and easier to think about with a divide-and-conquer methodology. There are fewer moving parts in each part of the problem, and any externalities created through “solving” the problem are somebody else’s problem, or a problem we can put off until later.

Ultimately, we create more work for ourselves—problems beget more problems, and these ill-conceived “solutions” don’t adequately address underlying problems, if at all. Antoine de Saint-Exupéry once noted, “Perfection is attained, not when there is nothing more add, but when there is nothing left to take away.”

Let’s stop fixing problems and start removing them.

–

*On that note, hasn’t anyone else found it weird that we take up half of our available roadway in New York City for the storage of empty cars?

The United States Postal Service needs to introduce electronic mail if they want to survive.

Despite private e-mail and private delivery services, The Postal Service remains top courier. However, as electronic correspondence increasingly cuts postage revenue and smarter private distribution centers enable Fedex and others to compete on cost and services, USPS needs to adopt modern messaging paradigms if it wants to protect its business viability.

I urge John E. Potter (current Postmaster General) to realize that USPS is in a unique position to do things that no one else can, and can accomplish them thusly: sell electronic post office boxes that take regular mail.

USPS e-mail: a certified, electronic, and virtual mailbox run by USPS which can get away with doing things that neither e-mail providers nor private companies can directly compete with:

1. Charge for Message Delivery. Credit card companies and healthcare companies, for example, need to notify you by mail of any changes to your service offering or plan. An official address that officially (as in governmentally-official) ties citizens to a mailbox. These companies are used to paying for this type of correspondence.
2. Charge Different Rates Depending on Message Type. Credit Card offers: $1.00 a message. Not-for-profits: $0.01 a message. Et cetera.
3. Strong-arm Government Agencies to Adopt Electronic Messaging Capabilities. Offer free message delivery for all government agencies, the cost reduction in the first year’s postage alone would likely pay for the entire implementation.
4. Automatically discard Junk Mail. I’d pay $20/year for that.
5. Automatic Package Redirection. Order something delivered to your e-USPS address and packages are automatically routed to the nearest facility for delivery, no matter where you move.
6. Official Mail Segregation. Identify and differentiate government/very important mail from everything else at delivery.
7. Certified Delivery. It can never get lost in the mail… and they can charge for delivery/opening confirmation. And you know it got to the right person.
8. Physical Address Privacy. I know where Elvis lives.

I’d also love it if they adopted a scanned-mail service offering similar to Earth Class Mail, so my experience with my mail is the same regardless of how the initial sender sent it. Backwards compatible mail.

I’m sure you can think of other things that could be accomplished with this setup.

Inasmuch, USPS’s unique features make this proposal particularly compelling. Firstly, they are one of the few agencies explicitly authorized by The US Constitution—the country’s politics (likely) won’t let it fail. Secondly, the government and agencies rely solely on USPS to correspond with its citizens—and the government’s co-dependence on it (likely) won’t let it fail, either. Therefore, it’s permanent and environmentally-friendly, to boot.

Unless USPS can’t find a way to stop borrowing from The Treasury to pay budget deficits, we’re going to have another persistent taxpayer liability on our collective hands. Ultimately, USPS needs to take a fresh look at how it can play within modern communications paradigms.

This is my suggestion.