Pokemon go did not come from nowhere. Itâ€™s been waiting for us.
I've been involved in the Augmented Reality community for well over a decade now. I've stalked forums and blogs, snuck into standards meetings at the w3c and discussed protocols. Iâ€™ve met a few of the key people from the many different companies helping to drive AR tech forward - leading us to a half virtual world.
Through all of that Pokemon was often mentioned. It was always a example of perfect game for the AR experience. Since long before Ninantic existed, let alone dreamed of getting the license, it was considered a potential â€œkiller appâ€ in the community. It was one of the things that represented what AR could become.
So now we actually have a world wide, Pokemon AR game how will it develop?
What exactly will AR become?
While I have no insight into anyone's specific plans, I am aware of many AR related technologies coming up, and believe I know what will be feasible, and indeed likely, to play a large role in the years ahead.
1 to 3 years -
...to the stars above.
The most obvious changes over the next few years will be purely gameplay related. Unlike Nintendo, who follows a "finish the game and release" philosophy, Niantic follows the more common mobile model of constant updates with new features.("Release early, release often")
What was once a spartan game at release is becoming more familiar, with trading,battles,and numerous other features added and refined to keep people interested.
Towards the end of the 3 years, however, things might get more exciting thanks to developments far above.
Since 2011 the European Space Agency has been launching a large number of satellites to form the "Galileo Constellation"...
This array of 30 satellites will provide a more accurate European alternative to the US GPS network.
The existing GPS public signal is accurate to 4 meters on a good day with clear reception.
By, comparison the Galileo satellite signal will be accurate to just 1 meter - even in cities and built up areas that might struggle to get good GPS reception.
It gets better though: Because the more satellites your phone can receive signals from the more accurate it can place itself, it's possible to use both gps and Galileo at the same time.
This means if you have good reception of the existing gps network and Galileo, you could even position your phone to within a few centimeter.
What does this mean for Pokemon players?
Well, it means more accurate placement of everything. Pokemon should be able to appear on the correct side of a road, house or building. Or even on top of a water fountain.
In short, things will be less "fuzzy" in terms of where they are.
It will won't yet be true "fixed to the real world" AR, but it would be a lot better, and it should happen automatically as phones start to support the new signals. Ninantic doesn't even have to work for this one.
3 to 5 years -
...teach Pokemon to understand.
Things should get really interesting over this period, as technology available *right now* should be both commonplace and reliable enough to use in-game - and the tech is pretty exciting.
When you look around your room right now, you instantly know what things are. Computer, Table, Chair, Elephant.
But computers till recently did not; they only knew a array of pixel values without any context.
However, thanks to pretty recent AI developments, computers can now understand and interpret those pixels to know what it's actually looking at.
For example, right now you can send a image to google and it will tell you whatâ€™s in it:
And their AI will work out what it contains, together with how sure it is.
While obviously not perfect, it's incredibly impressive.
Now imagine how this would tie to Pokemon. Imagine if your phoneâ€™s camera knew what it was looking at. It knew what was a building, a tree, a car, a mountain lion.
It could potentially make the Pokemon behave and react correctly around those objects, or even give a much more fine-grained "attraction" making Pokemon hang around where you expect them to.
Pokemon could even react, at least somewhat, to real life pets and wildlife making them seem more alive and part of the world around them.
The only thing stopping this tech being used now is speed and cost - processing the odd image uploaded is one thing. Processing tens of millions of Pokemon players images every few seconds? Considerably more effort.
Tango/Other 3d sensing:
At the moment Pokemon are displayed by simply pasting their pixels on top of the camera image.
Like a sticker on the screen, they can never go behind anything.
This is because phones at the moment have no sense of depth. They don't know what pixels are in front of what other pixels - so they could never place anything in between. Augmented reality at the moment consists just of pasting things over the camera feed.
Thankfully this will change, because phones are becoming equipped with 3d sensing systems.
One sensing system which seems mostly likely to catch on is Google's project Tango.
The very first â€œTangoâ€œ equipped phones are just starting to come out, so it will be awhile before they are mainstream, and awhile longer before Apple presumably has their own version.
Systems like Tango will enable Pokemon to look vastly more like they exist in the real world. They could wonder behind objects - even cast and receive shadows as real objects.
Aside from looking spiffy and increased sense of immersion, it would also allow brand new gameplay. Pokemon could actually hide themselves, requiring skill to spot. This might even lead to whole new games. Pokemon Go Snap anyone?
5-10 years -
...escaping the phones
For the moment despite the improvements above, the Pokemon inhabiting the world will be trapped behind the glass of our phones. We get a small window into another world, but no more.
This will change.
Many companies are working on head mounted displays, not to replace reality, but to add to it.
Microsoft's Hololens being arguably the furthest ahead in this field at the moment.
For $3000 you can get a complete standalone head mounted display (hmd) able to mix the real and virtual together, synced perfectly. It's still generation 1 hardware, limited field of view, and while it works outdoors the large form factor and limited battery makes it far from practical.
But itâ€™s today's tech. Developers can buy this now.
Following up behind Hololens is MagicLeap:
MagicLeap promises superior fields of view, and more natural focusing, but has yet to show of any hardware so it's likely a few years away.
There's also Meta - likely significantly cheaper then Microsoft's and offering similar features. Then there's Vuzix, ODG, Epson, Atheer - as well as almost half a dozen other companies working in this space - each one trying to be the very best.
Provided these companies can show a viable market, it's safe to say this tech will get smaller,better,cheaper and lighter. After all, when has tech not?
So in less than a decade we should have outdoor AR specs, which while not perfect, will let us live in a hybrid world of real and virtual. For gaming, entertainment, information or practical uses we will have virtual objects and information whenever we need - potentially replacing computer monitors completely.
And one of the key things leading us to this half virtual world will be Pokemon - although they won't be pocket sized monsters anymore..
11 years - Team Instinct wins Pokemon go. Game shuts down.
The world changed recently. It gained a new layer.
A layer seen only in glimpse’s via tiny screens – acting as windows into a barely-synced parallel reality.
This layer provides no useful purpose, does not expand human ability’s, and is something people have to go out of their way to experience. And yet, it represents a compelling enough of an idea for tens of millions to be experiencing it on its day of release – even in regions where it was not out.
Significant enough to give the company that licensed it its biggest share bump since 1983. And while it trailed off a bit since release, it still topped 500 million downloads 1 since launch in just a few months.
While not the first of its kind, Pokemon Go is by far the most significant Augmented Reality game ever made. It has made visible what was invisible before.
Pushing not just AR into the mainstream, but also its very nature revealing also how mainstream gaming itself is.
It does this by feeding a two addictions humans have;
- humans are infovores. We crave new data feed to us regardless of value.
- humans love clearly defined goals they can achieve. “points” or in this case “collecting stuff”
And it feeds these addictions without tying us down to one location. (Over 4.6 billion kilometers have been walked so far by pokemon players 1 ). It gives us a new experience, a new satisfying task to do, while being part of the world, not apart from it.
In short, Pokemon Go is not significant as its a masterpiece. Its not significant as its some revolution of technology. Its significant because its the start of something that is now inevitable. This is a collectively shared layer to our world that large numbers are now experiencing. For many this will be just the first of such layers.
Once that idea is in peoples heads it wont go away. We can have thousands, millions, of layers synced with the physical world, for entertainment, social, or practical uses. A half virtual world will envelope the globe, and the next generation will grow up used to the idea of physical reality being but one of many things they see.
A few no-longer pocket sized monsters are just the start.
Today I celebrate the state of AI and how fast it seems to be moving in recent years.
Did you know you can now send Google a picture, and its AI will tell you whats in it? (along with its degree of confidence)
You can test the system for free on that site, but if you want
to use it for bulk requests, they charge for the service.
Go on, try it now.
It recognized most, but not all, stuff I chucked at it – including drawings and CG artwork I made myself.
Its not perfect but it is very impressive.
Now whats particularly interesting is this service is a API. If you can afford it, you can write software that makes use of its recognition powers.
It will enable robot builders to give their creations some knowledge about whats in their environment. Not just the size or shape of a object – but what the object actually is.
AR apps would benefit similarly by this increased context. Whats being looked at? (is this a appropriate place for this type of Pokemon?)
Of course, Google Alphabet isn’t the only company in this space. Pretty much any company with any sense is working on machine learning of some sort.
To my knowledge though, this is first public API for this usage.
Additionally google have also recently used a AI to reduce their energy usage;
They are sharing the results with other tech firms, which is a pleasingly common trend. (almost as if smart companies realize that the condition of the world is more important then tiny competitive edges)
The details of the AI implementation isn’t so important here, but it shows we are starting to see a wide range of practical applications for machines that have been trained, rather then programmed.
I see a lot of what humanity invents as force multipliers.
Devices that let us have bigger results for less effort.
These AI’s I thus see as a force multiplayer for thinking – allowing correlations and deductions to be made automatically, with little or no effort from humans needed once set up.
As the use-cases spread further I see AIs being used to solve more and more of the worlds problems, the potential is vast.
And while of course there will be dangers too, they don’t come from the AIs themselves, but rather from humans training them for nefarious purposes. So we might well one day face a stock market crash due to AI activity, or a power grid shut down.
We will also, however, have deceases cured, Co2 levels reduced, and all manor discoveries and inventions we cant yet conceive of today.
Humanity has expanded its memory with writing, its sight with telescopes and microscopes, and “its voice” with telecommunications. Now, soon humanity will expand its thinking in the same way – and the effects might be just as far reaching.
Things invisible! (when in water) Porous Concrete! Sand that cant get wet!
Continuing my trend to post science/tech I find interesting, this time I thought I would post my amazement at some of the materials we have available these days.
Material technology is in many ways the building blocks of other technologies. We can’t make a skyscrapper without construction materials of a certain strength/weight ratio. We cant make a computer without semiconductors. We cant make a telescope without first inventing a glass lens.
Despite lacking a “mechanism” and appearing on the surface to be simple, many materials today are significantly clever and have huge research and effort put into making it. Sometimes they go on to be so casually used we forget that effort used to make them completely (ie, plastics).
Below I list a mix of materials available today that are interesting or appear “magic”.
Mostly I just looked for a interesting video demonstrating a material I knew existed. So some of these are from companies that sale them.
One that got fair coverage already but still cool, is Ventablack, a material that reflects almost no light – even if a lazer is pointed straight at it. This gives the impression of it being a “hole” in space, as its implausibly dark and gives no hint to its shape when looked straight on:
Hydrophobic materials are essentially ones that repel water.
Liquids will not naturally spread out on them, and rather be governed a lot more by surface tension.
As such they are really weird to look at and water on them seems to act quite unnatural. Pour hydrophobic sand into a liquid and the water forms a bubble around it rather then touching it – remove the sand and its still completely dry.
The benefits of hydrophobic materials, as seen in the second video, are essentially keeping things clean with little to no effort.
That’s it for now, stay tuned for more near future…or current….technology.
Aside from developments towards our half virtual world, humanity has made some other great leaps forward in recent months.
One of the most significant of these is in space travel.
SpaceX, Elon Musks rocket company, has recently successful safely landed its 5th rocket after delivering cargo to orbit.
Three of them landed on drone ships at sea, two on land.
Essentially, this is approaching a regular thing. Real delivery’s to the orbit without chucking most of the ship away each time.
The rocket being recovered costs $60 million to make so being able to reuse them is a huge step forward for making space infrastructure more affordable.
Making the means to get stuff into space as cheap as possible makes all future endeavors in that area easier. Its the start of a phase-change for space travel. Akin to how building a road makes traveling easier.
SpaceX, at the time of writing, hasn’t yet re-used a rocket, but it has tested them and their condition seems good. The first rocket re-use should be in the next few months.
On a semi-related note, a few months back 6 different spacecraft were docked at the ISS at the same time.
While it might not get too much coverage these days in the news, there’s actually a lot more traffic going into orbit then most would expect – and as prices fall its only going to increase.