Semi Automatic Artifact Generator

At the tail end of a post about training a network of sensory devices and learning machine programs to read sign language, I muse about the state of algorithmic art and how come I can't find any Alex Grey-trained Deep Dream programs to make meta-Grey artwork, and low and behold discover that neural style transfer has been a thing already for a year or so (although I have yet to find any Alex Grey training programs!).

People talk about technological mysticism, but this sh** gave me a temporary lobotomy when I first saw it.

It was a bit like hearing the sampler-sequencer on your phone for making electronic music (as a person with a real, tangible 707), or the robot landing on Mars. Up until a few years ago, you could  probably "put together two paintings" - a program that slices and dices a picture, not into pixelbits, but features like edges, curves, or gradients, and then splices together those features to make a hybrid picture. But you could not do it "in the style of Mondrian" or Miro, or Dali. This was a task relegated to the supreme conscious entities of this earth. No more.

And how did we get here? First you thought dreaming was a uniquely human thing, or at least a sentient creature thing (unless you're PKD), and then Deep Mind shows us that computers can in fact be set to dream just like us.* 

But that wasn't enough. The technique I'm referring to as neural style transfer is an extension of the Deep Dream project, where a deep neural net is turned on, but with nothing to do, and so it starts to get creative, and just puts things where it would like them to be, and seeing the things it knows about the world, even when they're not there.

Style transfer then takes this a step further. In a move than seems cruel as I write it, the "Dreamer" is given not a broad set of images of the world-at-large, as a contemporary, living, breathing human would see, but limits that Dreamers' world to one artist. Might be Dali, might be Mondrian (that's the cruel part).

Imagine having your entire brain erased, all but your memories of Mondrian's paintings. Being that we do not perceive objective reality apart from the interface of our previous experiences and current expectations, everything presented to us from that point on would look like a Mondrian painting. (Until we start incorporating the new non-Mondrian images we see into our corpus...). It's like that friend who is obsessed with comic books and superheroes, and everything that ever happens is a scene from Batman. You feel kind of bad, like he should go to a concert or take a walk in the park or something, don't you?

That's what we have here. We have created a thing capable of "seeing" the world, and then limited it's diet to only a bagful of paintings by a single artist.

And then we show it new things that it has never seen, and ask - "What do you see?"

Example - feed its dematerialized android eyes on a steady diet of Kandinsky and only Kandinsky. Then show it Neil Degrasse Tyson.

Neil Degrasse Tyson x Kandinsky

(content image + style image = generated image)

I found a pretty accessible description of what's going on:
Now remember- while doing style transfer, we are not training a neural network. Rather, what we're doing is?—?we start from a blank image composed of random pixel values, and we optimize a cost function by changing the pixel values of the image. In simple terms, we start with a blank canvas and a cost function. Then we iteratively modify each pixel so as to minimize our cost function. To put it in another way, while training neural networks we update our weights and biases, but in style transfer, we keep the weights and biases constant, and instead, update our image.
-Towards Data Science

Lastly, a bit about terminology. What I'm referring to here as algorithmic art is probably better termed AI art, to correctly co-locate it closer to fractal art. The distinguishing factor being the level of human agency. From a purist perspective, if you turn on an algorithmic art algorithm and leave the room, there is nothing on the computer when you return. Deep Dream is an algorithm that does the work while you wait, overnight while you sleep, while you both sleep, as it were. There's a new problem however in neural style transfer work - although much of what you're looking at is fully automated, the small role of the artist is disproportionately relevant to the overall impact of the image. I mean come on, a computer isn't meta-Escherizing itself! There's an artist at work there.

*Sure it is a prodigious act of artistic license to say that Deep Dreamers are dreaming "just like us," but can we not say that our dreams are a default, empty state filled with our life's worth of training data to run amok?


Notes:
Here's a place that will do it for you, but they have their own pre-designed style templates (called pastiches) - Deep Art.

There's also a few apps that do it, Prisma is one.

And here is a great survey of all the slight variations that you can get, like multiple transfer, masked transfer, color interpolation, and probably my favorite, texture transfer.

Vice article on Algorithmic Artist Dextro, 2014

Where's My Thought Translator
Network Address, 2018
Here's a few (what I imagine will become) classic style transfer artifacts, as well as a couple of the (already) style transfer classics, albeit from early 1900's abstract artists.

The Macro Image Series and the Dematerialization of Artifact
Network Address, 2013
Here's a presentation about the propagation of cultural artifacts in a digital medium.

Post Script:
I like to give Piet Mondrian credit for his prescience in foretelling bitmaps and pixelation, despite the fact that in his day, the primaries were still red-yellow-blue instead of our current day cyan-magenta-yellow.

Here, a painting of Napoleon, trained on Raphael's School of Athens:

Meta Napoleon Bonapart, unknown origin

Raphael's School of Athens

Jacques-Louis David's Napoleon Crossing the Alps

Post Post Script:

Same thing here but for music; I'm still waiting for the meta cover band that covers other bands as if they were already covering another band, for example, a cover band that does Radiohead doing the Doors, or if the Squirrel Nut Zippers covered Metallica, what would that sound like? Any tips would be much appreciated.

Ergobots

The question is no longer "Will robots take over the world;" it's more like "what will it mean to be human?"

Mobile handsets are the exocortex. Social media platforms are the virtual acropolis, populated with quasi artificial intelligentities. When you try to find a date online, you're fielding messages from semi-automated algorithms. When you call your network provider, you're talking to a fully-automated algorithm.

When you're heart gets too old to work by itself, you get a pacemaker. You're part robot. When you lose your eyeballs, for whatever reason, you get a headset that sees for you and translates visual stimuli into auditory signals that you can now perceive. You're an eyeborg.

So it's no surprise that your measly meatbody now gets support from an exoskeleton. These exosuits have been on the scene for a minute, but geared towards folks who can't move in the first place, due to spinal injuries, for example. Now they're in strenuous work environments. Tomorrow your mom will be telling you to put on your exosuit before you go out to shovel snow.

Notes:
Ford brings 'exosuits' to workers in 15 factories
Aug 2018, BBC

Livability

Downtown Cranford, New Jersey, circa 2018

A recent article in New Jersey's local news drew my attention to the subject of charming towns. In the town-planning world, the criteria by which we measure "charming" is called livability, and there's been plenty of towns seeking this status as of late.

Livability is related to walkability, like when you can get all your errands done with only parking your car once. And when you can walk from one store to another without crossing an 8-lane superhighway pedestrian deathtrap. It's also one of the reasons that lots of younger people are preferring to live in the density of downtown instead of on its spacious suburban fringes (which may be related to the reason for towns paying more attention to this stuff).

But livability goes further, as it is even more qualitative. It's not only "can I easily walk around town," but "is it a nice walk?" Livability measures how nice the greenery is, what the awnings look like on storefronts, and the feeling of open-ness vs closed-ness that one feels while walking around town or sipping coffee on a sidewalk bench after lunch.

There's ways to approach such qualitative measures quantitatively. For example, we can simply measure the amount of greenspace in square feet, or we can measure the ratio of the width of the streets to the height of the buildings (which is a good determinant of close-ness/open-ness). Or you can do what Haddonfield did, because they got notoriety.

Haddonfield adopted a business improvement district in 2004. This is a special distinction that allows certain properties in certain parts of town to get tax breaks, with the hope that it will "promote the economic and general welfare of the District and Borough." And words like 'general welfare' mean quality-of-life.

Forming these districts happens by way of a public/private partnership between property owners and a municipality. They are run via a board composed of professionals working in the town, retail business owners, property owners, and residents who are none of the above, that way their interests are diverse and vested.

The charge of this new entity is vast, but critical to a flourishing hyperlocal economy in today's world. They encourage self-help and self-financing, provide leadership, offer grant programs to attract new business, enhance the look of the downtown, sponsor networking events, organize special events, and market the downtown, all of which can help to raise the quality of living in these towns.

Improvements don't happen overnight. It took a good 15 years for Haddonfield to get this credit. Then again, what do I know, maybe it started being awesome a long time ago.

There's another town that established its own improvement district, and they've also been doing really well in the Livability Department. Cranford started an improvement district in 1992, and were the first to do so in New Jersey. If you drive through their downtown, right past the train station, you'll notice, or not notice, any Big Dog businesses. No fast food giants, no consumer goods warehouses, just mom-and-pop businesses.

This helps because the business owners are way more invested in what goes on locally than would be a multinational corporation. They care whether their sidewalks look nice. They care about quality of life issues.

Their District was created, among other things, to preserve historic aesthetics, establish a design criteria,  and ultimately to deal with funding and decision-making for the district so they could compete with the flow of commerce that went rushing through the recently-opened doors of the nearby Short Hills Mall.

Cranford recently took things a step further by hiring a town planner, because paying someone to think about all these little things makes a difference. There's a piano sitting in the middle of their downtown, all day all night. Sometimes they plan for an event there and invite a musician or a school music group, and sometimes, like on Friday nights, this random guy just shows up and plays old school Nintendo songs all night. They put up barricades and do yoga in the street. They get residents to loan their porch to local musicians who then travel around to all the different volunteer porches in town, followed by a crowd of drinking town residents (a crowd of residents, mind you, that is ok with paying somewhat higher property taxes to live somewhere that makes being an adult a little less soul-sucking).

Anyway, you can't just string up some Edison bulbs and call it a day. Livability requires planning. And it requires patience. And if you want your town to make it into the local news, or the national news, or you just want it to be a nice place to walk around, then you may want to take a stroll through these neighborhoods, take notes, and take action.

Post Script:
I neglect to mention that it doesn't hurt if you have a train station nearby.

Notes:
This town was named the 'most charming' one in all of N.J. We offer these pics as proof
Aug 2018, NJ.com

Crowdsourcing Reform: Community Participation and Land Use Regulation in New Jersey
Research paper for studies in urban planning, 2010

Haddonfield Business Improvement District

Partnership for Haddonfield 

Cranford Business Improvement District
adopted 1992

Cranford was recently named by The New Jersey Chapter of the American Planning Association as one of the 2018 Great Places in New Jersey.
Oct 2018, nj.com

Streets for People: A Primer for Americans
Bernard Rudofsky, 1969
(Read this book to learn the hows and whys of the enjoyable dowtown experience from renowned thinker, designer, and creator of the Bernardo sandal, yes it's true.)

The Death and Life of Great American Cities
Jane Jacobs, 1961
(The number one book about what makes a nice place to live by the number one person to write about such things.)


Post Script:
Here's a peek through the looking glass. An article just come from the BBC, about poor town planning. It describes the lives of many young adults stuck in this new autocentric anticulture nightmare. In America it's called "suburbia" and it's been a problem for decades.
[poor town planning]

Apparently developers did not consult planners when they plopped hundreds of residential units so far out of the city that you can't get to the market - or anywhere else - without a car.

Some developers buck this trend, and their successful recipe is detailed in the article as such:

“The secret is the layout of connected streets with interesting squares and courtyards, coupled with the way that offices, small shops, cafés, pubs and even a garden centre were integrated with the homes as in an authentic small town.”

Young couples 'trapped in car dependency'
Oct 2018, BBC

UrSci

From other flavor networks like this we can infer that Asian cuisine is vegetable-based and diverse, whereas North American cuisine is animal-based and limited to basically cow meat and butter. As for this flavor network, it shows us that cashew nuts are more like cheese than nuts.

Ur was an ancient city in Mesopotamia and the birthplace of Abraham, who is the birthperson of the Abrahamic religions (Jewish, Christian, Muslim). Because of its potent origin-inducing prowess, it has become a word that means 'the beginning of,' for example, an urtext is the original version of a text.

Every time I hear about network science, I think of the word ur-science, because it sure seems like it's underneath all the rest. If physics is the underlying science from which all else follows, then network science must go beyond even that. It's like the link between math and science, that's another way to look at it. Math is all meta, totally disembodied, no physicality involved. And it's more like a language than a science, a perfect language, God's language if you will. Network science, however, seems to reach in-between this space, in-between the body and the void.

As anyone even slightly interested in science would already know, there's some problems with physics, like finding a universal theory that explains both gravity and quantum mechanics. Although the theories used in both these phenomena are mature and useful, they don't overlap enough to make one single theory.

Then there's other problems, particularly with quantum phenomena, like 'spooky action at a distance,' also called non-locality, and also called teleportation, where a change to one particle that had been entangled with another but then separated by space and yet still entangled, will immediately change the other entangled particle, no matter how far away they may be, and instantaneously, he says again for effect to remind you, the reader, that Einstein's Relativity says that nothing happens "instantaneously" but as a result of its relationship within the greater context of spacetime. (This goes against all intuitive thought in the classical physics mindset, and is the reason Einstein called it "spooky.")

So we know that there's some work that needs to be done in terms of making Science into one thing, and to explain some of the things that we know to be true, yet we still cannot believe, albeit because of our feeble limbic-driven wetware.

***
This brings us to Network Science, this thing that became real popular with the advent of the Internet, although it's been around for quite awhile. My own introduction to this came from the book Linked by Albert-László Barabási which came out just before The Social Network (the thing not the movie) in 2002*.

In it, Barabási gives a rudimentary exploration that even a non-scientist can understand. There's hubs and there's spokes. And although The Internet looks like a mess, there is some order to it, and there are some rules to how networks form, how long they last, what kind of information gets passed through it and between who.

Plenty of work has been done on the subject of Network Science, both since Barabási 's 2002 book and long before it. Granted there is always a draw to something new. But there are also these moments that happen when a phenomena has been sitting around doing its thing for ages, and then all of the sudden an example comes to life so that everyone can see it and use it and think about it (and even write books on it). And that would be the dynamics of the social network that we now use to explain and understand everything from fundraising to depression.

When a thing becomes this pervasive in society, we can talk about it and wrestle with it in common language. And we can see anew a thing that once before we never even noticed. I guess this is similar to how we used to think the nervous system was like a bunch of pipes filled with a fluid that behaved according to laws of pressures etc., and then electricity comes along, and all of the sudden there's a new way of looking at the nervous system.

So perhaps it's possible that Network Science is a new analogy for things we have yet to understand. And then again, maybe it's just another one of those - when you're holding a hammer, every problem looks like a nail - and the Network is just the hammer of our day. 

*Physicist Mark Buchanan also wrote about networks in his 2002 book Nexus


Notes:
The dimension of a space can be inferred from the abstract network structure
Sep 2018, phys.org

Post Script:
Network Science is the Ur-Science
Network Address, 2015

URSCI - Undergraduate Research, Scholarship and Creative Investigations at Dominican University in Illinois

Flavor network and the principles of food pairing
Yong-Yeol Ahn, Sebastian E. Ahnert, James P. Bagrow & Albert-László Barabási
Scientific Reports volume 1, Article number: 196 (2011)

Smoke and Mirrors

Diamonds are forever.

If you don't know what a deep learning neural network is by now, I can no longer tell you, because there's enough buzz out there already.

If you like the idea of neural nets, quantum mechanics, photons, and metamolecular diamonds are going to take over our world and our bodies in the next 20 years, then you're in the right place, at the right time.

Because today makes an appearance of a neural net that uses light instead of electricity, and they even have a new name for their proof-of-concept device - a diffractive neural net. So get with it.

Image Source: The Crystal Dome at Swarovski in Kristallwelten, Austria


Notes:
Xing Lin et al. All-optical machine learning using diffractive deep neural networks, Science (2018). DOI: 10.1126/science.aat8084 

A neural network that operates at the speed of light
July 2018, Techxplore.com

On Statutes and Liberties

Original Statue of Liberty vs Robert Davidson's Las Vegas replica. Which one is more feminine?

The United States Postal Service made some stamps featuring the Statue of Liberty, but used the "wrong" reference, and now owes an un-attributed artist $3.5 million. I think I'm already using the wrong language here.

The artist, Robert Davidson, made a sculpture of the Statue of Liberty for a New York-themed casino in Las Vegas (can we all just read Baudrillard's Simulacra and Simulation right now?), and it was this statue, not the original, that the Post Office used to make its stamps.

There's a difference, as you can see above, where one is more feminized. Davidson did that on purpose, and that difference makes it original enough to be protected under fair use law.

Interesting situation here, in that a copy of a copy got in trouble for copying the copy instead of the original, and now owes the copy-creator. But the details get even better, because the USPS licensed the photograph of the Statue from Getty (well I guess they can't just take their own photograph, right?), but they didn't realize that the photo they licensed was of the wrong Statue. Hence Davidson, sculptor of the statue in the photo that was licensed, sues for unauthorized use.

And because the Postal Service sells lots of stamps, there's lots of money involved. There was a good case similar to this involving Artist Shepard Fairey's HOPE Poster, President Barack Obama, Photographer Mannie Garcia, and the Associate Press.


Notes:
Information on Fair Use, Copyright, Intellectual Property, etc.
Electronic Frontier Foundation

Post Office owes $3.5M for using wrong Statue of Liberty 
Jul 2018, Ars Technica

On Property Rights
Feb 2018, Network Address

Pedestrian Utopia

Anamorphic Crosswalks are the new thing, maybe.

Brains are expensive. It takes a lot for us to pay attention to stuff, and so, most of the time, even though we think we're paying attention to things, we're not.

Funny, but part of our mental resources are spent convincing us that we are paying attention to everything, so that we don't misallocate our attention-resources to things we don't need to be paying attention to. In other words, we don't want to spend time noticing the things that we're not noticing. Yes?

The literal blind spot in our field of vision is a perfect example. You probably don't even know you have one, but there's a sliver of your peripheral vision field that's invisible due to some wiring issues in our heads. But part of our brain makes up for that missing piece, filling it in, extrapolating from existing data, kind of like how photoshop can erase you from a picture, leaving only the background you're standing in front of.

Anyway, whether we are willing to believe it or not, there's a lot we don't notice. This is a problem for drivers, for sure, but it's much more of a problem for pedestrians.

If you, driver, pass the same intersection every day, and one out of every 1,000 times there's a person standing at the corner waiting to cross your lane, chances are you will not see that person. They are never, ever there, so why would you see them.

This is how the brain works. It predicts; it expects. It sees what it thinks should be there based on previous experience. We can't treat every situation as brand new, because it takes too much of bandwidth. And we will not look for a pedestrian that might be there in a 1-in-1,000 chance.

Fortunately, there's some really smart people thinking about this problem. University of Toronto engineering researchers working in Human Factors and Transportation say it's a consistency and predictability issue, and I am so happy to see someone with a bit of authority say this.

They say the problem is not that it's not the driver's behavior that needs to change, it's the infrastructure itself. And it's because of this energy-saving feature of the brain.

You can ask drivers to 'look both ways' all you want. They have way too much to worry about, and if pedestrians do not make up enough of their sensory environment, they will not waste precious attention on them.

I'm not an urban planner, but there's one thing I'm really adamant about it when I comes to traffic design. In the US, crosswalks are at the corner. In Europe, however, they are about 10 or 20 feet away from the corner. Maybe not everywhere, but in enough places that I certainly took notice of this change.

This is one of those things that makes travel and exposure to other cultures so valuable. The moment I saw this difference, I couldn't believe why we were doing it the wrong way, and that nobody ever says anything about it. Studies like this support the European design; there is absolutely no way that a driver is paying attention as they turn.

And until then, another simple solution would be to just get more pedestrians! If every intersection had a wall of people waiting to cross every time you pass it, you will certainly pay attention to them.


Notes:
More than half of drivers don't look for cyclists and pedestrians before turning right, study finds
Aug 2018, phys.org