Tuesday, July 19, 2011

Photoshopping the Filter Bubble

Eli Pariser first noticed the phenomenon that he calls the “Filter Bubble” when his conservative “friends” began to disappear from his Facebook page.  Now you might think that for Pariser, co-founder of the unabashedly liberal website MoveOn.org, this was good news.  Not so, he asserts in his book titled, not surprisingly, The Filter Bubble. Pariser is apparently a throwback to the days of the founding fathers when savvy politicians believed in keeping their friends close and their enemies closer; when you learned by studying the perspective of your adversary. So, he wanted to find out why the Tea Party had left his Facebook party.  Not their choice, Pariser came to discover – they had been filter bubbled out.

What Pariser learned in the course of researching his book was that Facebook had shown his conservative friends the door because Facebook, or rather Facebook’s filtering algorithm had decided, based on the fact that Pariser did click on his Tea Party buddies less often than his MoveOn cronies, that he really wasn’t all that interested in them.  So why clutter his page with them?  “Off with their heads!”

I advise you to read the book.  It is interesting, and more than a little chilling.  Basically, here’s the Cliff’s Notes version of what is going on:  Most of THE INTERNET makes its money from advertising.  The ubiquitous little ads that pop-up on the pages you go to while online.  The hosting page – be it Google or Lands End, the Gap, Spotify, whoever – gets a bigger piece of the ad revenue if you actually click on an ad.  Hence the more the page "knows" about you the more it can push – in split seconds – ads onto the page that are tailored for your very personal profile.  Try this – do a Google search for Labrador retrievers, play around on dog pages for a while.  Now go to some other site – like Amazon or Yahoo news.  Look at the ads.  Seem a little more “doggy” than usual?  I told you it was a touch creepy.  There are very large, very wealthy companies that do nothing but gather our “click streams” and sell them to the algorithm-makers.

You can actually understand it from a business perspective.  Advertisers are simply trying to place their products in front of people whose own Internet behavior indicates that they are interested in the product.  Seems harmless until we remember the case of the vanishing conservatives.  THE INTERNET isn’t simply tracking and constructing filters based on the products we like, it is also building filters that keep out the ideas that we don’t like, while foregrounding our proclivities.  Internet algorithms try to construct, and lead us to, our vision of “a perfect world.”  You know the saying – someone asks you a question and you respond, “Well, in a perfect world .  .  .”  What we mean is in our perfect world,” the world as we would like it to be. 

In the movie Heaven Can Wait – the 1978 Warren Beatty version – the welcoming angel tells Beatty’s character that heaven is “a product of your image and that of those who share your image,” a perfect world, defined by what we, and our “friends” believe a perfect world should be.  That is a very prescient “internet-algorithm-esque” concept for a 1978 chick-flick!

There are, however, problems inherent in letting Internet algorithms define a perfect world for us, based upon their perception of our behavior.  I am reminded of the elementary schoolyard where I played as a child.  It was, by contemporary standards, a death trap.  Asphalt paving everywhere except on the fields where we played baseball and football.  Those were dirt, not grass, dirt.  The slides were really tall – you sort of had to lean back to see the top.  They were shiny steel with four-inch sides.  Sliding down on summer days was a delicate balance.  The heat seemed to increase your speed, but if you were too light to get all the way off the end, you stuck.  First degree burns on your butt.  So you leapt off to the easier embrace of the landing area, which was, remember, asphalt.  Similarly, sliding in a baseball game was a decision to which one did not come lightly.  You measured the transient heroism of victory against the possibility of major abrasions and a tetanus-shot trip to the nurse’s office.  All in all we had a good time.

My daughters grew up as playgrounds were transitioning into “a perfect world.”  Everything is now low and slow, plastic and padded.  No doubt injuries still occur, though it seems you would probably have to put some planning and effort into it.  According to the TV ads, successful injuries are dealt with by a phalanx of perfect moms welding spray-on antiseptic and instant bandages.

The point is this – sometimes the world that is constructed by others for our “own good” damages the depth of our experience and compromises the legitimacy of our conclusions.  Internet filters that show us only that which we already believe and desire, destroy the opportunity for the serendipitous discovery that comes from going somewhere we have never been before.  They deny us the opportunity to learn from those who think differently than we.  They create a perfect world in which everything seems low and slow, plastic and padded.

But wait! There is a software fix for this world in a bubble that might even increase Internet profits.  Listen up, moguls.  Photoshop has a feature that lets you select parts of an image; either parts that you click on, or parts that share a color.  Point is that it lets you select part of an image based on certain criteria.  Once you have selected those parts of the image you can go to the “Selection” menu, where among the options is: Select Inverse - which means "select all those elements that I have not chosen."

You see where I am going here?  If the algorithm can decide what it thinks I want, can’t it also decide what I don’t want?  Wouldn't it be cool if I could tell Google to “Select Inverse?”  Create a search based on the notion that what I haven’t experienced might be more intriguing than what I have already done?  Think about it as a clock face.  You are standing in the middle and facing 12 o’clock.  Noon is “a perfect world.”  Midnight is what the algorithm predicts you want it to find.  Six o’clock is “Select Inverse.”  Why can’t I ask for that “six o’clock search” instead?

And let’s not forget the numbers in between.  Photoshop also has a slider attached to many of its functions called opacity or intensity.  Essentially, it a function rheostat.   You move from, say, 100% opacity, where you cannot see through an image at all, to 0% opacity where the image disappears and only the background is visible.  Why not a “Search Slider” that lets you move the algorithm around.  Say 1:00 o’clock equals a search with your characteristics intensified 30%, 2:00 o’clock is “you” intensified 60%.  And 11:00 o’clock reflects a search with your characteristics deflected by 30%, 10:00 o’clock and “you” are deflected 60%.  Fader Bar Search. Why not?

There are obvious and intriguing existential implications in the fact that moving in both a “positive” and a “negative” direction will eventually bring us to the same 6:00 o’clock “Select Inverse” world that stands in algorithmic opposition to the perfect world that the Filter Bubble seeks to create for us.  But, in the final analysis, shouldn’t we be allowed to choose the direction and intensity of the journey?  Isn’t that really what “searching” means?

Saturday, July 9, 2011

Making Light of Art or Art with Lytro?

If you were an “image maker” back before 1800, odds are you worked in paint, ink, clay or stone, and your objective was representation.  Etch the critter you want to kill on the shaft of a spear or the wall of a cave and you increased your chances for a successful hunt.  If you were both a good hunter and a talented artist, maybe Og would slip you an extra hunk of Mastodon haunch to do some painting on his wall or weapon.  Fast forward 20,000 years, and the ability create a particularly godly rendering of the monarch or prelate could get you a pretty cushy stay in Athens, Alexandria, or Rome. 

Come the 14 and 15 hundreds, and the dominant artists of the Renaissance were employing canvas, oil paints, and occasionally ground glass from the thriving foundries out on Murano to craft portraits and landscapes that seemed real enough to stride off the wall and stroll around the piazza.  Brush and palette-type image-makers continued to thrive well into the 1800s.  Anybody who was anybody had a cluster of oils adorning the walls – like some eerie foreshadowing of today’s refrigerator doors or cubicle walls.  Consider James Whistler’s homage to his mother; still mainstream when painted in 1871.  But by then the game was definitely afoot.

As far back as 1826 Frenchman Joseph Nicephore had used a camera obscura to burn a permanent image of the view from his garden onto a doped pewter plate.  He called it heliography, or sun-drawing.  It was, in reality, the beginning of photography.  By mid-century image-makers were toting boxes of various shapes and sizes in to the halls of congress, the homes of the high and mighty, the backwaters of jungles, and onto any battlefield to which they could finagle access.  Photography arrived and changed image making for forever.

Photography was a double-edged sword.  It certainly loped the legs out from under the portrait painter.  Who needed a “close to real” painting when you could have something else that was “picture perfect”?  On the other hand, from this newly mandated supine perspective, artists began to look beyond the blinders of representationalism.  It is not by chance that Impressionism sprang onto canvases in the late 1800s, just as photography was claiming control of portraiture. 

Almost without exception, significant innovations in our ability to create images change the entire spectrum of human expression.  Consider these three connected innovations:
Point and shoot digital cameras, their inclusion into cellphones, and, the Internet. The result? At the end of May of this year, Facebook announced that its users photo archives had exceeded 100 billion images.  That is a lot of refrigerator doors, and an incredible number of dumb cat pictures.  But it is also a phenomenon that we should not ignore.  The iconic photographer, Ansel Adams, used to distinguish between “scenic beauty” and artistic beauty.  Scenic beauty is what you see on postcards.  Artistic beauty is what you see in images created by Adams, and Margaret Bourke-White, and Yousuf Karsh.  I would hazard to guess that the majority of Facebook photos contain neither scenic nor artistic beauty.  They are images of convenience posted as a shorthand peek into our experiences.  Neither the intent nor the impact is aesthetic.  They are reportorial – because the capabilities of the technology incline us to that usage:  Point. Click. Dog. Child. Vacation. Flat tire.  Photography trivialized.

But there is a new game in town.  Its name is Lytro.

The short explanation is that Lytro is a company founded by a freshly-minted Stanford Ph.D. by the name of Ren Ng that is going to put the light field camera in to our hands.  Whether it will be in our phones or a stand-alone unit is still up for grabs, but we will have it.  Dr. Ng intends to see to that.

“Cool.” I thought.  .  .  .  .  “What’s a light field camera?” [And, I’ve taught photography in a very respectable university!] 

I now know that a light field camera is one that grabs and stores all the light being reflected by whatever is in front of the camera.  That, too, seems something of a yawner until you think about what it means to have ALL the light that is reflected by whatever is in front of a camera.

Remember Joseph Nicephore and his heliography? Sun drawing?  And remember what it became? Photography?  That’s right – photo = light.  Photography is “light drawing.”  Photography has, until Lytro, grabbed SOME of the light being reflected by whatever is in front of the lens and makes a drawing with that little bit of light.  Without getting all warm and geeky about this, being able to grab ALL the light, store it, and play with it all you want AFTER clicking the shutter is, well, incredibly warm and geeky – and artistically awesome.  I understand that it used to take scads of cameras and computers to generate those “ALL the light” images.  Lytro does it with cellphone size technology.  Truly wonderful.  Why?

Just three examples:

Depth of field.  The depth of field in any photograph is the range of the image that is in focus.  Think of a photo as a loaf of sliced bread.  You are looking in one end of the loaf and the picture ends at the other end of the loaf.  The depth of field is the number of slices that are in focus.  Depth of field is one of the things that photographers who are concerned with artistic beauty agonize over.  Focus draws the eye, de-focus feathers attention.  Deciding which slice or slices of the loaf you want to be in focus is an artistic decision.  BL [Before Lytro] you had to decide on the desired slices beforehand and adjust light sources and the opening of the lens and the shutter speed that would – depending on the light sensitivity of your film – allow just the right amount of light into the camera so that it would draw the loaf of bread so that only the slices you wanted to be in focus, would be in focus.  Clear?  Of course not, and it no longer matters.  What is important is that an “ALL the light” image let's you determine depth of field AFTER you take the picture – and you can keep changing your mind and/or create multiple depths of field.  Make slice one clear, fuzz out 2 through 6, 7 and 8 clear, and so on.  Old images quickly become, well, toast.

Example two: Shutter delay.  Gone.  You had to push the shutter halfway so that, WAIT, the autofocus could decided which slices of the loaf you wanted in focus, adjust the light and shutter speed and NOW you can take the picture.  Oops.  Baby smile gone, speeding athlete out of frame, whatever.  Not with Lytro.  One click of the shutter gathers ALL the light – gobbles the whole loaf.

Example three: Flash.  Surely you jest.  Lytro grabs ALL the light, and rarely needs more.

I better stop now.  I’m feeling a little lightheaded.  Oh, I know, people will do silly things with it.  We’ll see images with more focus highlights than anyone could absorb.  Those will be horrible.  But the others?  Oh, the possibilities .  .  .  .

All right, way down here is the link to the Lytro picture gallery.  I knew if I put it up higher on the page, you’d never get here.  Enjoy.