By Michael Salu
CONCEPTION AND PERCEPTION
The Hyper-Mythology of Images Today

Image © House of Thought
The following is an excerpt of a transcript from a talk I gave in 2017. I was a contributing speaker at a conference held by The Institute for Narrative Journalism in Amsterdam, Netherlands. The four-day conference was themed "Amplify your Story", for which I took as an opportunity to discuss the invisible mechanisms determining how and what we see and who's stories will have prominence within this gamification. It has been interesting for me to look back at this research, particularly after the evidence we now have of today's society grouped in opposing factions and the perilous influence of mechanised misinformation.

I've included a selection of slides from the much broader talk at the end, to give you an idea of what was explored.

We've discussed the weaponisation of images, but we're not yet talking about the next stage of image. Or, rather, the stage that is already upon us, but invisible. The ability for images to speak to each other through their dynamic groupings and the subsequent aggregation and analysis of their information.

Like-minded images find each other through groupings that occur in how we share and interact with these images. Still, this information that these images accrue about us and our behaviour does not return to us.

This quote is from Douglas Hofstadter, a mathematician who uses language metaphors to explain complex mathematical ideas about consciousness and perception. In general, I thought his work to be interesting when thinking about how to decode this current moment. In particular, his ability to work out complex ideas with storytelling.

We're emitting vast amounts of information beyond the storytelling itself. Be that through video, image, or text. This information is perpetually analysed to understand us and to group us in social and cultural terms.

I began to wonder, if the stories we're creating are now used to understand us, then eventually and inevitably used against us–as we've seen in the explicit manipulation of public opinion and the blanket noise and disinformation–then is there a case for fiction writing to be a more powerful form to communicate clearly with each other? I started to think whether our perception has altered significantly through all this manipulation and these feedback loops of aggregation. Has our ability to perceive been altered?

Altered in the way we now look at each other and how we create stories for and about each other. Has the scope of this comprehension shifted or narrowed?
So we will begin to perceive the world as the machine does, given this feedback loop of information?
A slide looking at a possible shorteneing and manipulation of perception through the use of a dynamic digital layer.
To understand the world, we create a system of categories. These categories exist and grow within language and tropes of experience that embed themselves within images. I wondered if because we're now mostly communicating through a digital framework, whether we are communicating with each other through an increasingly simplified set of categorisations (here I am referring to the diagram I created). The front end taxonomy (User) for how we comprehend each other is shrinking in scope, whilst the back end (Data), is increasingly vast. What of us is compressed within this modern mythology? What have we actually seen before, or we think that we have seen before? Who has access to the real us?

We can now easily create video narratives of our day to day lives, in which we all use the grammar or vernacular of film. Our collective film education enables this, given we have spent a significant chunk of our lives watching films, which are very controlled narratives in their own right. So that repetition, the restructuring of what we understand as reality and that use of imagery, and the grammar of emotion and event, then filters back into how we see each other, into a tightly-controlled narrative, into an unruly unreality.

Through the middle here, you'll see the digital layer. I am not necessarily just talking about artificial intelligence, but what also occurrs in this space is an unfolding layer of dissonance between us and our comprehension of both the conscious and unconscious world. The dissonance is the tighter feedback loop, which shackles and reigns in our scope of understanding.

On the one hand, there are two main aspects, aggregation and analysis, on the other, obfuscation. This is now clearly apparent in the weaponised fake news, a layer of distraction. As a result of this shortened perception, these tighter and tighter feedback loops are geared further to optimising profit and shrinking our ability to critically perceive and tell our own stories.

As a result, do we now have a more simplified perception of each other? And are we merely telling narratives that we expect to hear of each other, narratives presented to us through this aggregative swarm?

We are guided invisibly into thinking we are thinking for ourselves.
My thoughts then on digital augmentation. I guess now we can see it moving the unconscious into the conscious reality.
I find this mechanism dangerously restrictive and prescriptive, eventually leading us to a dark place. The invisible rails now monetising our consciousness, shut us off from potential other narratives that we might need to be exposed to, shrinking empathy and criticism. And I'm concerned about that disconnect and is the space I have been trying to explore through my fiction.

We are telling more stories with video, what does this mean for language, are we even moving into a post-literate space? Or is this simply an evolution of language? The digitally native generation is exploring this intuitively.

But I'm stuck on the inability of the majority of us out there who do not code (or even those who do), to understand what this digital layer means, and we seem to rely heavily on old, outmoded tools and what we perceive to be true in an old-fashioned manner.

But that truth is difficult to engage with without actually understanding how this environment works, how artificial intelligence already works and how the mythology or the way we use images today has already taken effect.

When I say the digitally native generation, I mean the 20-year-olds born into the internet and now building companies generated primarily on the currency of social taxonomies.

So they'll build an audience very rapidly through engagement with irreverent imagery, for revenue through aggregating information. We all understand the frontispiece of this machinery with memes. But those condensed moments are quite interesting in now become the binary drivers of stories. Their front end leads us into a mythological reference that we can reframe and identify, whilst the back end essentially does similar work on us.

Traditional media needs to take more ownership of how their images are read and what these media outlets can contribute to critique this layer of obfuscation and misinformation coming from machines reading and processing images, and building, ultimately, a higher but as yet untethered layer of comprehension.
So we come to look at the conscious and unconscious. I see the unconscious being the intangible world, the world that we're trying to capture and tell stories about and the conscious being the world that we see in front of us. So this is where I'm interested in how this area be affected by deep learning systems. I don't know how much, or how many of you know about deep learning or machine learning, but with respect to images, systems will develop algorithmic processes enabling analysis of thousands of captured images and representations of our 'conscious' world and then, as a result, begin to see the world as we do, or not.

So through an amassed number of images and number of times the machine will see a picture of a dog or sees an audience sitting in the auditorium, the computer is then able to deduce what these images are and in a sense mean and how-to, and how to read, or yes, perceive them. I am interested in what this means philosophically.

If machines themselves can read and see the world as we do, are they going to eventually be in a position where they have read and seen more images than we have? And what will this mean to the way we have traditionally understood, certainly at least the 'conscious' the world?

This information will then be used to develop and even alter our perception. So we will begin to perceive the world as the machine does, given this feedback loop of information? Will we have to relearn how to perceive? If so, surely this is a situation ripe for a lot of cognitive and societal wreckage along the journey of this realtime thought experiment?

So do people who create images, media organisations, photographers, and artists need to take ownership of their work on this information level, the metadata level behind the images? Because if they don't, will we see only a small percentage of what our work is doing?
A screenshot of some of the imagery generated for my ongoing world-building research project S.O.U.L.
So onto the unconscious. Through my writing, I am beginning to explore some of the machinations behind the way images could manipulate us.

I'm using fiction as a device, in a very straightforward way I am trying to say the stories I am writing are not stories. Rather they are theoretical explorations of some of the key themes I am discussing here with you today. Where fiction plays the role of a tool, and I'm using the tropes of fiction to engage with this digital mythology layer. So the development of this project is a a world-building building exercise of mine.

S.O.U.L. which is an acronym standing for. Systemic Obfuscation of User Liberalism is a frame around how we currently live and the increasing distance we take from objective physical reality.

It is a framework to try and decode the way we move back and forth between real spaces and digital spaces and how that affects our behaviour and changes the way we see ourselves regarding identity, gender and sexuality, exploring digital isolation and isolationism.

Some of the world-building exercises are visual; mostly it is all driven by original texts, all the stories within S.O.U.L. slot into a predesigned world. I mean, you can probably explain to you better than I can. (a digital voice takes over the presentation for a moment)

"S O U L is an organisational body realised to ensure an individual has the capacity for self-generated classification to safeguard the virility of future dreams. Members of S O U L worked side-by-side with dream personnel at locations around the world to ensure seamless integration of simulated experiences with less tangible aspirations."

I didn't understand that either, but basically what's happening here is a sardonic take on digital life and of course social media infrastructure and algorithmic guidance very much at the helm of this new lived experience.

I was interested in what happens when your own dreams are actually determined for you without you knowing about it. So this slightly rogue organisation sets out to do this for you. The basic premise of this project is that you, a user, can visit this dream-based world. And to find out more, there will be some F.A.Q.'s, as per a typical website, which attempts to explain what S.O.U.L. does and how it can benefit you.

But as the user goes through these questions, they become more confused due to the increasing obfuscation. Each story in the book is related to a different question in the world of S.O.U.L.

I've listed a few of them here. Each story takes tiny moments of digital life and extrapolates.. So for example, at the bottom there, you see one titled "Mr. Interracial".

I am interested in how we can read data to tell truths about ourselves. If you look at data from the porn industry (which is becoming more and more mainstream), you can determine a lot about our behaviours, the information we're privy to and our desires.

For example, one of the most prominent search terms in the U.S. on porn websites was 'Arab', which I found interesting especially when considering the role of the media concerning what gets us off and what this data does in terms of determining what that narrative is creating the narrative of the other..

And then how does this relate to desire? Obviously, our desires are often related to what we consider to be, you know, a little bit distant, or other from ourselves.

So I used some of these porn-based data points to create characters within the story. It was an opportunity to talk about the reductive tropes built out of economic histories that have such a profound effect on us today.

And this is one example of many ways I am trying to tackle the mythology of images that I don't think we ask enough questions.

I am very interested in progress. I am very interested in where we can be with storytelling, but I don't believe that we can evolve storytelling if we don't understand the language.
This talk was originally given in spring 2017 at Pakhuis de Zwijger in Amsterdam. You can follow Michael Salu on Twitter or Instagram.

Selected talks and articles


Selected Projects

Both complete works and experiments, thinking never breaks stride.
Check back in as we add more.