The decline of visual literacy

Hey guys, Andy here. In a week where AI has made headlines again with the release of a silent protest album, I have been thinking about how technology has changed our visual literacy and what is an acceptable standard for an image or visual.

My premise is this; I am a photographer and my craft includes the before, the during and after to produce a visual, photographic, result. This includes planning (storyboards, reference images, lighting diagrams), the shoot (ie composition, lighting, exposure, depth of field) and the post production (cropping, white balance, black or white vs colour, shadows, highlights) where I assess all the previous elements again (and again and again). There are multiple touch points of assessing the image to ensure the story I am telling is as good as I can make it. The processes involves a level of skill but what is overlooked is the care and, often, the anguish in regards to opportunity cost; do I keep this detail at the expense of that one? Do I crop X at the expense of Y? Do I convert this image to black and white and lose this glorious pop of colour? All of these elements are in my control and I give thought and consideration to each one.?

I posted this image earlier this week; it is a portrait of a shoe patina expert, PJ, sitting next to a window in a shoe shop; note the shoe lasts on the wall behind him and the shoes on the shelf to his left. I arrived at the location early, looked at various options of where I wanted PJ to be, contemplated how he would be lit and how to incorporate elements of the store to give him context and background. The final image was the fifth frame I took, using the first four fames to figure out frame five. All the elements in the frame are all deliberate and within my sphere of control.

Compare this to AI and the text prompt of “photo of man in suit sitting in chair next to a window in shoe shop” both in colour and black and white. I feel chalk and cheese do not adequately describe the differences between my portrait and this AI generated content (note the gentleman missing his left leg in the far left black and white image).

What I see as a professional is not what a layperson sees; there is a reason I am a photographer and a reason a layperson is not. Moving onto the realms of video, while I am not a videographer, there is an underlying truth of both pictures and moving pictures; both are visual means of story telling. And the quality of both is getting worse and will continue to decline as our visual literacy falls.

My rationale is that people emulate what they see; if you are what you eat, you shoot what you’ve previously seen. Camera phone ‘snaps’ and short videos have become the norm and are a new visual baseline in both what people aim to achieve and also what they’re willing to accept in regards to standards often driven by enticing cheap, or free, costs. AI content has lowered the bar even more. AI content will improve, I am sure (no more missing legs). However quality of output from AI by a layperson will, in regards to composition, lighting, narrative, depth etc… deteriorate as a layperson does not understand or have the same visual literacy a professional does and will not notice gaps in AI generated content.

It is already happening. AI slop, the term for low-quality media made using generative AI, akin to “spam” is increasingly being used to accompany social media posts. I recently found myself pointing out what I thought was obvious in regards to a concept trailer for the upcoming Frankenstein movie on Netflix; not only are the visuals and the voice over inconsistent and clunky, the video notes clearly state it is a concept trailer using AI. However it was presented to me as an official trailer and, when I pointed out it was not, the reply was, ‘Some of the footage is taken from genuine cinematic releases.’ It would appear that it is acceptable to present an AI generated trailer, using some genuine footage from other releases, to promote a different upcoming film.?

And in regards to non AI content compare this 90s Levi’s ad with an homage released this week with Beyonce. The original has a clear narrative and story, the bad guy, the good guy, the potential romantic interest, the crowd, the set up, the ending; it's all there backed by a classic track by The Clash. There is a build up of tension and anticipation and the inclusion of the boxing footage, first in the background on the TV and then featured with the knock out punch, is a lovely touch.

The modern version is a Beyonce add with Levi's product placement and is instantly forgettable; it lacks the thought or care of the original and it is over before it begins. This is by design, contemporary content is being driven by social media, ie the 16 second Beyonce Levi’s advert is a quick, attention-grabbing clip, designed to engage with brevity and rapid delivery. It can not set up an engaging narrative as there simply is no time. Beyonce, both appearing in the advert and also with her own song as the soundtrack, is doing all the heavy lifting.

Such content erodes visual literacy; if content like this is all the public is fed, this will become a new baseline of what is acceptable. Consider this, have you noticed increased video footage in a tall and thin format with the rest of screen blurred? This format is 9:16, TikTok’s favoured aspect ratio. Increasingly footage is being cropped to 9:16, primarily for TikTok, but then also being promoted on other, non TikTok, channels. The crop cuts out all content outside the middle third originally taken in a landscape format; that is to say all the detail to the left and right of the middle third is discarded. All detail, context, background or potential segues are being cut in order that the middle of the frame, the focus of the footage, can maximise the format of one app. We are sleep walking into losing sight of the bigger picture.?