A Culture Stuck In The Rear-View Mirror: The Banality Of Machine Learning Content

get ready for endlessly repurposing the past

I was writing a post a few weeks ago, and I wanted to see if there was a special WordPress tool that might facilitate whatever it was that I was thinking of doing. I think it was some kind of form submission I might have been looking for at the time, but in the rare moment of actually scrolling through the “block editor,” as it’s called, and looking at the ever-increasing plethora of tools which I normally ignore I came across something that knocked the wind out of me.

I was distressed to see that WordPress had now included machine learning writing and image generation tools [in Beta] in the dashboard of WP that I use to create this blog. Given that most of the comments that I flush from the spam filters on a daily basis here for the last eight months were touting E-Z ways I have to divorce myself from the tedium of actually writing this blog and using the latest “AI” tools that everyone and their pet ferret are ga-ga over, I find this ironic and dispiriting.


I knew what I wanted to do with my life from an early age. It was some time in 7th grade when we were asked what career goals we might have and I answered “commercial artist;” the mid-70s term for graphic designer. I joined the school paper in 8th grade to better understand the printing process. The focus has always been there for me. I have always been entranced by the blend of communication and aesthetics that it entailed. By the same token, I’ve always been a computer geek from my teenage years. The first computer I saw was a Radio Shack TRS-80 in 1979 and I was drawn to the notion of using this machine to make images. Which, back then, was down to programming them, pixel by pixel in a form of BASIC.

Meanwhile, the graphic designers who were ushered in on the Post-Punk era such as Peter Saville, Malcolm, Garrett, Neville Brody, Martyn Atkins, and Keith Breeden [to name just five] became titans to me. The work they were doing to give face to the music I was wrapped up in was extraordinary by any metrics I’d care to use. The music combined with the images achieved a rare synergy for me. The sense of movement and progression was palpable.

Due to the relatively primitive computers at my disposal, and the very real hardware cost of creating and displaying images, it wasn’t until I first encountered the Macintosh in 1984 that I felt that computers were finally getting viable for creating images. By 1985, I was using a MacPlus to completely revolutionize my work as a graphic designer as I moved my production over from the toxic stew of chemicals that was endemic to a stat camera and a phototypesetter, to the newfangled, “desktop publishing” paradigm. Within 3 weeks I became about 250% more productive. And that was with just 1-bit images to use and manipulate.


8-bit, 8KHz-24KHz frequency response. and up to one second of recording capability with 16K or RAM

As the 70s became the 80s synthesizers underwent paradigm shifts on many fronts. As monosynths became polysynths, and decreased drastically in price the march of all forms digital technology, as it had in did in graphic design, made an impression on the arts. I can recall the astonishment that accompanied the news of the Fairlight CMI, the first sampling keyboard that I had heard about. The notion of controlling the pitch and envelope of a digitally manipulated sound recording seemed like an incredibly futuristic breakthrough at the time. The first album I can remember hearing with the Fairlight in its credits still stands tall as one of the best albums of the 80s…and beyond. Peter Gabriel’s third album used the Fairlight to manipulate sounds recorded specifically for the album in order to achieve effects that could not be had any other way. I thought this was a shining new era opening up.

But I was wrong in that I didn’t extrapolate what the result of primates getting their hands on this technology would lead to. Humans, being a lazy breed, in a few years mostly used this technology not for making something new that had not been heard before, but primarily for taking the hooks of old, existing records to construct new pieces around in the hopes of having some of that old magic rub off. It must have been dispiriting for Clyde Stubblefield to hear all of those bitten samples of “Funky Drummer” that littered the music of the 80s and beyond.

As the 80s became the 90s, it became apparent to me that the vast majority of the new digital technology that was sweeping through the creative arts was generally being purposed to sample and repurpose earlier art once the limitations of memory capacity fell by the wayside. In the same way, the impetus of Modernism that had been driving music since my childhood, became dissipated. The 80s saw the rise of consciously retro style movements and I’ve sensed that Pop music has become content to rifle through the trends of the past as many artists are now content to plow-and re-plow existing stylistic furrows. It never would have been acceptable for youth in the late 70s for us to suddenly become fascinated by the most popular musician of forty years earlier; Bing Crosby. Yet in 2023, there’s no shortage of musicians climbing over one another to emulate even older music… Fleetwood Mac!


Late last year, I began hearing about CHATGPT, the front end for a machine learning system that when fed text prompts, could generate new “content” based on the parameters of the text prompt used. It was soon followed by DALL-E which creates images in the same way. And how exactly does this software work? By scraping the internet and sucking up everything that’s visible; 24-7. It’s like an unceasing robotic vacuum cleaner with a black hole for a bag and a god complex for a CPU. And based on what happened in music, I can extrapolate a road of stagnation ahead.

It was also very telling that Microsoft’s early experiments with AI twitbots quickly pivoted to racism and hate speech, because there’s there’s no shortage of examples on the web that these systems are trained on. And of course, Microsoft [and Google, and …fill-in-the-blank…] is all-in on this tech since it will provide them infinite ways to focus advertising on the topics that your personal human brain obsesses on. While that’s depressing enough, the bigger picture is that the horizons for art will shrink under the AI engines that are looking backward for raw materials. I’m looking towards a future with no humans endlessly refashioning previously extant works of art.

These tools are owned by multi-billion dollar constrtiums, answerable to no one, yet they are scraping the entire web for words and images that they do not own, and will be used to train systems designed to increase the wealth of their owners by cutting out the working class previously relied upon to create these words and images. It’s “interesting.” As little as a few years ago, the jobs of computer coder and graphic designer were seen as bulwarks against the inexorable push to replace workers with software due to the sophistication of the thought processes used to create clean code or a compelling piece of communication. Now? Viable software code can be generated at a text prompt and high resolution images in the style of fill-in-the-blank are a few words away at a prompt window. Songs can be generated. Videos synthesized from text prompts.

Will any of this be worth writing about on a blog like this one? When I can already replace my writing with chatbot output, why are we even here? I’ve barely begun to scratch the surface with this hasty post, and we’re spiraling out of the orbit of art and into the political. Because, sooner or later, everything is political. For more on that score, I can highly recommend an article written by Naomi Klein and published in The Guardian today. Great timing as I started writing this post on Friday but could not finish it. But these issues are only beginning to smolder. We should all stand back once they burst into flames.


About postpunkmonk

graphic design | software UI design | remastering vinyl • record collector • satire • non-fiction
This entry was posted in Uncategorized. Bookmark the permalink.

12 Responses to A Culture Stuck In The Rear-View Mirror: The Banality Of Machine Learning Content

  1. I read and enjoyed Kleins article in the paper today (I say in the paper, I read it on the app on my phone) and I read your post. So much of it chimes with my own concerns, and I too idolised record sleeve designers from my desk at art school while trying to grasp how I could earn a living “making pictures” as technology rose around me (far less successfully than you, I clung to the old ways for too long). The AI content lacks wit and I am yet to see it do absurdity with a human heart. The day AI can judge a dad joke from a typo that’s been kept as it sounds like a dad joke… well, I’ll be out of business as a blogger.


    • postpunkmonk says:

      steveforthedeaf – I lost the plot in writing this post, [so many rabbit holes to poke into!] but 39 years in, I’m here to tell you that computers have destroyed graphic design. By removing the highly technical processes necessary to achieve print [I could try to describe “stripping” to a lay person, but I won’t!] the barrier of entry to graphic design was dramatically lowered in ways that led to a tsunami of mediocre work. Hell, even most of my graphic design idols struggled to come to grips withe the technology I described. Most of them never got within striking distance of the peaks of their work “on the boards.” I’ve tried to stay “pure” and to eschew software processes and use a computer like the worlds cleanest stat camera. The whole retro slant of digital culture bothers me and Simon Reynolds [who wrote the pointed “Retromania” a dozen years ago]

      and that’s me… Post-Punk Monk, talking! My proviso is this: The Man will eliminate any jobs that they can with any technology, no matter how expensive [but the more expensive, the better – that provides barriers of entry against their competitors] since the goal of Capitalism is not to spread wealth but to concentrate and stop it from spreading.

      Liked by 1 person

      • I will endeavour to get a copy of that quick-smart as you are talking my language. And I was taught by Andy Vella. He despaired at my mediocrity and rightly so


        • postpunkmonk says:

          steveforthedeaf – At least I’ve heard of your instructor. I despaired at my instructor’s [naming names here: Jagdish Chavda] mediocrity who dismissed the works of Saville and Garrett as unworthy of interest. He wouldn’t even view them… they were “record cover trash.” In fact, I regret my entire college education. My university was a terrible place to learn. The only class I felt challenged and inspired by was Aesthetics. That was the only instructor I had in college where we were encouraged to think. And it was philosophy course! At least my education was cheap and I graduated without any debt back when that was still possible.

          Liked by 1 person

  2. Jon J says:

    I’ve found my experiments with ChatGPT and Dall-E anywhere from banal (mostly ChatGPT) to laugh-out-loud WTF (mostly Dall-E). Toys, but not ready for serious work. And really not wanted for serious work. I got on email from Adobe about their similar tool, still in beta, but I’m not as excited as I once was. I’m in the same business as you, reading about AI coming for graphic design jobs and I don’t see it yet, but maybe some people will be satisfied with mediocre simply for the price point.

    I’m going to put the Simon Reynolds book on the list to check out. Thank you for the recommendation.

    Have you run into Vaporwave graphics? It puzzles me to stumble across people using this style. I remember the first time seeing it, maybe ten years ago–it was art David Byrne, of all people, had done completely in PowerPoint, and it was as ghastly as you can imagine.


  3. Rupert says:

    Well said. I’ve been amazined how totally uncreative the AI stuff being puffed up is. 20 years ago AI research seemed to be to do with the ida of creating thought, or even being, now it’s reduced to simplistic scraping and mash-up. I’m a big fan of collage and appropriation but only when it is developed and reinvented by the author/artist. I don’t thnk we should worry yet: at the moment AI is used for simplistic tasks and crap art & writing. Humans remain humans, computers remain computers and only do what they are instructed to do.

    Liked by 2 people

  4. SimonH says:

    As the developments in AI will purely ‘follow the money’ there’s little hope sadly. I don’t want to go all apocalyptic of course:) I think though music has been heading that way for a while anyhow.

    A lot of people will be quite happy to have their music made that way. The rest of us will be like trad jazz fans saying how great it was when flesh and blood creatures made records…

    Liked by 1 person

  5. JT says:

    >8KHz-24KHz frequency response

    You mean 8KHz-24KHz sampling rate. Frequency response is dictated by sampling rate; cut the max sampling rate in half to get the maximum reproducible frequency (in this case 12kHz… or almost an octave lower than the nighest sounds healthy humans can hear).

    Yer proof-readin’ pal – JT


  6. It’s a wild west time for sure…and anyone who would choose to listen to “A.I. Drake” gets precisely what they so richly deserve. But I’m also heartened by the genius of composers like Taylor Brook, who uses machine-learning algorithms in software of his own design to jam with A.I. Check out his latest album here: https://taylorbrook.bandcamp.com/album/dichroma-guitar-and-a-i

    P.S. I recently moved to the Substack platform – no built in ChatGPT…yet!

    Liked by 1 person

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.