On A.I and the rise of Automation

A few weeks ago, I was contracted to create an album cover. It was a simple enough gig. The client sent me some photos and asked, that I make alterations to them. In graphic design terms I had to do some ‘compositing’: Adding elements from other photos, blending them together, overlaying effects that masked the rougher edges and adding some grain for that super cool, grungy, urban aesthetic. All in all, the work took around a day plus the time needed for revisions. The client was happy. And once I got paid, so was I.

While I did not take any photos myself, I still showed the finished product to some friends and family, when they asked “What are you up to these days?” and presented it as my own artwork. Rightfully so, most would say. But I didn’t do it alone.

During this process I used multiple programs aimed at photo-manipulation, most recognizable of which was probably Adobe Photoshop. This incredible software comes loaded with a plethora of tools, just waiting to be awkwardly utilized by an autodidact, who has seen too many YouTube tutorials. With ‘Magic Wand’ and ‘Quick Selection Tool’ I can isolate an object from the background in seconds – A task which might have taken hours to carefully execute in the good old days of analogue media. With a few simple filters and automated actions, I can turn a blue daylit scene into a moody, warm sunset. And using Photoshop’s new ‘Neural Filters’ I can add anything from accurate depth-of-field blur to literally changing the entire scene to be covered in snow (if I wanted to). It’s amazing to someone, who – as a kid – used to sit and draw ugly smiley faces in Windows 98’s Paint. And to someone who did physical image manipulation 60 years ago, I can only imagine it would be mind-blowing. Shrewd readers will already see, where I’m going with this.

Today, one of the biggest topics around the world is AI-generated art. The human-like writing of ChatGPT bordering on uncanny-valley. The mind-blowing artworks that come out of Midjourney and DALL-E. And, most unsettlingly to me, the fake voices of ElevenLabs and the like. With a few simple prompts anyone can churn out a product in minutes, that – to most but the very scrutinizing – resembles a human work of multiple days, months or even years. Just a year ago, it would seem overly optimistic to think, we could already be so far along. Everyday these AI models get better and more precise, and everyday more and more artists, creatives and writers worry how this will affect their business.

Already, I’ve seen co-workers and friends in the media-industry adopt these tools. A friend of mine, who works in strategic communications, shared how he had used ChatGPT to alleviate the tedious task of writing the company newsletters, and simply added final touches and corrections to them.  A graphics artist shared how he used Midjourney to create mock-ups for client specifications, and then created a final piece based on the one they liked the most. While these are just anecdotal stories, the internet journals at large are also becoming filled with stories of students using ChatGPT to write their essays, and tutorials on how to get it to write your company communiqués. And so, the worlds of academia, media and corporate communications are all struggling to figure out, how much they should adapt to this new technology and where to draw the line on ‘original’ work.

But they are not alone. The programmers, statisticians are accountants are also in the danger zone, as AI models evolve to write code, analyse large portions of data, and handle everything from receipts to payroll-sheets. It is starting to seem as if anything that is mostly done in front a computer, will be done by a computer within the next decade. And while surgeons, farmers and chauffeurs are eyeing the robots in their own fields, maybe the physical world will be the last to be conquered by automation. And so, the response these innovations seem to fall mostly into two camps: The AIs are coming to take our jobs and leave us penniless, or these new tools will be a great aide that will helps us work better, even faster and finally push those stock-prices to unimagined heights.

Personally, I am probably partial to the former of the two. I think it stands to reason that, as the hours needed to create certain products are reduced, so too will be the number of people needed to work in the industry. And, the way our current global economy works, it seems unlikely that this increased production will equal increased means for the people who used to produce. Much more believable is it, that the upper management of any corporation, will see a way to get the same output as last year, but with a fifth of the employees - hey, someone will still have to input the prompts, and it sure isn’t going to be the boss, who’s got more important things to do.

This might seem a little reductionist and definitely anti-capitalist, but at the same time I don’t believe there is any use in trying to fight these tools. All the effort in the world cannot shut Pandora’s Box after it has been opened, and while the AI models may seem like little black boxes of incomprehensible magic, they are still just ideas. Ideas that human beings had, tried and put out into the world. And much like the steam train, the phone and Photoshop, you can’t un-think an idea. A graphic designer in the early 90’s might have railed against the rise of digital manipulation software as much as they liked, but it wouldn’t change a single thing. Progress’ gonna progress, and you can’t stop it.

So where does that leave us? Will AI render all our jobs obsolete, offering us no alternative way to survive? Are we doomed to watch, as the global elite gathers up the remaining resources and leave the rest of the population out in the cold? Will Elysium finally become a good film, simply on the merit that it turned out prophetic? Maybe. Maybe not.

There is another way. Because it is also true, that these tools can really be a great aide in our work. Much like I could create in a day, something that might have taken a week in older times, maybe in the future, I could create it in a couple of minutes. And maybe that’s not a bad thing. And maybe the really scary realisation isn’t that our technology needs to stop evolving, but that our work-life needs to start.

If I’m feeling particularly optimistic – Or particularly inebriated – I can imagine a world, where we have reduced the working hours needed to make the things we want, without feeling the need to conjure up some other busywork to make up the rest of the traditional 8-hour work day. We could simply recognize, that we have freed up that much more of our lives to pursue other interest, or perhaps those same interests but on personal projects. Bear with me because it is about to get utopian, but if these AI models, these tools sprung from ideas, can vastly increase our work-output, could we not just reduce our working hours? Is there any reason, the only response to the assembly line needing fewer hands is to reduce the amount of them, instead of just reducing how many hours of the day it runs?

In these times, people talk of lofty economic goals like Universal Basic Income (UBI), where everyone is paid a set amount, simply for being part of society, and any work done adds additional funds on top of that. Often the response to such ideas is, that it is unjust for the essential workers of society to let others coast on the back of their work. And if everyone is getting paid without working, where is the government – being funded by the taxes of workers – getting the money to pay them?

But in a future where every job can be automated, it seems a lot more feasible, that the robots and the AI models could be the foundation on which we build a society that requires a lot less but gives a lot more to its citizens. Besides, if the only result of rising automation is that business owners make more money and widen the gap to the working class, I think we’re missing the full potential of AI.

There are countless ways this could be implemented. Maybe UBI is the answer. Maybe the proletariat need to seize the means of productions, overthrow the bourgeoisie and finally create a communist utopia (this time it’ll finally work I’m sure). Or maybe anyone profiting off of generative AI should be taxed more for that work, since they’re really profiting off the work of countless other people.

I’m no economist. Nor a politician, sociologist or in any other profession, that allows me to guess at the feasibility of changes to our current economic model. But I am a person, who works in an industry that is increasingly having to deal with, how rapidly evolving technology affects our current economic model. And right now, it seems more and more likely that the prospects are looking bleak. Across the media, academic and tech industry veterans and novices alike fear the repercussions of unleashing a technology that has the potential to cut away the need for their services in a single fell stroke.

But maybe they don’t have to. If we are prepared to make changes, maybe there is a feasible future, where AI-powered automation can allow us the freedom to continue the rise in living standards, that has followed the technological advancements since the Industrial Revolution. And we can sit in the sun, eating machine-grown and harvested crops, and think about the days when creating a simple image took a full day. Now we can leave the lesser tasks of necessity to the machines we once feared. Hopefully the robots won’t mind.