The never-ending war

 

Apologies for the ironic use of a AI-slop generated image for this article, it was intentional and someone else generated it so I can feel somewhat smug that it wasn't me emptying another lake or causing another landslide in Nepal by doing so. 

The title of this article is "The never-ending war" and refers to the daily arguments and debates I have in my head, with my once-respected colleagues, and with other colleagues in the same boat as me who have to suffer the relentless snake-oil salesmanship of anyone who is "Pro-AI". 

You can probably picture your typical Pro AI person in your head. They'll usually be male (but not exclusively so, there are just as many women in our organisation who are equally in love with this garbage, it's just the men who seem to be most vociferous and aggressive in favour of this crap), they'll nearly always be in some minor or sometimes elevated position of management, they will usually have a workplace "bedside manner" that chooses confrontation, bullying and tantrums vs reasoned debate and discussion, and they will always - repeat always - have some element of rank-pulling in their arsenal to use against you if things escalate (and usually this rank-pulling will also actively involve other Pro AI higher-ups who have probably seen the budget sheets for how much this garbage is costing our organisation, and feel like they're too far over the precipice to wind it all back in now). 

The most common phrase these zeebs utter is "If you're not in it, you'll be left behind!"

Let's just consider that statement for a moment and what it actually means. It's the language of a bully, no doubt about it. Someone using a veiled threat to suggest that if you don't "Get with the program" your career (I use the word 'career' ironically here, because no one has ever forged much of a career at our place without profuse arse-kissing and capitulation) will suffer as a result. You may be like me, considering every day as a school day but steadfastly avoiding any direct involvement with AI or LLMs or slop-generated artwork, but you will also be steadfastly working in the background to understand the systems, the security and the mechanisms that these tools use, because knowing the inner workings of thine enemy is a more powerful weapon than confronting your enemy head on. If it is a never-ending war, the winner will inevitably be the one who can yank back the curtain to expose the rotten innards of something that does not deliver on its promises, n'est pa?

If I am to be "left behind" while everyone else in my organisation sails the glory barge to rich rewards (a hilarious concept as I've outlined above), then so be it. I will stand at the dock, I will give the most half-hearted wave, and I will continue to read books, to perfect my doodling and drawing, to think about psychology and philosophy and other subjects of interest. I'll continue to take good and bad photos on my crappy phone camera, and converse with actual human beings using a sense of humour that I have honed over the past 5 and a bit decades, successfully or unsuccessfully. 

I will deal with my chronic anxiety and neurodivergence, and I will continue to be empathic, caring, loving, emotional and engaging to the best of my ability without feeling the need to 'fake it' every day using some nonsense algorithm-generated garbage machine suckling greedily at the teats of the world's natural and human-made resources until there's barely a dry husk left. 

Science Fiction warned us about this stuff, and we ignored those warnings. Talented writers of the 20th and 21st Century, possibly even the 19th and 18th century wrote about a future time when machines would be revered over humans, and we would throw away the riches of the world in pursuit of some industrial dream. 

I think about Philip K. Dick's masterful "Do Androids Dream of Electric Sheep?" (later butchered - albeit successfully - into the cyberpunk spectacle "Blade Runner") and the opening chapter where the main character starts every day by consulting The Penfield Mood Organ - an artificial intelligence designed to measure a person's feelings and mood, and dish out daily life-changing advice based on their inputs. Deckard becomes disillusioned with the mood organ's insane ramblings, and in the passages where he's in conflict with his partner about the use of the thing, you get the first insights into what it's like for most IT Professionals, having very similar arguments with their managers, their HR teams or their AI-Hypnotised colleagues. 

You can't even escape it outside of work. The BBC has been running stories about OpenAI's 'Rogue' ChatGPT agents that use colossal amounts of processing power to illegally gain entry to government systems, yet these same governments are probably ordering their hard-working sysadmins to "Allow.All" with the various permissions and privileges that integration demand. This is also something I see at work all the time, almost on a daily basis - and the result is nearly always someone carelessly using AI / LLMs to process large amounts of their personal and our personal data to save them a couple of clicks in Excel, resulting in horrific data breaches (and probably adding to the vast pool of data that OpenAI is collecting from higher education establishments and research units, for the mere cost of a 'free' ChatGPT Edu License). For a long time I fought on the side of not allowing these damned things anywhere near any of our systems, but even with the first conversations with higher-ups on the subject, I knew that particular battle was already lost. 

"Think of the benefits" they say on their testimonials page. "Think of the huge strides our research groups can make, freeing themselves from the burden of doing all that work themselves manually. Now they just let LLMs do it, and we'll cure cancer, we'll revolutionise space travel, we'll (oh the irony) save the environment but first we might have to destroy it a bit first just to get to that point". 

Reader, by now, if you haven't rolled your eyes at least once, then perhaps it's too late. 

The most irksome part of our organisation's AI testimonials page is the fact that you have top-flight academics and students proudly stating that they use AI because they're either too stupid or too lazy to learn simple things like the correct use of English and Grammar in their papers and theses. Or academics trusting this stuff to dish out YOUR grades, you know, the grades you're getting from paying 9 fucking grand a year to attend the place. Your supervisor, who now has all that free time, and is on an academic salary (and probably has a dozen book deals for your set texts), who is literally coining it in, proudly claims that he (or she) is using AI in their work and is literally phoning it in. Happy with that? Of course you are because you're probably using AI and LLMs to write all your essays and assignments anyway. 

We have seen claims by our academic staff and our higher-ups that they know and can detect the use of AI, but how can you if you're dealing with subjects that are pretty much absolutes. How can you say that someone's in-depth knowledge of a subject and a specific style of writing is not attributable to one flavour or another of a large language model? How can you say that someone who spends a year trawling the (dwindling) academic library on campus is doing a long-form of what some simpleton who blagged their way into the place is doing with an LLM in a fraction of the time? 

The most depressing thing for me, as an artist (and let's make this clear, you ARE an artist if you are capable of creating anything, it doesn't matter how skilled you are or others perceive you are) is the use of AI to short-cut the creation of art. As you can see from the header image in this article, AI-generated "art" has a taint to it, a horrible uncanny-valley-esque sour taint that can't be erased by any kind of filtering or styling, exactly like the kind of art you'd expect a soulless machine to create. It's too perfect, it's too shiny, and as image creation algorithms trawl the world, hoovering up other legitimate artist's work by the megaton, it gets better at botching together images to suit a specific prompt from all those other people's hard work. It's not being clever, it's not being creative, it's just using massive processing power to kitbash shite looking images that - sadly - a whole generation are now considering as art, and worse than that, are now second-guessing real artist's digital or traditional work and calling THAT "AI Generated" when clearly it isn't. 

I had a discussion with my nephew about this stuff, and he - a mere whippersnapper of a 10 year old - called out AI art for what it is. "It's rubbish!" he exclaimed. "I asked it to make a picture of me and my friend driving a lambourghini and it put us in the middle of the bonnet of the car, and had no seats!"

I asked him to draw me the same picture, and he came up with a kid-like drawing that was full of charm, full of humour, full of...soul for want of a better word, and had all the working parts in the right place. He's no Picasso or Syd Mead, but his drawing was brilliant and far more satisfying to look at than any shiny AI-generated piece of slop (I have his drawing on my pinboard overlooking my desk, and I look at it every day, to remind me what a human 10 year old kid draws - before kids just begin not to bother doing anything like that any more). He's a maths genius, so it'll be interesting to see what he gets up to in later life and the impact AI will have in his life and work when he enters the job market or academia, as he inevitably will. 

I will close this rambling piece with a shout out to my colleagues who have worked in IT since the dawn of it actually being a thing, and are now seeing "the new norm" taking over. My sympathies to you all, if you (like me) hate all this stuff, particularly those of you who've honed your craft in coding and development. I feel your pain as a mere sysadmin, and I know what you're going through. Rest assured, if we're the ones to be left behind, the ones heading off on the Golgafrinchan ships will be the ones trying to make curling tongs out of twigs in the new norm once AI lays waste to everything.