The Great Binary Layer Cake Collapse (contains AI)

I am not a philosopher. Not because I wouldn’t like spending my spare time pondering questions without real answers (“real” like in the simpler terms of physics and math). A good philosopher knows their peers, has a solid foundation of the classics and actually applies scientific craftsmanship. You know, reciting correctly and doing real research, falsifying their own theories, advancing the specific field of science and stuff.

That’s not me. I am spewing individual, unscientific opinion here, shooting from the hip.

“It will be a bloodbath.” (Elena Verna performing an excellent sales pitch, Product at Heart 2025)

There’s a lot of buzz going on about AI, you probably have noticed, and naturally a lot of emotionality and guess-work is involved (qed). Personally I am equally excited and scared shitless about what might happen in the next 1 to 20 years. Not because I expect Skynet deploying killer robots (which still is a possibility, I mean those LLMs are trained on our culture), but a fundamental shift will occur that we, as a global society, will collectively experience. We will have to deal with this somehow and as the little egoistic and panicking monkeys that we are, I expect us to be screaming and flinging poo at each other.

Let me explain: Somewhere in the 19th century science kicked into gear and started what I just phrase the “30 ms machine age” - 30 seconds because we, as humans are quite slow in signal recognition and reaction time - thinking not included, mind you. We also have massive problems with scale and complex data. And as the world filled with more and more people due to fertilizer and industrial livestock farming and science got so massive that the era of universal geniuses came to an end - it became obvious that humans needed helpers.

Binary Machines on the rise

Introducing Herman Hollerith (S'ouvre dans une nouvelle fenêtre), inventing the punched card tabulation machine in 1884. It’s is the starting point of an unparalleled success story for automation and data processing at scale. So successful that it helped us, as in us Germans, to organize the genocide of millions of Jews and to enable the logistics for a world war. Thanks for supporting fascists comitting atrocities, dear IBM™.

This basically outlines why it was such a horrible success story: If one can automate calculation, data acquisition and research at such scales and even translate it into actions, it is a damn jackpot for any mass production economy. I once read a paper how Nazi Germany only was able to launch WW2 because they basically stole the money from the people they enslaved and killed. Sometimes I am wondering if we, Germans, were so pissed that we sucked at imperialism and slavery (in contrast to let’s say, the Brits, Dutch, Belgians, Russians, Spanish, Portugese and French) that we needed to show the world that we could equally excel and even surpass other nations at being a dick. I digress.

“Because it’s simple, it’s prone to cascades, and because it’s complex, you can’t predict what’s going to fail.” (Dr. Praxidike Meng, The Expanse)

For 150 years we, humans, have been refining those machines. We made them more flexible through digitization. We made them really small. We abstracted the most complex phenomena and translated it into computational matter (weather, 3D calculus, etc.). We established paradigms and languages and architectures and standards and it was glorious. It is glorious. The very core of our human being, our identities are now defined by the interaction with machines. Estimated 5.6 billion people (S'ouvre dans une nouvelle fenêtre) are using the internet. Every day, every second. Our societies and economies are built upon layers of layers of layers of understanding, more or less binary, machines. Layers for instructing machines. Layers for making machines easily accessible. Layers for maintaining machines and breaking them. Layers how to teach interaction with machines and coaching how to deal with other layers. Layers for getting entertained by machines. Layers for organizing all those layers. Layers for giving them a strategy. Layers for reaping loads of money from layers. Layers for hiring people for layers. Each layer maintained by dedicated layer specialists. It is a House of layer cards, or better, a universe of it (hey, just making sure that our monkey brains really cannot grasp the scale).

Let that sink in: 150 years of building layers potentially scrapped in just 3, 5 or 10 years. And I am not even yet talking about other rather loosely related professions which will be impacted by AI, e.g. lawyers reading, memorizing, understanding and applying legalese. I am not talking about e-commerce platforms losing direct customer access, crashing with a disrupted business model. Not talking about the plethora of people dropping out of creative jobs.

“Doors are closing, new windows are opening.”

Yes, yes, the practical optimists. They will be right until they are wrong and the same applies to this text in reverse. Only that there will be too many people who will not be able to pay rent and bills and mortgage and medical security and food. In the European Union alone over 10 million (S'ouvre dans une nouvelle fenêtre) people were employed in the information and communication sectors in 2024. Not all of them will be replaced, yes. But that’s still massive even if it’s seem like a drop in the ocean compared to ~197 million workforce. Especially because this calculation does not count in the cascades rippling through other sectors when high-paying jobs just die like flies because AI will be faster and cheaper and, most of the times, better. It will be 2008 again, but that was just a hiccup. Or the Dotcom bubble bursting but like this really sucks for everybody. I firmly believe, that this time, when the door smashes closed, many of us will be standing in the door frame and ultimately Elena’s quote goes proverbial.

Mankind faces two irreversible turning points, human-made Climate Change and the rise of Artificial Intelligence. Maybe they cancel each other out, maybe one fuels the other. I cannot believe, simply due to history and current events and rampant, narcissist apes, that individual, national, cultural and clerical divides can be overcome to unite for giving all of humanity (or the living beings on earth) a perspective to survive this gruesome cataclysm.

Some will. At some point in the future they will describe their survival as virtue and destiny and exceptionalism - but right now it’s just Sovjet Roulette.

If you disagree with my perspective, I welcome your thoughts. I'd love to see those windows of opportunity you envision. Otherwise, I recommend reading the paper AI 2027 (S'ouvre dans une nouvelle fenêtre) - and maybe also my upcoming, follow-up post why this change will happen, faster than many of us think.

In case this is all too much unprofessional doom oracle for you (it’s okay, I tend to see too much in binary), read Christina Wodtke’s piece “I Love Generative AI and Hate the Companies Building It” (S'ouvre dans une nouvelle fenêtre). It’s much more in the present, it has decent research and gives a far better picture which means some people justify for a unparalleled power and money grab.