The missing soul in the digital feed

I have been hearing a lot about AI slop recently. Our digital feeds are becoming a landfill for the “AI slop” with a relentless, synthetic sludge of text, code, images and video clips. As AI continues to generate more and more synthetic data to saturate the internet with its own outputs, AI continues to “learn” its own synthetic data repeatedly. This is the inbreeding of information. This is leading to the quality of Generative AI outputs depleting. While Big Tech markets Generative AI as a pinnacle of efficiency, we are in the midst of a silent, catastrophic trade: we are surrendering our cognitive inheritance for the convenience of an automated void.

As more people outsource their tasks ranging from writing computer code to producing documentation and multimedia artefacts, to the machine, the supply of raw human data is drying up because the more people rely on AI to generate content, the fewer people produce raw human data. This will lead to the decrease of human creativity and curiosity. This isn’t just about bad prose; it is about the systematic erosion of human creativity and the richness of human language history since the ancient time. We are standing at a precipice where original thoughts and ideas are being replaced by recycled patterns. If we do not do something about it, I’m afraid we might find ourselves unable to think and to speak in a voice that is truly our own.

We are drifting toward a world where data feeds on its own echoes – a sterile environment where every new piece of content is merely a reshuffled, degraded version of what a machine previously guessed a human might say. When we stop producing new human raw data, the machine stops “learning” new things and subsequently the digital landscape flattens into a featureless desert without any newly grown plants.

AGI can think like human, not!!

AI does not think, even though Big Tech claims AGI can think like humans. However, AGI does not exist yet and I do not think it will ever exist, as AI does not have emotional feelings like humans. AI does not know what is right or wrong, or what is good or bad in a human philosophical sense. This distinction is vital. Human thought is an act of will, rooted in ethics, empathy and emotion. AI, conversely, is a feelingless and emotionless mathematical prediction machine. When we mistake a statistical probability for a thinking mind, we devalue the very philosophical inquiry that makes us human. We are trading our judgment and common sense for calculation.

Are humans training AI or being trained by AI to think like machines?

AI is a pattern matcher, not a mind reader. My biggest concern is: whilst we humans are trying to improve our prompt engineering skills to be more effective when we communicate with AI through the chat window to issue our commands to AI, are we trying to train AI to understand human language structure better? Or is AI trying to subtly train us to think, speak, and write like a machine? Are humans going to lose our creativity, empathy, emotion, curiosity, and philosophy if our human brain is programmed by AI?

This is a cognitive atrophy in real time. By constantly refining our commands to suit the rigid and predictable structures of machine learning, we are subtly training ourselves to communicate with the AI chat window without empathy, curiosity or emotion. We don’t even expect any surprises, out-of-the-box or light bulb moment outputs generated by AI. We are in fact re-programming our own brains to think like machines. We are losing the very faculties of our own philosophy because our brain is being optimised for machine interaction. We are not just using an AI tool; we are being reshaped by it.

Linguistic challenges: the acceptance of mediocre standard

I am already seeing more and more poor-quality LinkedIn posts, and some of them are pretty obviously AI-generated. I am not saying AI should not be used, but there is no soul and human touch in those posts. The choice of words chosen by AI is based on mathematical and statistical probability, which means the more popular the word, the more likely AI will predict it for the next token. Therefore, it is not the best words AI is choosing as they lack the statistical weight, but rather the most used words, so the best words are being ignored. I am also seeing more and more vocabulary disappearing due to not being as popular as those mediocre words.

In this new linguistic economy, vocabulary is disappearing. Rich, specific and sophisticated words are being discarded because they lack the statistical weight of their more popular counterparts. We are witnessing a trend where linguistic diversity is sacrificed at the altar of mathematical likelihood and statistical popularity, resulting in a standardised and soul-less form of Generative AI outputs. It is a race to the bottom of the most common denominator. It is quite heartbreaking to watch the vibrant colours of human language fading into a dull, algorithmic grey. Perhaps the most challenging impact of the AI slop is that we human are becoming more accepting the mediocre outputs generated by AI

The way forward: Critical Thinking and Validation

The way forward is a radical recommitment to critical thinking to validate the outputs generated by AI. We must become the ultimate arbiters of quality and standard. We must refuse to be fooled by the slick convenience of AI slop. We need to protect our literacy and must be able to distinguish between a mathematical prediction and a genuine human insight. We must demand more from our cognitive self than the machine can ever provide.

As the digital world becomes increasingly synthetic, we must ask ourselves a question: How can we continue to produce our raw, human spark data full of creativity, curiosity, emotion? If we don’t fight for our own mind, the algorithm will be more than happy to fill the void for us!