Can We Trust Them?

I got one of those promotional emails from Google Workspace recently, pushing "Ask Gemini in Drive." Their pitch: get synthesized answers pulled from multiple files, no need to actually find and read them yourself.

Sounds convenient. But here's the thing I believe strongly in this age of AI slop and post truth internet you must always check your sources.

If there's a conspiracy or cartel among tech giants, it works in two parts.

First, they're slowly weaning us off our own capabilities. Our own PCs, our own software, our own data, our own creative work. Everything gets pushed toward a thin client model where it all lives in the cloud. Their cloud. Their software. Your data, despite endless assurances it stays private (except for training purposes, except for testing, and so on, have you actually read the updated terms of service lately). That means they control everything. Sound familiar? Too much like Orwell's Big Brother?

Second, while some features stay free for now, it quickly becomes clear you'll need to subscribe, and not just to one service either. The real hidden cost will be cloud storage, and eventually being locked out of what's actually yours. I'm not exaggerating. OneDrive has already started doing this, and not just to me.

So we'll be handed answers without needing to think, but we could pay dearly for the privilege.

Is this nirvana, or is it dystopia?

Proof read by Claude Sonnet 5

Jibberish on Gibberish?

 When Copies Aren't Perfect: A Visual Warning About AI LLM Training

We tend to trust that digital copies are exact replicas of the original. But what happens when we copy a copy, then copy that copy again?

Using my cartoon alter ego Alvin, I wish to demonstrate how the compressed JPEG image format, using its "lossy" compression method, degrades with each generation of copying. While JPEG saves disk space and speeds up downloads, it achieves this by discarding data each time you save.

Starting with a clear image of Alvin shouting "gibberish" at a screen, created in Coreldraw. I repeatedly saved and resaved it as a JPEG:

• At 80% quality (the common standard), the first 5 generations showed minor deterioration at high-contrast edges.

• Switching to 70% quality (used by many social media platforms), problems became obvious by generation 10

• At 50% quality, by generation 20, the image was almost unrecognisable. Even the mispelt word "gibberish" Alvin was shouting became illegible

Why This Matters

This isn't just about image quality. It's a powerful analogy for what's happening with large language models trained on "synthetic data", a dodgy term used by the LLM enthusiasts for AI-generated content fed back into AI systems.

Just as each JPEG generation compounds tiny adjustments until the image becomes gibberish, AI systems trained on their own output accumulate biases and inaccuracies. The feedback loop doesn't make things more accurate, it amplifies what's wrong.

When we assume digital processes are perfectly reliable, we miss how errors compound through iteration. Each cycle reinterprets the previous one, carrying forward and magnifying small mistakes, biases and fake stuff. Eventually, we're left with output that bears little resemblance to the original truth.

The lesson? Whether it's image compression or AI training, recursive copying without fresh input leads to degradation. Garbage in, garbage out, feeding this back in and the garbage out just gets worse.

Proof reading and summary assisted by Claude Sonnet 4.5 (AI)

Living in the AI Slopocene

We're living in a new era some call the "Slopocene" defined by overproduced, low-quality AI content flooding the internet. This isn't just about poor posts cluttering our feeds; it represents a fundamental erosion of information reliability.

The problem is multifaceted: people's trust in science, journalism, and credible sources is crumbling while tech platforms prioritise profits over truth. Meanwhile, AI-generated "slop" low-effort, high-volume content; dominates online spaces, creating what the Conversation article (see link above) describes as a potential future where "recursive training collapse turns the web into a haunted archive of confused bots and broken truths."

As AI systems increasingly train on AI-generated content, we risk a feedback loop of degrading information quality, like making photocopies of photocopies until meaning is lost.

This isn't the digital future anyone wanted. We're caught between technological dependence and growing awareness of these systems' limitations and misuse. While there are no simple solutions, being more intentional about information consumption, more critical of sources, and more supportive of genuine journalism and research becomes crucial. 

In this age of AI slop, authentic human insight has never been more valuable or more at risk.

I’m using this post to hold an interesting experiment. There are actually two blog posts one here and the other on Meet the People. They both express the same ideas, but one is an original written by Norm and the other is a reinterpretation by an AI.  Could I ask you to read both? Then leave a comment on the article you believe is generated by the AI (Claude Sonnet 4) and the post written directly by me.