Building a Living Newsfeed for the Age of Automation

I use my website as a living playground – a place to experiment with systems that push the boundaries of what's possible for clients. Every idea is tested in motion, built not just to function but to learn.

The news page started as one of those experiments – a test of how far automation can go before it begins to resemble discernment. The goal wasn't to publish more, but to understand whether a machine could care about relevance. Whether it could tell the difference between information and insight.

"The goal wasn't to publish more, but to understand whether a machine could care about relevance. Whether it could tell the difference between information and insight."

The Idea

We're long past the age where the internet needed more content. The challenge now is to extract meaning from the noise.

The modern search landscape no longer rewards volume or clever keyword placement. It rewards coherence. Search systems are evolving from indexers into interpreters – they don't just look for mentions; they look for relationships. They infer authority not from how often you speak, but from what connects the things you say.

So the idea was simple: build a news system that mirrors this shift – one that curates the best of the technology, design, and AI worlds, using the same logic that underpins how search itself now thinks.

How It Works

The system functions like a living editorial organism.

Sources → Adapters → Filtering Pipeline → Categorisation → Cache → UI

Each source – Guardian, NYTimes, arXiv, Product Hunt, YouTube, and others – is normalised into a common format and scored against quality weights, inclusion rules, and conceptual density.

Articles pass through a series of filters that remove consumer tech, politics, and superficial trends, leaving only professional and research-grade material. The categoriser then classifies each piece into one or more of seven pillars – Technology, Design, AI, Innovation, Strategy, Future, and My Articles – mirroring the mental architecture of my own practice.

But the final step – the one that makes this system different – is the AI curator.

The AI as the Last Filter

After every mechanical filter has done its work – scoring, categorising, normalising – the AI layer reviews what remains. It doesn't measure by keyword or frequency; it measures by meaning.

It asks:

"Is this aligned with the intent of this site?" "Does this represent thinking rather than noise?" "Would this deepen understanding for the kind of reader who belongs here?"

"This AI doesn't create content; it judges it. It's a digital editor with no ego – a synthetic sense of taste."

This AI doesn't create content; it judges it. It's a digital editor with no ego – a synthetic sense of taste.

Where older algorithms optimise for clicks, this one optimises for coherence. It decides what belongs in the conversation – and what doesn't.

That's the future of content: not more, but better chosen.

Why I Built It

Because automation isn't the enemy of creativity – it's the next medium for it. As search evolves, visibility is no longer earned through performance metrics but through structured trust. Search engines now reward patterns of thought that are consistent, contextual, and grounded.

The more a site reflects a clear domain of understanding – with data, structure, and credible curation – the more it becomes part of the web's collective intelligence.

So I built this system as both experiment and statement: A website that teaches algorithms what I know – not through claims, but through consistent demonstration.

From SEO to AEO

Traditional SEO tried to make websites more discoverable. Answer Engine Optimisation (AEO) makes them understandable.

In the emerging landscape, it's not about "ranking first," but about being referenced – cited by models that synthesise knowledge.

To do that, your site must become a structured dataset of expertise. That's what this system is: a self-curating, semantically consistent, always-fresh dataset of what matters in my field. It doesn't fight for visibility; it trains recognition.

Each article is stored with schema, entity tags, and relational metadata – a web of structured context. When an AI crawler arrives, it doesn't just see content; it sees competence.

The Benefits of a Living System

• Perpetual relevance – The site updates itself with curated, topic-specific content.
• Authority signalling – Structured, trustworthy data builds semantic weight over time.
• Editorial automation – The machine filters the world, leaving only what aligns with the brand's intelligence.
• AI discoverability – Well-structured data feeds reasoning engines, not just search indexes.
• Sustainability – The system grows, learns, and refines without constant human labour.

It's not a blog. It's a knowledge engine.

Why It Matters

We're entering a new phase of the web – one where everything we publish teaches something to the systems that interpret us. Our websites are no longer just for humans. They are for the machines that explain us to humans.

To exist online now is to participate in the training of perception – in the ongoing dialogue between human voice and artificial understanding.

My news page is just one manifestation of that idea: A self-curating reflection of what I value, what I read, and what I believe is worth attention.

It doesn't speak for me – it thinks alongside me.

Experience It Yourself

If you're curious to see this system in action – visit the news page.

"Automation isn't replacing human judgement. It's mirroring it – giving our thinking a way to live autonomously in the digital world."

– Leantonio