The Tech Industry’s Illusion of Innovation

Martyn Rhisiart Jones

Oza-Cesuras, 4th December 2025

There’s a moment in every hype cycle when reality taps the industry on the shoulder and whispers, “This isn‘t working.“ And every time that moment arrives, tech responds with the grace of a cornered raccoon.

Suddenly we’re informed by people who say “paradigm” without blinking. They claim the miracle failed to materialise because the data wasn’t spiritually exfoliated. They argue the talent wasn’t elite enough to breathe the rarefied air of innovation. The budget didn’t ascend into the fiscal stratosphere. Furthermore, a clandestine sect of “haters” projected bad vibes that knocked the innovation chakra out of alignment. Anything, anything, to avoid confronting the sacrilegious notion that the tech itself is a glittering, investor-soaked dud.

We experienced this during the dot-com frenzy. At that time, half the startups were websites in search of a reason to exist. Then big data swaggered in. It promised a revolution but delivered mostly slide decks. Hadoop clusters became mausoleums for CSVs. Data lakes flooded. Data lakehouses tried to slap lipstick on the pig. Then they blamed enterprise IT for failing to appreciate the avant-garde architecture. Data mesh took the chaos of org politics and gave it a manifesto.

Now the industry has discovered generative and “agentic” AI. It is a hype engine so powerful it might register as a controlled substance. And the excuses when it underdelivers? Weaponised. Institutionalised. Spiritualised. The gospel of “maybe if you spent another $12 million fine-tuning it, you’d see ROI” is the new prosperity theology.

But let’s cut the crap: the problem isn’t your data. The problem isn’t your engineers. The problem isn’t the sceptics.

The problem is an industry that long ago decided selling fantasies is more profitable than building tools that work.

Tech has become a self-licking ice cream cone of hype, an ouroboros swallowing its own press releases. Real problems are boring. Real engineering is complex. We end up with magical thinking. It is capitalised into a product roadmap. It is pushed out with the urgency of a hostage video: “This changes everything.” No, it doesn’t. It barely changes the PowerPoint template.

Consider the current crop of AI “agents.” They are promoted as autonomous digital workers. These AI agents are seen as tireless cognitive athletes. They will sprint through your workflows while you sip an oat-milk latte. In reality, most agents are brittle little goblins that hallucinate their way into a ditch the moment they encounter ambiguity. They don’t replace employees—they generate more work for them. They’re not assistants… they’re chaos monkeys with a UI.

And yet the sales pitch remains unwavering:

If the AI fails, it is you who is unworthy.

• Your data is impure.
• Your infrastructure is medieval.
• Your expectations are naïve.
• Your enterprise is insufficiently “transformed.”

This is not innovation. This is gaslighting.

Tech’s dirtiest secret is that hype isn’t a byproduct, it’s the product. The goal isn’t to build systems that solve problems. The real goal is to build belief that salvation is always one release cycle away. Just keep paying. Just keep deploying. Just keep praying to the roadmap.

The industry’s “thought leaders” serve as hype sommeliers. They pour fresh glasses of jargon to dull your scepticism. Terms like foundation models, synthetic corpora, autonomous cognition, multimodal synergy, and emergent agent collectives fill the air. It’s not language; it’s a smokescreen.

If any other industry behaved this way, for example, in finance, medicine, and aviation, we’d call it malpractice. In tech, we call it 4QTR.

The tragic part? The real breakthroughs, the real engineers, the real craft get buried under the avalanche of bullshit. The industry is so addicted to spectacle that genuine progress is suffocating under the weight of its carnival barkers.

Maybe that’s the real scandal. It’s not that tech keeps getting it wrong. It’s that it continues to insist on getting it wrong loudly, expensively, and with a straight face.

The cure is apparent, but unfashionable: stop worshipping the baubles. Stop mistaking demos for destiny. Stop forgiving the industry for selling fantasies wrapped in jargon.

Until then, the hype machine will keep spinning. The excuses will continue to mutate. Reality will keep dragging the industry back to earth. It will kick, scream, and blame the datasets.

In the end, the hype is inflated. The only thing more inflated is the certainty that the hype must never be questioned.

Many thanks for reading.