AI Perception Gap: The Journey Concludes

AI Perception Gap’s Aim

This blog concludes the six-part AI Perception Gap series I introduced in November 2024. In that first entry, I expressed my astonishment at Gen Z’s perception of AI — observed during a backpacking trip to Latin America — with this reflection:

Educated individuals, from my very generation (not my grandparents!), who were born and raised in wealthy economies where AI is booming, and no one had a clue about the monumental positive impact the technology can and will have on society at a global scale! The perception gap was born…

The problem? Most of Gen Z’s perception of AI was built by mainstream media’s doom-and-gloom obsession. In Latin America, I found myself discussing AI’s positive potential with backpacker after backpacker. Watching their attitudes transform from fear to fascination was so rewarding that I felt compelled to amplify this mission through blogging. My goal was crystal clear: make my friends and the rest of Gen Z aware of the positive potential of AI and prove that “AI” doesn’t automatically mean “killer robots destined to destroy humanity”.

Positive Impacting AI

Since announcing my aim in November 2024, I’ve addressed the AI Perception Gap by showing that, when aligned with human goals, AI can:

1. identify human fragments in the aftermath of a plane crash (edition 1),
2. assist with discovering critical minerals (edition 2),
3. decode the structures of over 200 million proteins, reinventing drug discovery (edition 4),
4. uncover the atomic blueprint of materials that could reshape entire industries (edition 6).

In each case, AI accomplished what human cognition alone couldn’t — or what would have required the coordination of millions of superhuman brains. The world’s most brilliant scientists spend a lifetime mastering just one domain, often with little exposure to others. AI, on the other hand, achieves superhuman performance in one field at one moment, and seamlessly shifts to another field — with equally impressive performance — the next. This level of cross-domain proficiency isn’t just impressive — it’s the very essence of what makes AI revolutionary.

Negative Impacting AI

While most of this series explored AI’s positive potential, editions 3 and 5 addressed its risks. This was necessary since closing the AI Perception Gap requires the confrontation of Gen Z’s fears.

I categorise these into loud fears — AGI going full sci-fi villain — and quiet fears — like hidden algorithmic bias embedded in decision-making systems.

In edition 3, we explored how AI can quietly reinforce inequality (a quiet fear). PredPol, a policing tool built with good intentions, was trained on biased historical data and repeatedly deployed officers into already over-policed communities. The designers had noble goals but made a crucial oversight — they didn’t account for how their data reflected existing social inequalities. As AI systems become woven into the fabric of public life, we desperately need designers who understand both the mathematics of machine learning and the messy realities of human society.

Edition 5, on the other hand, leaned into the fear most often voiced by Gen Z: AGI with lethal autonomy. We imagined a system that identifies threats from satellite feeds, drafts tactical plans, and executes strikes with drone swarms — all without human input. Gen Z’s fear is valid, but this reality is not destiny. What gives me hope is that democratic nations are treating this like nuclear tech: seriously, cautiously, and collaboratively, with guardrails already taking shape. Just as nuclear deterrence has prevented catastrophe for decades, coordinated regulation and transparent safety efforts can do the same for AGI.

Did AI Perception Gap Achieve Its Aim?

That’s for you, the reader, to answer! If you’re still reading at my eighth (and final!) entry, I’d love to hear how your perceptions changed during the series. I’ve spent over 100 hours on this project; any sense of reward would no doubt put a smile on my face!

Anyhow, I hope you too were smiling whilst reading: AI Perception Gap, over and out.