The Intent Gap-How AI Exposes the Gap between Input & Intent

---

26 July 2025

“AI will make the lazy thinkers lazier… and the ambitious thinkers more ambitious.”
— Jim Hays

We are not in an AI revolution—we’re in a human one.

This isn’t the first time we’ve faced a cognitive reckoning.

More than a decade ago, in my doctoral research at Cranfield, I studied how fast-growing datasets in the retail and FMCG industries were outpacing teams’ ability to extract meaning. Even then, we were already producing more information than decision-makers could meaningfully absorb. The conclusion was clear: data was abundant, but actionable insight remained scarce.

That was 2014.

Back then, we were talking in terms of terabytes. Today, we’re generating over 120 zettabytes of data per year, with IDC projecting nearly double that by the end of 2025. In just one decade, we’ve moved from information overload to something more structural: a breakdown in the signal-to-noise ratio of modern work.

Yet human cognition hasn’t evolved in kind. Our working memory remains capped at around 7±2 items (Miller, 1956). Our decision fatigue kicks in after just 20–25 meaningful choices a day (Baumeister, 2002). And our ability to derive wisdom from information is increasingly compromised by the sheer speed and scale of what surrounds us.

Now comes generative AI.

For some, it’s a powerful amplifier of human insight.

For others, it becomes a shortcut to avoid effort, responsibility, and thought.

This is the Intent Gap: the accelerating divide between those who use AI to extend their capacity—and those who use it to escape their limitations.

Generative AI is, in theory, the most democratized tool of our age. Anyone with an internet connection can access it. But in practice, its impact depends entirely on the user.

Recent research from Stanford HAI shows that users who actively refine their prompts, verify outputs, and challenge assumptions produce significantly better outcomes than those who simply “ask and accept.” MIT studies have echoed this: a small investment in intentionality—just 10 minutes of structured prompting—can produce a 40% improvement in relevance and task effectiveness.

The difference isn’t in the tool. It’s in the intention behind it.

Lazy users get lazy results. Ambitious thinkers get compounded leverage.

The most advanced machine learning models can simulate language, analyze markets, and even draft strategic memos. But they still lack one critical thing: desire. They can simulate knowledge, but not will. They can process data, but not care.

And that’s where we come in.

We’ve long described class divisions by job types—manual vs cognitive, blue collar vs white collar. But in the AI age, the more important divide is mental, not occupational.

It’s the split between those who build with AI—and those who wait to be carried by it.

MIT economist David Autor warned in 2024 of a coming “Mad Max” scenario—not because jobs would disappear, but because valuable skills would be commoditized by automation. As AI systems replicate once-differentiating tasks—copywriting, forecasting, strategy synthesis—the edge no longer lies in what you can execute.

It lies in what you choose to focus on, filter out, and shape with intent.

This is not automation replacing workers. It’s automation revealing the hollowness of shallow work. It’s not just jobs that will be lost—it’s the perceived intelligence of people who allowed tools to think on their behalf.

In short: AI doesn’t close gaps. It amplifies them.

So, what do the ambitious thinkers do differently?

They don’t treat AI as a destination. They treat it as a launchpad. They use it not to replace thought—but to accelerate feedback loops, challenge assumptions, and explore edge ideas faster.

At Minor Hotels, along with our Technical AI Center of Excellence, we launched SAIGE—Strategic AI for Growth & Enablement within our Commercial team —not to follow a trend, but to shape a future. SAIGE is built on one question:

What are the uniquely human capabilities applied to commercial hospitaltity and global travel demand influence we are trying to augment with this tool?

From revenue forecasting to guest sentiment analysis to multi-market content generation, the goal is never just speed or automation. It’s enhanced decision quality. It’s using AI to do what it does best—pattern recognition at scale—so our teams can do what we do best: exercise judgment, creativity, and empathy.

And that’s the crux: the best users of AI are those who remain fully human in how they apply it.

As leaders, our job isn’t to chase platforms. It’s to cultivate intentionality.

This means building cultures where people are trained to:

• Ask sharper questions, not just accept faster answers
• Understand the limits of machine cognition—its biases, blind spots, and false confidence
• Pair technical fluency with strategic clarity
• Use AI to multiply their strengths, not mask their weaknesses

This is not about digital literacy alone. It’s about intellectual discipline.

And just like in my earlier academic work, the difference still lies not in the volume of input—but in the clarity of action that follows.

There’s a powerful illusion at play today: that more information equals more intelligence. That better tools equal better outcomes. That faster answers mean smarter decisions.

But none of this is true—unless intent leads the way.

Generative AI is an extraordinary breakthrough. But its greatest strength is also its greatest danger: it gives you exactly what you ask for, whether or not you understand what you’re asking.

In this new era, intelligence won’t be measured by what you know.

It will be measured by what you do with what you know—and what you choose to ignore.

“In the age of AI, the smartest minds won’t be the ones with the best tools—just the ones who refuse to let the tools do all the thinking.”
 

Same input.
Different intention.
Radically different futures.

1. IDC (2023). Global DataSphere Forecast, 2023–2027.
   → Cited for global annual data creation (120 zettabytes in 2023, projected to double by 2025).
   https://www.idc.com/getdoc.jsp?containerId=US50552423
2. George A. Miller (1956). “The Magical Number Seven, Plus or Minus Two.”
   → Foundational paper on limits of human working memory.
   https://psychclassics.yorku.ca/Miller/
3. Roy Baumeister et al. (2002). “Decision Fatigue Exhausts Self-Regulatory Resources.”
   → Source for cognitive limitations on high-stakes decisions per day.
   https://psycnet.apa.org/doi/10.1037/0022-3514.82.4.563
4. Stanford HAI (2023). “Measuring Human-AI Collaboration Outcomes.”
   → Demonstrated quality variance based on user engagement and prompting effort.
   https://hai.stanford.edu/research/ai-collaboration
5. MIT Sloan Management Review (2024). “The Secret to Better AI Results? Ask Smarter Questions.”
   → Cited for finding that prompt refinement boosts task relevance by 40%.
   https://sloanreview.mit.edu/article/the-secret-to-better-ai-results/
6. David Autor, MIT (2024). “Work of the Future” Initiative and public comments.
   → Referenced for warnings on AI devaluing cognitive labor and creating skill abundance.
   https://workofthefuture.mit.edu
7. Arthur Mensch, CEO of Mistral AI (2024). VivaTech Conference Interview.
   → Cited for comment on “deskilling” as the primary AI risk.
   https://www.businessinsider.com/mistral-ai-ceo-risk-ai-lazy-deskilling-dario-amodei-jobs-2025-6
8. Ian Di Tullio (2014). PhD Thesis – Cranfield University: “Improving the Direct Marketing Practices of FMCG Retailers Through Better Customer Selection.”
   → Referenced for original research on information-action gaps and decision-making limits in data-rich environments.