AI: Sword, Shield, or Shortcut? Why the Real Challenge Is Still Us

Scroll through the news, and it’s impossible not to run into flashy headlines about artificial intelligence. Some talk about the end of humanity, others promise a dreamy utopia where machines handle everything for us. But as Narayanan and Kapoor write in AI as Normal Technology1, the truth is far less cinematic—and maybe more interesting. AI isn’t a god or a monster; it’s just another technology. Powerful, sure, but still a tool, like the many before it.

What really matters isn’t what AI can do—it’s what we decide to do with it. The biggest challenge isn’t losing control to machines, but figuring out how to weave AI into society without losing sight of ethics, education, and human judgment. The media hype—whether utopian or dystopian—often distracts from what counts: finding ways for AI to support human values, knowing its limits, and preparing ourselves for gradual, meaningful change.

Let’s make this discussion concrete. I’ll bring in a few examples from Italy’s digital transformation2, some data on tech adoption, and reflections on how our habits as learners and citizens are shifting in this new era.

Thesis 1: Tools Amplify; They Don’t Rule

AI headlines love drama. Either it’s here to save us, or it’s going to replace us all. But as Narayanan and Kapoor remind us1, AI isn’t magic—it’s a Swiss Army knife. Versatile, yes, but only as useful as the person holding it.

If we look closer, most “intelligent” technologies—the ones used in government or public services—aren’t taking over our lives; they’re simply making things work better. Take Italy’s digital tax system, for example. AI tools help cross‑check databases and pre‑compile tax returns2. The result? Less paperwork, less tax evasion, more trust. When people see clear benefits—saving time, saving money—adoption grows fast. Technology doesn’t succeed because it’s shiny, but because it’s useful.

Above mentioned article suggests that AI’s growth won’t be sudden, but will unfold gradually—through a normal process of adoption and integration into existing technologies. Still, I believe that given how easily information spreads today, this growth will happen faster than it did for previous technologies, and will accelerate even more if AI continues to prove useful in solving real human problems. That’s why it’s crucial to move just as quickly on regulation and to build systems that are resilient by design. But while the question of how fast AI will grow is important, an equally pressing one is how much we should trust it once it’s everywhere.

Thesis 2: The Trust Paradox

Narayanan and Kapoor highlighted a certain pessimism toward artificial intelligence, but during my research, I noticed the opposite trend among its users: an almost blind trust. 

One study found that when Google’s AI Overview sits on top of search results, only about 8% of users click the links below. Without the AI summary, that number jumps to 15%. Even fewer people—less than 1%—bother to verify the original sources3. That’s not exactly a recipe for critical thinking.

AI is like a Ferrari that allows us to move fast, automating tasks that are often repetitive or of low added value. But, as with any high-speed vehicle, the faster we go, the greater the risks become.
The point, then, is not to stop or to go as fast as possible, but to learn how to drive. The human role remains central, as Narayanan and Kapoor emphasize in AI as Normal Technology from the Knight First Amendment Institute: it is up to us to monitor, correct, and educate. We must maintain control over the tasks AI performs for us, but also train people to use this tool consciously.

Thesis 3: Education and Regulation—Our Real Defenses

Every new technology cuts both ways. A hammer builds a house or breaks a window—it depends on who’s holding it. AI is no different. The point isn’t to step on the gas or hit the brakes, but to learn to drive responsibly. That means education—lots of it. Not just coding, but understanding how AI works and when to trust it.

The real danger isn’t that machines will take control, but that we’ll lose depth in how we learn. We’ve grown used to quick answers, short videos, and instant gratification. With AI tools, information arrives instantly—but deep thinking takes time.

I remember realizing this during a dance pedagogy course: real learning follows a rhythm. You can’t rush coordination, reflection, or understanding. It takes repetition and pauses. But now, in a culture of 30‑second content and AI summaries, those pauses are disappearing.

That could lead us down two unhelpful paths. Some people will use AI just to take shortcuts—skimming instead of learning. Others will reject it entirely, scared by the hype. Both miss the point, and both waste AI’s true potential.

In Conclusion: Becoming Wise Navigators

AI isn’t a runaway train—it’s a GPS. It helps us navigate, but if we follow it blindly, we can still get lost. The answer is balance: keep the GPS, but learn to read the map too.

The next chapter of AI won’t be about machines outsmarting humans. It’ll be about people staying curious, skeptical, and engaged. If we keep asking questions, verifying answers, and learning how things work, AI stays what it should be—a tool, not a tyrant. And that’s how we make sure society grows stronger, not weaker, with it.


Footnotes

  1. Knight Columbia: AI as Normal Technology ↩ ↩2
  2. Digitalization and cross-border tax fraud: evidence from e-invoicing in Italy, International Tax & Public Finance. SpringerLink ↩ ↩2
  3. Google users are less likely to click on links when an AI summary appears (Pew Research) ↩

Commenti

Posta un commento