Artificial Intelligence: The next technological revolution after the Internet?

The creation of the Internet at the end of the last century revolutionized our lives (for better or worse, depending on one's perspective…). But it didn't happen overnight; it took decades to profoundly transform our societies after its invention.

Artificial intelligence seems to be following a similar trajectory. Often overestimated in the short term—just look at TV shows or my LinkedIn feed—but largely underestimated in its long-term impact, AI progresses at a slower pace than we might imagine… yet much more profoundly.

Let's explore this timeline to avoid immediate disillusionment and anticipate the real disruptions to come!

A parallel with the Internet

In the 1970s, the Internet was a curiosity reserved for the military and researchers. Even in the 1990s, few saw this nascent network as a major societal upheaval. By the late '90s, companies sparked investor dreams, mostly based on hype… It took another decade before the rise of the tech giants. Today, we can't imagine living without it.

Artificial intelligence is experiencing a similar moment. It's on everyone's lips, in every debate, in my LinkedIn feed… but its visible effects in our daily lives remain limited, often anecdotal or concentrated in certain sectors.

AI today: Ubiquitous in discourse, discreet in Practice

Media headlines promise imminent revolutions: autonomous cars, virtual doctors, intelligent assistants. Yet, in reality, few of these promises have moved beyond experimentation or prototyping. AI is present, but its impacts remain diffuse.

It's the same on my LinkedIn feed; every day, I come across posts claiming developers are obsolete… Hmm, those making such claims probably don't understand the developer's role… or they're being disingenuous to sell a course… likely of questionable quality.

AI quietly integrates into our lives: in Netflix recommendations, automatic spell-checkers, Google translations, or voice assistants. But are these breakthroughs or merely incremental improvements?

Even as a developer, I can't say my daily life has been revolutionized…

Amara's Law: Overestimated in the short term, underestimated in the long term

American futurist Roy Amara summarized the dynamics of major innovations: "We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run."

This cycle is evident everywhere: in the history of the Internet, electricity, automobiles. All these inventions had hesitant beginnings, followed by periods of disillusionment, before radically transforming society over several decades.

The current state of AI

What's already working well

In several areas, AI performs quite effectively:

Automatic translation: A significant leap forward from the early days of Google Translate!
Personalized recommendations: Whether for videos, products, or content, AI sometimes knows our tastes better than we do.
Voice assistants: Useful for simple tasks, becoming more relevant each year.

In all these domains, AI doesn't replace humans 100% in every scenario but can meet a large portion of needs.

What's still out of reach

Authentic creativity: Generating text or images doesn't yet equate to original creation or contextual understanding.
Common sense and human judgment: AI can still be fooled by glaring errors or absurd reasoning that even a child would catch.
Reliability: Hallucinations from generative models (or algorithmic biases) pose problems; without domain knowledge, one might overlook significant errors…

Why the revolution will still take time

AI doesn't develop in a vacuum: it requires vast amounts of reliable data, massive infrastructures, regulatory frameworks, and robust industry standards. All of this is still under construction (Well, in Europe, we're ahead on regulation… not sure that's an advantage for our companies…).

We also need to transform organizations, educational systems, and even our mindsets… This takes years, even generations. Integrating AI into business processes, retraining professions, reorganizing companies around data… none of this happens with a snap of the fingers.

In 10, 20, 30 Years?

It's hard to predict precisely what AI will bring, just as it was difficult in 1990 to imagine a phone replacing our calendars, road maps (remember the hassle…), cameras (ah, film rolls… we'd find out a photo was poor quality two months later when developing it…), and libraries…

AI could one day become as commonplace as electricity: invisible, yet everywhere. We only realize electricity's ubiquity during a power outage—that's when it hurts. What will it be like for AI?

AI has experienced several "winters," periods when progress seemed to stall. We're not immune to entering another such winter for several decades… I don't believe in general AI anytime soon (not sure we'll ever achieve it, actually).

The role of Python in this transformation

If there were a language to narrate this revolution, it would undoubtedly be Python (Okay, I might be slightly biased). Simple, readable, accessible, Python has enabled thousands of engineers, researchers, students, and enthusiasts to build models, prototype quickly, and share innovations (No, I swear, I'm not selling Python courses ^^).

Just as HTML democratized web page creation in the late '90s, Python lowers the entry barrier into the world of AI. Thanks to libraries like TensorFlow, PyTorch, scikit-learn, or Transformers, it's possible to go from idea to functional model quite rapidly.

AI: An old promise already partially fulfilled?

We often forget that artificial intelligence isn't new; it's been around since the 1950s.

Since then, AI has experienced several waves of progress, often invisible to the general public:

Optimization of supply chains.
Automatic reading of handwritten documents.
Assistance in medical diagnostics.
Search engine queries.

So, it's not the first time AI has transformed entire sectors of society. It simply did so in the shadows. Today, what's changing is the declared ambition and the massive expectation of a global and rapid upheaval. AI is now targeting white-collar jobs.

Conclusion

We live in an era where technological impatience is the norm. But history shows that true revolutions take time. We must navigate between naive euphoria and sterile cynicism.

In short, time will tell 🙂