What past industrial revolutions teach your SME about AI
Every major tech revolution has followed the same script: hype, excess, a shakeout, maturity. Artificial intelligence is no exception. Understanding this pattern is the best way to decide what to do today.
Every major technological revolution has followed the same script: hype, excess, a shakeout, and finally those left standing change their market forever. Artificial intelligence is no exception. Understanding this pattern is the best way to decide what to do today in your SME, without being swept away by either fear or illusion.
It's talked about everywhere, and almost always unhelpfully. On one side, people promise that AI will solve every problem in your company by tomorrow. On the other, that it will wipe out work and steamroll anyone who doesn't adapt. For an entrepreneur who has to make the numbers add up every day, neither voice helps you decide.
There is, however, a more solid way to get your bearings: look at what actually happened in previous industrial revolutions. Because patterns repeat, and history tells us with surprising clarity who gained, who lost and why.
The pattern that has repeated for two centuries
Take the three great industrial revolutions: steam and mechanised textiles in the early 1800s, electricity and chemistry in the late 1800s, computing and the internet in the late 1900s. Three very different technologies, yet all three followed the same sequence.
First comes the hype: the new technology looks magical, everyone wants it, investment explodes. Then comes the excess: too much gets built, money goes into promises rather than real results, expectations run faster than the technology can deliver. At that point comes the shakeout: the bill comes due, many of those who bet everything on the wave of euphoria fail, the market reprices abruptly. And finally comes the phase that really matters, maturity: the technology, now stripped of illusion, becomes the foundation on which the real growth of the next decade is built.
The most instructive example is the British railway mania of the 1800s. Enormous capital poured into building railway lines, many companies were never profitable, and when the bubble burst thousands of investors were ruined. Yet the rails remained. And the English economy ran on them for fifty years. Those who had speculated lost everything; those who knew how to use the surviving infrastructure built fortunes.
In every technological revolution, betting on euphoria is one thing; using the technology to concretely improve your own work is another. The first is dangerous. The second is what separates those who thrive from those who fall behind.
Where we are with artificial intelligence
Today we are in the hype phase sliding toward excess. Investment in AI has reached unprecedented figures, the valuations of many companies in the sector are very high, and experts openly debate whether or not we are in a bubble.
For your SME, the question "will the bubble burst?" matters less than it seems. Even if there is a shakeout among the big tech companies, as happened with the internet in the early 2000s, the underlying technology will remain and keep spreading, exactly as the internet survived its own crash and then changed every sector. What matters for you is not guessing the moment of the stock-market repricing. It's deciding which side of the transformation you want to be on.
The part that really concerns your business: costs
Here we get to the concrete point. Behind every technological revolution there's an economic mechanism that touches your bottom line directly: the new technology lowers production costs, and whoever adopts it first gains an advantage over those who lag behind.
It happened with textile mechanisation, with the electrification of factories, with business software. And it's happening now with AI-based automation. Repetitive activities, such as document management, data control, reporting and part of administration and customer service, can today be automated at an ever lower cost.
This means something uncomfortable but important to say clearly: your competitors who automate their processes are lowering their costs. And if you stay on manual processes, every quarter that passes you lose a little margin compared with them. Not because they're better, but because they're using a tool you're not using yet.
It's not an abstract threat about the "future of work". It's a question of margin, here and now.
And people's jobs?
It's the question everyone asks, and it deserves an honest answer. The history of the industrial revolutions tells something more nuanced than the alarmist headline: automation has always eliminated certain tasks and, in the long run, created others, often more skilled. AI does not replace the entrepreneur or their judgement: it replaces the repetitive work that steals your hours today, freeing time and people for the activities that truly require a human mind.
What changes, more than the numbers, is what you need to know how to do, and therefore who you hire and how you grow the team. It's a topic that deserves its own space, and we've devoted another article to it: AI won't fire you, but it changes who you hire. Here, the essential point is enough: those who adopt AI to shed low-value work and focus on high-value work come out stronger.
The real watershed: data alone isn't enough
There's a widespread misconception to clear up. Having AI tools, or piling up mountains of data, is worth nothing in itself. The value isn't in the tool and isn't in the raw data. It's in the interpretation: turning the numbers into concrete decisions for your business.
A dashboard full of charts that no one can read is useless. Automation that produces reports no one uses to decide is a waste. Technology counts only when someone connects it to the real problems of the business: where am I losing margin, which customer is truly profitable, which process is costing me more than it returns.
This is the thread that ties all the industrial revolutions together: technology is the context, but it's always the human hand, the one who knows how to use, interpret and direct it, that decides who gains. It holds for steam, for electricity, for AI.
What you can do concretely, starting now
You don't need a big, costly project or an internal revolution. History rewards those who start early and in a focused way, not those who wait for the perfect solution.
The principle is simple: choose a single process to start from, the one that loses you the most hours or money, automate it well, and measure the result in concrete terms: hours saved, errors reduced, margin recovered. Automating one and seeing it pay for itself in a few months is worth more than ten projects started and never finished. If you want a method for choosing where to begin, we covered it in detail in Which processes to automate first in an SME.
And above all, keep the focus on the essential: the goal isn't to "use AI" for fashion. It's to free you and your team from low-value work, to devote your energy to what really grows the business.
In short
The industrial revolutions of the past tell us that technology, sooner or later, redraws every market, and that the winners aren't those who chase euphoria, but those who use it methodically to improve their own work. Artificial intelligence follows the same script. For an SME the stakes aren't philosophical, they're about the books: whoever automates low-value processes lowers costs and gains margin, whoever stands still loses it little by little.
The good news is that you don't need to be a tech giant to be on the right side. You need to start early, in a focused way, and always keep at the centre the thing history has confirmed for two centuries: technology counts only when a human mind turns it into concrete decisions.
Technology is the context, but who gains is always decided by whoever can turn it into concrete decisions. AFianco supports small and medium-sized businesses in Switzerland and Italy with digitalisation and process automation, starting from the company's real numbers. No hype: what actually works, explained by the people who implement it.
Frequently asked questions
Are we in an AI bubble?
We're probably in the hype phase sliding toward excess, with very high valuations. But for an SME it matters little to guess the moment of the stock-market repricing: like the internet after 2000, the underlying technology stays and spreads.
Is AI really worth it for a small business?
Yes, for a very concrete reason: AI-based automation lowers the cost of repetitive activities. Those who use it gain margin, those who stay on manual processes lose it bit by bit compared with competitors.
Will AI cost jobs in my company?
Automation eliminates certain repetitive tasks but, historically, creates others that are more skilled. AI doesn't replace the entrepreneur's judgement: it frees time from low-value work for what needs a human mind.
Where should an SME start with AI?
From a single process, the one that loses you the most hours or money. You automate it well, measure hours and margin recovered, then extend. Better one process that pays for itself than ten projects never finished.
Are data and AI tools enough to get results?
No. Raw data and tools alone don't produce value. What counts is interpretation: turning the numbers into concrete decisions, connecting the technology to the real problems of the business.