From dotcoms to artificial intelligence. Find out how to turn the AI bubble into a window of opportunity (and avoid being part of the 95% who get no return). (and avoid being part of the 95% who don't get a return).
In the late 1990s, the Internet was the new gold rush. Every ".com" company seemed destined to change the world, and investors competed to jump on the digital bandwagon. It was enough to have a website to attract capital. The Nasdaq index, the thermometer of technology companies, grew almost sixfold between 1995 and March 2000. It seemed that the growth would be infinite... until it was no longer so.
In the following two years, the Nasdaq lost 78% of its value. Thousands of companies disappeared, millions of investors evaporated and euphoria turned to skepticism.
But the essentials survived. From that collapse emerged fiber optics, modern communication protocols and digital habits that would transform the way we live. On top of those ruins, the Internet was consolidated as the most powerful infrastructure of the modern era.
From technology to culture
The difference between those who disappeared and those who prospered was not the technology, but the reading of the change. While some saw the Internet as a technical project ("having a website", "selling online", "appearing in search engines"), others saw a cultural transformation: a new way of searching, working, buying and communicating .
- Google did not invent search engines: it understood that knowledge would be the new oil.
- Amazon didn't invent commerce: it understood that trust in online shopping would be the new store.
- Microsoft did not invent the cloud: it understood that work would no longer take place within four walls.
In all these cases, the success was in betting on human behavior and not on the technology itself. That nuance changed the course of the digital economy.
The new bubble: artificial intelligence
Twenty-five years later, history is being rewritten, only with a different protagonist: artificial intelligence. Once again, investments are soaring, headlines promise daily revolutions and funds are looking for their next unicorn. Thousands of companies compete to launch AI pilots, yet most lack real impact or alignment with the business.
The parallel with the year 2000 is obvious: back then, many companies viewed the Internet from a technical angle; today, many do the same with AI. And we predict a similar outcome: a bubble that will not destroy value, but will lead to natural selection.
The mirage of integrations
Numerous companies have built their business on layers of third-party software, mainly APIs from large manufacturers. Dozens of startups that based their business on these integrations will lose their competitive advantage and simply... disappear. The recent appearance of OpenAI Agent Builder, a platform that allows you to create intelligent agents visually, integrate the Model Context Protocol (MCP) and automate processes without programming, makes it clear: when the manufacturer replaces you, your company ceases to make sense.
Tools like this are beginning to occupy the functional space of platforms such as n8n, Make or Zapier, reducing the competitive advantage of those who only offered integrations. The message is clear: if your proposal depends on a technological intermediary, no one can guarantee your sustainability.
The microproductivity trap
This problem can be transferred to the user companies, which have deployed a multitude of disconnected AI pilots , with no strategy or governance.
Chatbots that no one uses, dashboards that no one consults, automations that no one maintains... The initial enthusiasm has translated into micro-productivity without direction.
According to a recent MIT study, 95% of companies that have implemented AI have not obtained measurable economic return. The main cause is not technical, but strategic: lack of alignment with business objectives.
As with the Internet, the current AI "bubble" will not destroy value, it will redistribute it.
Those who only integrate technology will disappear; those who build competitive advantage on data, governance and vision will survive . The companies that understand this will be the Google, Amazon or Microsoft of the new cycle. The rest will feed the graph of a bubble.
From fashion to management
As was the case a few years ago, the real disruption is not in "using AI," but in rethinking how decisions are made, how processes are executed and how value is created with intelligence. A sound AI strategy does not start with choosing a model, but with understanding where it can generate sustainable impact.
This strategy must:
- Identify critical processes where AI can create savings or competitive advantage.
- Design the data and governance infrastructure to scale with control.
- Integrate AI capabilities (predictive, generative, cognitive) into real business flows.
- Measure impact and adoption with financial rigor and continuity.
When AI is integrated as a cross-cutting layer of decision and efficiency, and not as an isolated experiment, it ceases to be a fad and becomes a direction.
Where to start
An AI strategy does not start with technology, but with the right questions :
- Which areas of my business are bottlenecks or margin levers?
- What data do I have and what data do I need to automate meaningfully?
- What type of AI (predictive, generative, symbolic, etc.) can provide real value?
- How do I ensure governance, scalability and internal adoption?
Answering these questions requires methodology, transversal vision and knowledge of the business. To obtain a roadmap that combines immediate results with long-term structure.
Because, at the end of the day, it's not about having more AI, it's about having it well thought out. And that's where the work begins that very few are doing. Precisely for this reason, the window of opportunity has yet to be seized.
