In the same weeks, two headlines appeared that seem to completely contradict each other. On the one hand, Masayoshi Son, founder of SoftBank, talks about a wave of artificial intelligence with enormous numbers and a global economic transformation. On the other hand, reports on the MIT/NANDA 2025 study warn that a large majority of business AI pilots do not show measurable results in profits.
So what are we left with? Is AI the biggest opportunity in history or an expensive fad that almost always fails?
Both things can be true at the same time.
Son talks about the future and the big picture: where technology is going, how much capital can be moved and how deep the impact on the economy could be. That vision may make sense as direction.
The data attributed to MIT measures something else: the present, company by company. It does not say that AI is useless. He says that many companies are implementing it poorly or without connecting it to business results.
The practical conclusion is simple: The potential is enormous, but today many companies waste it.
A little healthy skepticism
At Ezelero we prefer cool-headed decisions. The numbers from technology investors usually come loaded with enthusiasm. They serve to understand the direction of the market, not to promise automatic returns to an SME.
The MIT data should not be read as a condemnation of technology either. It is a warning about execution. AI works best when applied to a defined process, with available data, clear rules and measurement.
What this means for your business today
Your company doesn't need to think like SoftBank. You need to take care of your cash, your team and your customers. For any business, the question is not how much AI will move in 2040. The question is much more specific:
- Where do I waste time today?
- What questions remain unanswered?
- What quotes cool down?
- What report is prepared manually each week?
- What process depends too much on one person?
The group that does win is not the one that invests the most
The group that gets results with AI usually has something in common: they don't start by buying technology. Start by choosing a measurable problem. Then design the flow, test small and adjust.
How we see it in Ezelero
We do not sell the euphoria of the billions or the pessimism of the headline. We sell something simpler: artificial intelligence applied with criteria and measured in results.
That means starting with your real problem, choosing the right tool, integrating it with your processes and demonstrating with numbers what it saved, what it accelerated or what it helped recover.
AI can transform the economy, but your company wins when it transforms a specific process today.
Let's start with the concrete
If you want to know where AI can leave a clear return and where it is still not worth it, the right thing to do is start with a diagnosis. Without smoke, without promising miracles and without buying a platform before understanding the process.
Ezelero editorial based on public statements attributed to Masayoshi Son/SoftBank and reports on MIT/NANDA 2025. Figures are used to explain the tension between global potential and actual business execution.
Frequently asked questions
Is AI an opportunity or an expensive fad?
It can be both. It is a great global opportunity, but it becomes expensive when a company implements it without process, without integration and without metrics.
What does the 95% figure mean for an SME?
It means that it is not advisable to buy AI for fashion. It is advisable to start with a small pilot, connected to a real problem and measured with simple indicators.
How can a company be in the group that does win with AI?
Choosing a specific process, using the correct tool, maintaining human control and measuring from day one.
Do you want to apply AI with numbers, not smoke?
We review your current process, choose a small pilot and define how to measure if it really suits your company.