PT
EN
Some technologies mark inflection points in history. The most recent examples were ERPs and dot-coms. In the field of methodologies, we can cite Scrum.
These practices are driven by commercial investments — especially from major consultancies. But not before they start gaining ground in companies organically.
At the height of the hype, it becomes almost impossible for an executive not to adopt the trending technologies. At the risk of damaging their own career.
I remember the dot-com era, when everyone wanted to build a website without quite knowing why. Many executives left positions at large companies to dedicate themselves to “dot-coms” — especially in retail.
Academia Follows the Birds, Not the Other Way Around
From an academic standpoint, when a topic gains critical mass of practical adoption, it becomes the subject of research. Curiously, in popular culture, academia is seen as an unquestionable source of knowledge. The popular logic is: “a scientific article states that birds fly.”
In reality, it works the other way around: first the birds fly, and only then does the phenomenon become a research subject. Science prefers to investigate the points where empirical knowledge gets stuck — like certain types of birds that, without the anatomy to fly, fly anyway.
The Classic Cycle of High-Potential Technologies
For technologies like AI, the massive market attention turns them into panaceas — promises to solve all of a company’s problems. That is where academia steps in and tries to answer four fundamental questions:
- Seminal research: do we have a critical mass of real cases to study quantitatively and qualitatively?
- Sociometric research: do we have a volume of papers generated on the topic that allows us to identify patterns?
- Identification of falsifications: are there pattern breaks?
- The productivity paradox: under what conditions does the technology in question bring competitive advantage?
In the last stage, the real value of the technology for competitive results becomes clear — and, most importantly: under what conditions that value is actually captured.
What Is the Productivity Paradox?
The productivity paradox refers to the phenomenon in which highly promising new technologies are widely adopted but do not immediately result in the expected productivity gains.
This concept was popularized by economist Robert Solow, who observed that despite advances in computing during the 1970s and 1980s, the impacts on company productivity were not evident in economic data.
The paradox occurs because, although the technology has potential, its full efficiency depends on factors such as:
- Adaptation time: companies need to change processes and train people.
- Complementary investments: technology alone is not enough; integration with other systems and methodologies is required.
- Side effects and inefficiencies: in the short term, adoption can create disorganization or increased costs before benefits are perceived.
Will AI Follow the Same Path?
I believe so. In reasoning and content generation, AI is impressive. In image generation, it is still quite rudimentary — although improving.
In AI’s case, we are possibly watching this cycle repeat: initial hype, mass adoption, implementation challenges, and only afterward, real productivity gains under specific conditions. One of the biggest obstacles on this path is silent and overlooked: the quality of the data feeding these systems — a topic explored in The Silent Error That Sabotages AI Projects.
The interesting part is that in Brazil, older methodologies and technologies tend to be adopted after some time delay. It will probably take a while for the full cycle to materialize here. Only when they are old enough to decide to live in a tropical country.
Meanwhile, my suggestion to the sugarcane juice vendor: advertise “sugarcane juice with AI.” And if you want to understand what truly determines whether technology generates value or just cost, it is worth reading Technology, Information and Value.