When the initial enthusiasm fades, it’s the right time to understand what truly works in generative AI initiatives. Are we really taking advantage of its potential?
That was precisely the conversation we wanted to have on Thursday, November 6, during our event “Post-Hype AI: what works, what doesn’t and what’s coming next,” held at Utopicus La Habana (Madrid).
A smaller format than in previous editions, but still among clients and friends from companies such as Iberia, Acciona, Repsol, OEPM, and Prosegur. This allowed for a closer exchange of ideas among those who are living through AI adoption in their organizations.
The conversation was moderated by Julián Gómez, Chief Digital Officer at LedaMC, and featured three outstanding speakers: María José Peral, CEO of the Instituto de Inteligencia Artificial; Manuel Rodríguez, CIO at Agroseguro; and Carlos González Jardón, Director of IT Governance and Architecture at ABANCA.
From initial enthusiasm to purposeful use
The session opened with a reflection by Dácil Castelo, CEO of LedaMC, which came up several times during the debate: “AI needs human wisdom.”
Because the key does not lie in the technology itself, but in how we combine it with critical thinking and purpose. “The question is how we use its power to create real value,” added Dácil.
From there, the panel delved into one of the greatest challenges faced by any new technology full of promises: the impulsive adoption that many companies make without a clear strategy.
“When you get carried away by the hype, you forget that things must make sense,” said Carlos González Jardón (ABANCA). “Sometimes, as technologists, we’ve lost the battle to management pressure, and we need to win it back.”
For Manuel Rodríguez (Agroseguro), the fear of being left behind, the well-known technological FOMO, is still strong: “Fortunately, we work on projects we believe will provide real value. Part of ROI is understanding the true potential of each initiative.”
From a more educational perspective, María José Peral (Instituto de IA) emphasized the need for ‘corporate re-education’: “Companies are beginning to realize that they need to evaluate the impact of each initiative before being dazzled by the WOW effects.”
Real use cases with quiet value
We continued the discussion on one of the most interesting topics: bringing visibility to real use cases, those in which generative AI is already delivering value.
María José Peral explained how her organization uses generative models to improve the learning experience of students “without losing human interaction.”
For his part, Manuel Rodríguez shared a tangible example of impact in his company: a fraud detection system, operating at Agroseguro for a couple of years now, with “phenomenal returns.”
Carlos González Jardón highlighted how, in the banking sector, “trust is our most important asset. That’s why we focus AI on internal or support processes, not on elements that could compromise the customer relationship.”
He also emphasized how user behavior can change when introducing new technologies, reminding us of the importance of observation and continuous learning when implementing new solutions. “That’s what happened when we launched a tool to review application documents: users began skipping the reading of requirements and submitting almost empty forms, expecting the AI to tell them what was missing.”
The role of human wisdom
The speakers agreed that generative AI only makes sense when combined with human judgment and critical thinking. “We must not forget that models are probabilistic. They give you the most likely answer, not necessarily the correct one,” explained Manuel Rodríguez.
“The real problem is not AI’s hallucinations but our own: when we trust it without understanding how it works or what biases it carries,” added González Jardón.
For María José Peral, the recipe is simple: education, curiosity, and continuous learning. “Rather than protecting yourself from AI, you should read, learn, and stay up to date.”
Security and data: major challenges
The conversation then moved on to other highly relevant topics in AI implementation, such as security and data reliability.
“As a bank, data security is our main concern,” said González Jardón. “We know total protection doesn’t exist, but we strive to get as close as possible.”
Peral also pointed out that, in her consulting experience, “the lack of data culture in many companies is still a major obstacle.”
Meanwhile, González summed up the consequences of this lack of data in a phrase worth remembering: “AI without good data produces bad data.”
There was also time to discuss the AI bubble and the breakneck speed of technological change. “AI is here to stay,” stated Carlos. “There will be adjustments, but it won’t disappear.”
María José also mentioned the educational sector, where “there’s definitely a training bubble, there are courses for everything. What matters is teaching how to use AI critically and responsibly.”
And Manuel Rodríguez, regarding technological change, reminded us once again that “it’s not about jumping on everything new, but about measuring value, budget, and opportunity cost.”
Towards purpose-driven AI
The event closed with remarks from Julián Gómez, who summed up the path forward: “AI has democratized technology, but now it’s time to make it truly useful, integrating it where it creates value and measuring its real impact.”
At LedaMC, we’ve been applying that philosophy for some time, integrating Generative Artificial Intelligence into software estimation, requirements improvement, test case generation, and productivity analysis within Quanter.
Our goal is clear: to help organizations achieve more with less, combining the power of AI with the experience and human judgment of our team.
A big thank you to everyone who joined us and participated with their questions and especially to María José Peral, Manuel Rodríguez, and Carlos González Jardón for sharing their knowledge and inspiring us to keep our curiosity alive.
See you next time!
