Categories:
Interview published by Computer World
Sarah Harmon arrived in Spain from the United States thinking it would be something temporary. She has been here for 30 years, leading teams in tech giants such as LinkedIn, Sngular, or Microsoft, in addition to being a board member at CaixaBank Payments & Consumer. For exactly one year now, she has been managing the Spanish edtech Odilo, a learning platform with a potential network of 170 million users on five continents and a verified resource catalog of more than 7,300 providers, built over more than a decade alongside publishers, experts, and specialized magazines. Although originally, this startup created in Cartagena by a group of engineers sought to apply the concept of the cloud to libraries, the company has evolved to become a global educational ecosystem.
The executive has been in the sector since 1990. She has seen the rise of the internet, mobility, and the cloud. All these waves brought the same promise of transformation. This time, she says, is different. Not because artificial intelligence is more powerful, but because for the first time it absorbs the employee’s knowledge. What used to require a specialist is now executed by a machine. That changes the question: it is no longer about how to use the tools, but what to learn when the tools do more and more.
If she had to give a grade to the current impact of AI on professional skills, she would give it a five out of ten. “We tend to overestimate the immediate impact of any innovation,” she explains. That five is not reassuring. “We are only using a small part of what these tools can do,” she says.
What is happening in fields like technology will eventually spread to the rest of the sectors. Developers are seeing it first. AI is already writing code, debugging errors, and generating solutions that previously required years of specialization. What today is an aid tomorrow may be a substitute. In healthcare or banking, that moment will arrive later, slowed down by regulation, privacy, and cultural resistance. But it will arrive. The question is not if it will happen, but when.
Most organizations, for now, are using it to do the same thing faster. They automate tasks, reduce times, and gain productivity. But that doesn’t change anything fundamental. “We are not approaching it as transformation. We are approaching it as improvement, as efficiency.” The difference is profound. An efficient organization does the same thing better. A transformed organization changes what it does and why it does it. This goes beyond skills. It is also a question of method. Of how an organization learns and with what discipline. “The logic is no longer accumulating knowledge for years and becomes how to constantly update oneself,” Harmon states.
Sarah Harmon has been in the sector since 1990. She has seen the rise of the internet, mobility, and the cloud, waves that brought the same promise of transformation. But with AI, she tells us, it’s different, not because of its power, but because for the first time it absorbs the employee’s knowledge.
The clearest sign that a continuous learning culture does not exist is usage. Mandatory courses work. Afterwards, participation plummets. “Employees log in when required and disappear as soon as they can. It is not a platform problem, but a cultural one,” Harmon maintains. Corporate training does not only compete with other learning platforms. It competes with YouTube, with ChatGPT, with Instagram. And it almost always loses. The reason is that most are designed for whoever buys them, not for whoever uses them to learn.
It’s not that corporate training is failing. It’s that it was never designed to transform anything. Organizations that confuse efficiency with transformation are not only wasting time; they are accumulating a debt that AI will eventually collect. Harmon advocates continuous learning as part of the job, but rejects long-term planning. “The world changes too fast. Business changes too fast. Skills change too fast.”
En un mundo donde todo cambia demasiado rápido para planificar, la única respuesta posible es cultural. “El aprendizaje continuo no nace de un plan. Es cultura”, dice Harmon.
In a world where everything changes too fast to plan, the only possible answer is cultural. “Continuous learning does not stem from a plan. It is culture,” says Harmon.
It is not about adding more courses, but assuming that learning is not something that happens outside of work, but within it. That is why she insists on integrating learning into employee evaluations. “There should be two or three learning objectives in the performance evaluation. For the employee to understand that learning is part of their job.” As long as that doesn’t happen, training will remain an accessory, understood more as an obligation than an opportunity.
Changing that is the only real answer to the question of how not to get left behind. Once the employee internalizes that learning is part of their job, the focus shifts to how and what is learned. The starting point is self-knowledge: knowing what you do well, what you want to learn, and where that intersects with what the market demands. Hence her insistence on the figure of the sherpa: a system capable of guiding and adapting learning to the real context of each person, not to that of the platform. Content quality matters as much as guidance. In contrast to platforms that learn from unverified public information, Odilo has spent fourteen years building a catalog with publishers, renowned experts, and specialized magazines. Content validated before reaching the user.
As for the what, Harmon insists on focusing on what technology cannot replicate. She rejects the term soft skills and prefers to talk about durable skills: communication, critical thinking, negotiation, creativity, or curiosity. “Skills are what will differentiate us. AI can execute. It cannot replace judgment.”
“Continuous learning does not stem from a plan. It is culture.”
But knowing is not enough. The employee needs to practice those skills within their own organization before they count. Digital learning prepares the ground. Real practice consolidates it.
But the hardest question is not what employees learn. It is what managers do with the time that AI promises to free up. If that promise is fulfilled, between 30 and 40% of the work dedicated today to low-value tasks will disappear in the coming months, according to estimates circulating among consulting firms. The question is not how to fill that space, but what to do with it.
For years, middle managers have accumulated more tasks, more control, and more responsibility. Now they can dedicate that time to other things. But it is not clear that organizations are prepared for it. The optimistic scenario is that this time is used to lead in a different way: more coaching, more relationship building, and more time for strategy. The less optimistic one is that the roles that AI empties simply disappear. “I don’t know if companies are thinking about how to relocate those people,” Harmon acknowledges.
At Odilo, they have decided that to lead this change, they must start with themselves. They have halted part of their activity to train all employees in AI, with mandatory sessions that are part of the job. “We want to set that example,” says Harmon. That bet has an additional dimension. 80% of their clients have Spanish as their primary language. In a market where large global platforms seek to scale in English, Odilo has built over years a knowledge of Spanish-speaking educational systems and markets that is not easily replicated.
The learning platform may be adapted and the company may tell you what it needs. But, as Harmon concludes, “the responsibility of keeping yourself relevant must be yours.”