AI, AI! Okay: it's the big topic of the moment. But the enthusiasm and/or panic may be getting out of hand. That's why, among the daily flood of emails on the subject, one with the suggestive subject caught my attention: "Marc Torrens demystifies the real potential of artificial intelligence for business leaders." That's it: the problem/danger lies in mythifying it. Torrens, a Data, Analytics, Technology, and AI professor at Esade, gave a lecture on how AI really works, what its opportunities are, and what challenges companies must address to leverage it. Rigor, detail, and analysis. Perfect. Very necessary. But I wanted to delve into the problem, portray it well, more than the solution. So I asked him directly if business management is getting too carried away with AI. "It is necessary to lower the volume and tone down the hype. AI is not a magic wand, but an infrastructure, a tool. Like with the arrival of electricity or the internet, the real impact does not come from using it because it's 'trendy,' but from redesigning work thanks to it. The current obsession is a symptom of strategic myopia: the focus is on the tool and not the problem to be solved." Clear as water. Essentially, it's laziness when the most challenging part is missing: "In technology, it has been shown many times that developing 80% of the solution requires 20% of the total effort to make it truly useful and secure. The difficulty lies in the last mile." Similar to self-driving cars, "currently autonomous in highly digitized environments and under optimal conditions: extending this solution to any environment and circumstance requires many more technological advances." And when we get there, let's see what we find. "Generative AI is a disruptive technology, but we still have to see to what extent it will change businesses, and how. In general, we have a bias heavily influenced by Silicon Valley's marketing efforts with a clear corporate objective: to secure funding. The scientific world critical of this technology does not have the media platform that those large corporations have." Ander Serrano, Innovation Manager at Evercom, predicts that "the transformation is irreversible," but admits that "the narrative tends to precede reality. Technology evolves faster than organizations, especially the larger ones with more structure to mobilize. Culture, processes, and value proposition depend on changing human habits." That's why we are at a crucial moment, where "the challenge lies in moving from exploration to obtaining significant advantages, whether in terms of business returns or service improvements. The most agile companies are already taking advantage of much of what AI can do for them, but this is just the beginning." Serrano points out "obvious risks: legal, training/education, or technical, based on implementation and connection with reliable data." But he agrees with Torrens that "there is a more important problem that is talked about much less, focusing on the tool and forgetting the purpose. The real challenge of AI is not technical, but in the purpose: Why do we want it and how far are we willing to go? Is it for our team to better reconcile? For the customer to receive faster service? To offer something new?" Isaac Cantalejo, Director of Netmind Managing Director and Vice President of BTS, is critical of our approach: "He is right to point out that there is an inflation of expectations around AI and that it is advisable to lower collective anxiety, but falls short if we reduce it to 'AI is not all that.' It is a structural technology and will have a profound and lasting impact." Having said that, he acknowledges that there is a problem, but "it is not its importance, but the place it is being assigned in organizations. In many of them, it has become an almost ideological object. There is an obsession with 'not falling behind' that leads to hasty decisions," because AI is used "as a mental shortcut to solve problems that are actually related to organizational design, leadership, and ways of working." Serrano has even heard talk of "a trend towards AI Washing," which consists of "calling something AI that is not really." Cantalejo does not like the cliché of "bursting a supposed bubble. What we need is to reposition AI: realistically assume what it can contribute and what it cannot, and stop projecting expectations onto it that correspond to human and organizational decisions."
The great hangover towards the "Trough of Disillusionment"
Marc Torrens argues that, regarding AI consumption, we are in the phase of the "enthusiasm hangover." The closest thing to a breathalyzer (the metaphor is mine) in these matters would be "Gartner's famous Hype Cycle." By blowing into generative AI, we see that it is descending from the 'Peak of Inflated Expectations' towards the 'Trough of Disillusionment' due to the clash with the operational reality of companies. So maybe we think that some Sundays it's better not to get up.