# Thinking through AI: from uncritical adoption to organisational and educational transformation Author: Carlos López Ariztegui Original publication: 2026-02-25 Spanish original: https://estrategiabyaleph.substack.com/p/estrategia-126-pensar-la-ia-de-la English URL: https://elcontemplador.github.io/estrategia-english/essays/126/ Status: Published translation This is a translation of the original Spanish essay published on 25 February 2026. Its claims, examples and forecasts retain that historical context. English publication: 2026-09-29 > This week, estrategIA welcomes **[Carlos López Ariztegui](https://www.linkedin.com/in/clariztegui/)**, Dean of ESIC Business & Marketing School. > > With almost four decades at the technological frontier in organisations such as Apple, Nokia and AT&T Bell Labs, [Carlos](https://substack.com/@clariztegui) has consistently combined innovation with critical thinking. In his academic leadership at ESIC—a key strategic partner in several of ALEPH Educational Institution's major programmes—he now promotes a humanistic view of digital transformation that goes far beyond adopting tools. > > This week's article poses an uncomfortable but necessary challenge: stop asking what AI can do for our organisations, and start asking whether we have the judgement, processes and culture needed to guide it purposefully and turn it into real value. Today's conversation about artificial intelligence is dominated by an obvious paradox: never have so many organisations claimed to use AI, yet few manage to turn that adoption into real value. Investment grows, pilots proliferate and tools become popular, but the impact on results remains limited. Technology advances; the organisation stays where it is. This gap is not a technical failure. It reflects a mistaken approach. Many companies see AI as a capability to buy, rather than a system that requires them to rethink how they work. The consequence is predictable: new tools are introduced into old structures, in the expectation of a change that never arrives. A familiar phenomenon emerges. Fear of missing out pushes organisations to adopt solutions before understanding them. Competitive pressure, amplified by suppliers and media narratives, creates a race to attach “AI” labels to processes. The result is a flood of almost indistinguishable applications: polished interfaces connected to the same model, applied to problems that needed operational clarity rather than probabilistic methods. The organisation is not transformed; it acquires a new coating. It automates irrelevant reports, accelerates poorly framed decisions and multiplies existing mistakes. The technology works, but the system into which it is introduced is unprepared to absorb it. The problem is not using AI, but using it without first defining why the organisation exists and how it creates value. The same pattern appears across the economy. The sector's growth is driven more by expectations of future productivity than by present returns. Massive infrastructure investment, concentration among a few players and dependence on still-immature demand suggest that technological enthusiasm is preceding organisational maturity. Again, this is not a technological crisis; it is a crisis of interpretation. Thinking through AI means interrupting that momentum. Start with questions, not the tool. Which decision do we want to improve? Which process needs to change? What knowledge is missing? Which error are we amplifying? Without that framework, any pilot can look successful while providing no competitive advantage at all. AI does not replace organisational thinking; it exposes it. Introduced into an environment of inconsistent data, fragmented processes or ambiguous objectives, it does not remedy those shortcomings. It reveals them at scale. That is why many projects fail without any actual technical malfunction. The model works exactly as intended; the organisation does not know what to ask of it. This also requires redefining the expert. For decades, professional value was associated with having answers. Today, any system can produce them in seconds. The distinctively human contribution shifts elsewhere: asking relevant questions, interpreting ambiguity and taking responsibility for decisions under uncertainty. Expertise ceases to mean accumulating information and becomes the capacity for judgement. Knowledge is no longer a static store, but an interactive process between people and systems. Professionals do not compete with machines on speed or memory. They complement them with judgement, context and purpose. The capabilities gaining value are therefore precisely those least amenable to automation: interpretation, creativity, ethical evaluation and systemic understanding. AI does not eliminate these functions; it makes them more important. The more accessible answers become, the more it matters to know which problem is worth solving. This shift directly affects higher education. The challenge is not to teach students how to use particular tools, but how to work with an external cognitive system. Three transformations are needed. First, teaching staff must learn before their students do. Introducing classroom software is not enough. We need to design experiences in which AI supports reasoning rather than substitutes for intellectual effort. Second, integration must extend across disciplines. AI is not a subject; it is a condition of the environment. Marketing, finance and operations must rethink their questions in the light of a new kind of available information. Third, learning must include friction. Deep understanding comes from confronting open-ended problems where the tool does not guarantee a solution. Students need to experience the cycle of building, making mistakes and revising in order to develop their own judgement. Education cannot merely produce efficient users of automated systems. It must develop professionals capable of directing them. That means cultivating critical thinking, responsibility and interpretive skills, because AI can scale mistakes as readily as successes. The implication is clear: once AI becomes ubiquitous, it ceases to confer a competitive advantage. The difference will lie in understanding it, not possessing it. The organisations that prosper will not be those adopting the most technology, but those that have redefined their processes, decisions and culture to work with it. AI is not a shortcut. It is a multiplier. It amplifies clarity or it amplifies confusion. The real challenge, therefore, is not mastering the tool but mastering how we think with it. Only then does technological adoption become transformation. **[Carlos López Ariztegui](https://www.linkedin.com/in/clariztegui/)** *Dean, ESIC Business & Marketing School* [![Infographic mapping uncritical AI adoption to organisational judgement and educational transformation.](https://elcontemplador.github.io/estrategia-english/assets/images/7f82919f-96f8-4c06-a3a7-35c7ad7b0959_572x1024.png)](https://substackcdn.com/image/fetch/$s_!-_4f!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F7f82919f-96f8-4c06-a3a7-35c7ad7b0959_572x1024.png) *English reading of the original infographic, “Thinking through AI: from uncritical adoption to organisational and educational transformation”, credited to Carlos López Ariztegui:* - **Today's paradox: adoption versus impact.** Mass adoption is a “gold rush”: record investment, proliferating tools and pilots, and fear of missing out. Real impact remains limited as organisations retain old structures. Technology advances; the organisation stands still. The mismatch comes from a mistaken approach. - **The mistake: coating, not transforming.** Capabilities are bought without rethinking work; “AI” becomes a label. Irrelevant reports, poorly framed decisions and previous errors are automated or accelerated. The system is unprepared. This is a crisis of interpretation, not technology. - **Thinking through AI: interrupt the momentum.** Start with questions: which decision should improve, which process should change, what knowledge is missing, and which error are we amplifying? AI acts as a mirror to inconsistent data and fragmented processes, exposing rather than fixing deficiencies. A small label beside the brain graphic is malformed in the original and has not been reconstructed. - **Redefine the expert and human capabilities.** Move from “having answers” to “asking questions”, and prioritise judgement over speed. The graphic labels interpretation, creativity, ethical evaluation and systemic understanding as key, non-automatable capabilities. AI complements rather than replaces these; people provide context and purpose. - **Transform higher education.** Teachers learn first and design experiences that support reasoning; integration is cross-disciplinary because AI is a condition of the environment, not an isolated subject; learning includes friction through building, making mistakes and revising. Develop people who can direct systems, not merely use them efficiently. - **The future: understanding as an advantage.** Ubiquitous AI no longer provides a competitive advantage by itself. Understanding and redefinition lead to thriving organisations with redesigned processes and cultures. The closing message repeats the article: AI is a multiplier of clarity or confusion; the challenge is mastering how we think with it.