<p>In a changing world, where new artificial intelligence tools arrive every day, the promise of speed and productivity seems to be the new frontier for CTOs and IT departments.</p>
<p>The mantra resonating in managers’ heads these days is:</p>
<pre><code class="language-text">we need to be faster
</code></pre>
<p>“VibeCoding,” the promise that anyone can build software without knowing how to program, is the flagship tool of this speed, and coding agents are its vassals: a platoon of virtual professionals generating code and technical debt at impressive speeds.</p>
<p>However, this acceleration might hide an insidious danger: the creation of a generation of “virtual seniors”—professionals who, while capable of delivering products rapidly thanks to AI, lack the deep skills necessary to tackle the most complex technical challenges.</p>
<h2>A Disservice or a Mockery?</h2>
<p>If until today professional growth—from junior to senior—was a long path made of errors, corrections, and manual learning, AI risks skipping this crucial phase, giving everyone tools that allow them to bypass “paying their dues” and become hyper-productive immediately.</p>
<p>A perverse mechanism disguised as productivity, which could turn into a weapon pointed at the very company demanding speed, creating “apparent professionals”: people capable of achieving excellent results immediately, but then unable to adapt what is required to complex business needs.</p>
<p>The question we should ask ourselves is: how can CTOs avoid this drift? How can we counter the process of de-skilling currently underway?</p>
<h2>The Competence Collapse</h2>
<p>There is great enthusiasm for how AI accelerates everything, but this acceleration does not allow for one very important thing: error.</p>
<p>One error teaches more than a thousand lines generated by AI. When we delegate everything to an agent for too long, we risk losing the ability to learn, because there is no material time to assimilate, study, make mistakes, and learn.</p>
<p>The data speaks clearly: according to the Stack Overflow Developer Survey 2025, 84% of developers are using or plan to use AI tools, with 51% using them daily. However, trust has plummeted: only 33% trust the accuracy of AI output, while 46% actively distrust it. The main frustration? 66% complain that “AI solutions are mostly correct, but not entirely,” and 45% find that debugging AI-generated code takes more time than writing it from scratch.</p>
<p>If we think about the next 5 years, the risk is that skills will no longer be formed because AI will do the work for us. But at that point: who will evaluate the quality of the result? Who will understand if that product is suitable for the corporate context?</p>
<p>As long as we have people raised with the old learning mindset, perhaps we will manage, but in 5 years? And in 10?</p>
<p>We must think about how to prevent tech companies from finding themselves with a generation of hollow seniors, incapable of managing complex systems.</p>
<h2>Digital Outsourcing</h2>
<p>The direction we are heading is digital outsourcing: we are selling off culture in exchange for an appearance of speed.</p>
<p>Are we ready to delegate our know-how to software whose workings we do not know? The moment that software makes a mistake, how do we understand where it went wrong if we do not have the skills to do so?</p>
<p>Let’s reflect on this:</p>
<pre><code class="language-text">Startups can launch with AI-generated code, but they cannot scale without expert developers
</code></pre>
<p>If we need to create a POC: AI is fine; if we need to build a solid company with robust foundations, we cannot rely on those who do not have the skills to understand what they are doing.</p>
<h2>Proactive Strategies for CTOs: Let’s Create Antibodies</h2>
<p>The solution is not to ban AI, but to rethink the growth path of our developers. We must design positive frictions in the workflow that force the brain not to switch off.</p>
<h3>1. AI-Free Zones</h3>
<p>Once, humans had to run after animals to hunt; now, to stay fit, they spend hours on a treadmill. We need to define boundaries where the use of AI is deliberately forbidden.</p>
<p>Designate specific work areas, perhaps in critical contexts like legacy systems, as cognitive gyms. In these zones, AI use is forbidden or limited to consulting documentation. This forces one to read, understand, and describe logic, creating that muscle memory that would otherwise be lost.</p>
<h3>2. Rethinking Code Reviews: Ask “Why?”</h3>
<p>Any software change must not be evaluated only on whether it works: the new question must be “Why?”.</p>
<p>Why did you choose this specific pattern instead of another? When the answer is “AI suggested it,” the change will be rejected. We must evaluate the ability to defend architectural choices, not just the ability to produce output.</p>
<p>I remember a university course where I taught: after evaluating assignments, I always asked the reason for the choices made. Sometimes students answered “Because ChatGPT suggested it.” In that moment, I knew the work was produced by a machine, and I insisted that it be rewritten, analyzed, and understood.</p>
<p>The ability to argue one’s choices will be even more important in a world where AI can suggest a thousand different alternatives in seconds.</p>
<h3>3. Let’s Play Wargames</h3>
<p>AI is great at diagnosing common errors, but what happens when the system collapses in a way the model has never seen?</p>
<p>Organize sessions where parts of the staging environment are intentionally broken in subtle ways (network problems, race conditions, memory leaks) and ask the team to solve the problem without AI tools. The ability to trace an error through the “mental stack” distinguishes a professional from a simple prompt engineer.</p>
<h3>4. Digital Archeology</h3>
<p>Assign tasks in contexts where AI has no information, so that the answers are largely wrong and people are forced to study better.</p>
<p>An example is managing legacy code on proprietary platforms that are poorly documented. This forces people to study how the old programmer thought—an exercise in technical empathy fundamental for designing robust systems.</p>
<h3>5. Let’s Talk About Design First and Then Prompts</h3>
<p>Let’s force people to first write a design document, a high-level architecture that describes the overall vision.</p>
<p>Thought must precede the generation of the result. If AI writes the code, the human must write the architecture. If the human abdicates this too, they become useless and probably harmful.</p>
<h2>Recruiting 2.0: Hiring for Cognitive Capacity</h2>
<p>Old algorithmic tests are now useless: any AI model solves them in seconds. What should we look for?</p>
<p><strong>The ability to ask questions</strong>: my 7-year-old son bombards me with questions, but from how he asks them, I understand what he is learning and in what direction he is going. Do the same with candidates: describe a vague problem. Do not evaluate the solution, but the questions they ask to clarify requirements. AI is terrible at handling human ambiguity—it doesn’t clarify, it just follows the most logical path. For sci-fi fans: it’s like talking to a Vulcan.</p>
<p><strong>Debug broken code</strong>: give candidates a complex system that fails intermittently. Observe the mental investigation process.</p>
<p><strong>Offline architectures</strong>: chalk and blackboard. I know it’s very “boomer,” but it works. Have them draw an architecture, then increase the requirements and follow the reasoning they use to adapt it.</p>
<h2>Labeling and Quarantine</h2>
<p>We must distinguish what is produced by a machine from what is produced by a human. A person has a specific goal, while a machine relies on probability and iteration.</p>
<p><strong>Labeling</strong>: if your code or documents are more than 50% generated by a machine, introduce “AI GENERATED” annotations and indicate the model involved. This helps trace the source of problems in the future.</p>
<p><strong>Quarantine</strong>: avoid letting AI directly handle the product CORE. Use it for tests, utilities, analysis, ideas, but not for the core business. AI as help is good; AI managing the core business risks transforming it into a shapeless mass without a true owner.</p>
<h2>What Will the New Seniors Have to Do?</h2>
<p>The role of the senior changes. It is no longer the person who writes code fastest or knows APIs by heart. Speed will come from tools.</p>
<p>The new seniors will have to:</p>
<ul>
<li>Develop transversal holistic skills</li>
<li>Understand if they are facing a hallucination or truth</li>
<li>Comprehend human specifications beyond literal text, understanding implied implications that machines cannot grasp</li>
<li>Prevent technical debt generated by AI</li>
</ul>
<p>We must reward those who prevent technical debt, not those who commit more features. KPIs must shift from speed to stability and complexity management.</p>
<h2>We Must Not Be Faster, But More Resilient</h2>
<p>Let’s fall out of love with speed: it is not the goal of our work. As CTOs, we must think about sustainable growth, a company model capable of being expanded and understood. We must avoid being slaves to technologies we are unable to govern.</p>
<p>Let’s invest in training, in understanding processes and architectures. Let’s ensure that people specialize in humanistic subjects capable of understanding people, making them capable of intelligently critiquing what is proposed by machines.</p>
<p>If AI is to enter corporate processes, let’s do it consciously, not like a wave that overwhelms everything. Let’s ensure that AI is a tool for growth and not for destruction of skills: it is not enough to create; it is necessary to explain how and why it was created, otherwise, we risk building sandcastles destined to collapse at the first breath of wind.</p>


Matteo Baccan

AI coding tools promise unprecedented speed but risk creating a generation of developers who lack fundamental problem-solving skills. While 84% of developers use AI tools, only 33% trust their accuracy, and 45% find debugging AI code takes longer than writing from scratch. CTOs must implement strategies like AI-free zones, design-first workflows, and cognitive testing to prevent skill erosion. The focus should shift from speed to resilience, rewarding developers who prevent technical debt rather than those who ship features fastest. Success requires treating AI as a tool for growth while maintaining human expertise in architecture, debugging, and critical thinking.

The Illusion of Speed in the AI Era