A Signal from University Admissions

When I learned that there was talk of roughly thirty applicants for each place at the University of Bucharest’s Faculty of Foreign Languages and Literatures, my first reaction was:

If they all stay in the country, we shall become a nation of potential polyglots.

The joke conceals a serious question. Why are so many young people turning to foreign languages? Out of an unexpected passion for philology? From a desire to know other cultures? Or because foreign languages promise mobility, access to a wider world, and perhaps the possibility of working anywhere without another professional specialization seeming indispensable?

Competition in Computer Engineering at the Polytechnic University, always intense, appears especially high this year. At the same time, technical faculties want to introduce courses, tracks, or programs related to artificial intelligence.

The phenomena seem different, yet they express the same orientation. Young people seek fields that promise access to the future. Foreign languages open geographical borders. Computer science and artificial intelligence appear to open professional ones.

Can a Field Be Created by a Decision?

A distinction must be made here. A specialization is not invented at the whim of a faculty. It must be recognized within the national framework, while the organization of a program entails external evaluation, authorization to operate, and later accreditation. The institution must demonstrate that it has the staff, infrastructure, and quality-assurance mechanisms required.

In brief, authorization allows a program to begin under controlled conditions; accreditation confirms, after a further evaluation, that it can operate on a stable basis. The procedure is necessary and far from simple. Yet even the most rigorous procedure cannot compress intellectual time. It can verify the existence of the declared resources; it cannot confer, by official act, the maturity of a scientific school.

A specialization may be recognized and a program may be authorized. A field, however, matures through research, continuity, results, and the formation of a community.

A Return to the 1960s

When I entered university in 1961, education in automatic control was still being established in Romania, within the Faculty of Power Engineering. The Automatic Control specialization was formalized in 1963, and in the 1966–1967 academic year the creation of the Faculty of Automatic Control and Computers was decided.

Looking back, I do not believe Romanian automatic control was born solely from an urgent demand clearly articulated by industry. There was not yet a multitude of enterprises asking for specialists in an already well-defined profession. There was, however, the conviction that automation would become indispensable.

The impulse came largely from specialists’ desire to keep pace with developments in the industrialized countries and, under the political conditions of the time, especially in the Soviet Union. As Caragiale might perhaps have put it, “we too must have our own bankrupts.”

The ironic phrase captures institutional imitation, but it does not exhaust the phenomenon. In automatic control and, a little later, in computers and computer science, the top-down decision did not remain without consequences. People were trained; laboratories, courses, institutes, methodological centers, and eventually applications appeared. What began as anticipation then helped produce the anticipated reality.

The University That Anticipates Demand

The precedent of the 1960s is useful only if we also recognize its limits. At the time, the impulse came largely from above, in a planned economy, while the university was preparing a modernization that industry had not yet adequately articulated. Had everyone waited for industry alone to express a fully mature demand, automation and computerization would probably have begun much later.

The creation of departments, the training of professors, and the emergence of graduates gradually made the application of the new technologies possible. The university anticipated the future and, to some extent, brought it about.

Today, in the case of artificial intelligence, pressure comes simultaneously from several directions:

• the desire to keep pace with major international trends;
• the strategies of the state and the European Union;
• competition among universities for applicants, projects, and funding;
• the real needs of enterprises that have undergone digital transformation;
• the expectations of a public already using AI through services available on smartphones;
• young people’s hopes concerning the professions and incomes of the future.

Demand now comes from everywhere, but it does not have the same substance everywhere.

The analogy lies in the university’s anticipatory role: both then and now, training specialists can accelerate economic transformation. The difference is that AI is not entering an almost empty field. It encounters a society already digitally connected, enterprises that are very unevenly prepared, and commercial tools that change before university programs can be stabilized.

From a Common Core to Specializations

The comparison between automatic control and artificial intelligence is tempting because both are cross-disciplinary fields.

Automatic control belonged neither exclusively to power engineering, mechanics, chemistry, nor machine building. It had a theoretical core of its own—systems theory, regulation, control, and stability—which was then applied differently in each branch of engineering.

At university I encountered both foundational courses—among them Theoretical Foundations of Automatic Control Systems—and application-oriented subjects: automation in power engineering, machine building, chemistry, telemechanics, and other areas. A structure gradually emerged: a shared theoretical center and several specializations tied to fields of use.

Specialization is not a sign of inadequacy. It is the consequence of the maturation of knowledge. No one should feel at fault for being unable to absorb, day after day, everything that appears under the label of artificial intelligence.

The university’s problem is to decide what belongs to the core and what is merely a transient tool. A student must know the technologies of the present, but the value of university education lies chiefly in the capacity to understand the next generation, not merely to use the current one.

A Society Digitalized Before Its Institutions

The smartphone has brought digital services and, more recently, artificial intelligence to the level of the personal interface: communication, maps, banking, translation, summaries, and almost instantaneous answers. This individual experience creates a natural social expectation: public administration, hospitals, schools, and businesses should offer the same speed and ease.

The smartphone has digitalized the human relationship with the world without digitalizing, to the same extent, the structures of the world with which people interact.

A paradox therefore emerges. The public demands artificial intelligence precisely because it has experienced it directly, yet some of the fields upon which it exerts pressure have not even completed their digital transformation. An institution may have a website and an app while documents still circulate on paper, information is repeatedly re-entered, and databases do not communicate with one another.

Façade Artificial Intelligence

When digitalization is superficial, artificial intelligence risks becoming a modern label applied to an old organization. A physician may consult an AI system, a teacher may generate materials, and a civil servant may draft a text—but these individual uses do not in themselves transform an institution.

For AI to transform a field, an ordered reality must first exist: defined processes, accurate and interoperable data, and clear responsibilities. A system cannot coherently analyze data that do not exist or cannot be correlated. It cannot optimize a process that has not been described. It cannot make responsible decisions in an institution where no one knows who is accountable for the data and the results.

Without structural digitalization, the danger of façade AI appears:

• a modern chatbot in front of an administration that still works mainly on paper;
• a sophisticated model fed with incomplete or incompatible data;
• an “AI strategy” in a company that has not described its basic processes;
• individual use of public AI services for confidential information;
• a modern label attached to an old organization.

Partial use of AI tools is not useless: it can save time and improve individual work. But using an AI product must not be confused with transforming an organization through AI. Genuine transformation requires infrastructure, data governance, and clear mechanisms of accountability—not merely access to a model.

From an Approved Specialization to a University Program

Uneven digitalization inevitably carries over into education. Universities are under pressure to train AI specialists, but employers do not demand a single kind of “AI specialist.” They need different competencies: model research and engineering, data engineering, systems integration, security, and deep knowledge of the application domain.

For this reason, artificial intelligence should not remain confined to a single faculty. In adapted forms, it can reach power engineering, mechanics, transportation, medicine, economics, public administration, law, or the humanities. Interdisciplinarity, however, is not achieved simply by adding one course. It requires shared subjects, projects led by professors from different fields, practical experience in real organizations, and, where possible, combined educational pathways.

A faculty cannot invent and open a specialization by itself overnight. The specialization must exist within the recognized national framework. An institution may apply to organize a program in that profile if it demonstrates that it has teaching staff, infrastructure, research activity, and quality-assurance capacity. External evaluation, authorization to operate, and later accreditation follow.

Does demonstrating the resources needed to organize a program also guarantee the existence of a mature academic community?

Evaluation can confirm the presence of professors, courses, laboratories, and quality mechanisms. It is harder to measure the continuity of research, the depth of experience, and the team’s capacity to separate the foundations of the field from tools that are currently fashionable.

Here the risk of university branding also appears: an attractive name can become an admissions argument before the profile acquires substance. Prevention requires not only administrative control but also public transparency. A program should clearly show who teaches, what research the team conducts, what laboratories exist, what interdisciplinary projects it offers, and what outcomes its graduates achieve.

The risk is not the absence of all control. The risk is the rapid multiplication of programs in a field that changes faster than the teams capable of teaching it in depth can be formed.

The State That Creates and the State That Restrains

In the case of automatic control and computers, the state played a decisive role. It created departments, faculties, institutes, and enterprises and funded the training of specialists before the entire economy was ready to employ them.

In the case of artificial intelligence, state involvement is even more necessary, but its role is now twofold.

It is not enough to speak only of “ethical regulation.” Ethics formulates principles; law and institutions establish obligations, audits, accountability, and avenues for appeal. The European AI Act follows a risk-based approach: some practices are prohibited, high-risk systems are subject to special requirements, and certain uses carry transparency obligations.

Artificial intelligence may penetrate medicine, justice, education, public administration, defense, energy, and communications. In such fields, an error is no longer merely technical. It can become a social, legal, or political decision.

The more serious the consequence, the less acceptable the explanation “the algorithm decided” becomes. A person and an institution must be accountable.

But the state is not only the arbiter of artificial intelligence. It is also one of its most powerful potential users. The same systems that can streamline administration, detect fraud, or prevent accidents can be used for excessive surveillance, classification of citizens, manipulation of information, and political control.

State involvement does not automatically guarantee responsible use of AI. The state itself must be controlled through law, transparency, audits, an independent judiciary, and public debate.

The university has an additional responsibility here. It must train not only builders and users of systems, but also people able to ask where AI should not be used, who is accountable for its consequences, and how an automated decision can be challenged.

What Young People Seek and What the University Owes Them

It would be unfair to view admissions competition only with suspicion. Young people seek what every generation has sought: a profession, an income, a social position, freedom of movement, and the feeling that they belong to the present rather than to a world that is disappearing. Artificial intelligence seems to offer all of these things; it is natural that they should be attracted to it.

The university, however, must not sell them hope alone. It owes them:

• an honest picture of the field and of real professional opportunities;
• foundations that endure rapid changes in tools;
• practical experience and contact with real problems in digitalized organizations;
• an understanding of the data, processes, limitations, and risks of AI systems;
• the capacity to continue learning and to specialize as the field evolves;
• responsibility for the social effects of the systems they will build or use.

It is not enough for a graduate to know how to use the tools of the moment. Graduates must understand the field in which they apply them and be able to judge their results. Artificial intelligence is not a single profession, but a set of methods that will combine with medicine, energy, industry, public administration, law, culture, and education. Tomorrow’s specialist will be formed at the meeting point between a common core and deep knowledge of an application domain.

Between Anticipation and Imitation

The university cannot wait until artificial intelligence has completely stabilized. If it did, it would arrive too late. But neither can it turn every novelty into a course and every fashion into a specialization. It must anticipate without merely imitating.

Automatic control and computer science offer a precedent: they were introduced partly through decisions from above and through the desire to keep pace with the technological world, even when domestic demand was not yet fully formed. By creating courses, laboratories, and specialists, the university contributed to the modernization of the economy and, over time, produced the demand for which it had been preparing.

In the case of artificial intelligence, the situation is more complex. Demand already exists, but unevenly: the public has encountered AI through the smartphone and asks for it everywhere, while institutions and enterprises are not all prepared to integrate it. Some have not even completed structural digitalization. The university therefore stands between a technology in ferment, an impatient society, and an unevenly digitalized economy.

How, then, is a university field born? Not through the mere accumulation of courses, nor through the success of a name in university admissions. A field is born when enthusiasm is transformed into method, tools are placed upon foundations, communities capable of researching and teaching emerge, and applications begin to differentiate.

In the 1960s, automatic control appeared as a promise and an act of synchronization. It needed time to form its professors, language, books, and applications. Artificial intelligence is now undergoing a similar process, but one that is far faster, broader, and harder to control.

The role of the university is neither to stop this movement nor to race blindly after it. Its role is to introduce order into knowledge that is still unstable: to separate foundations from fashion, methods from products, competence from mere familiarity, and progress from the uncontrolled use of technological power.

And the role of the state is not only to help the field be born, but also to ensure that society can live safely with what it has created.