The Person Who Could Replace You Is Already Using AI

The statement may seem provocative, but it describes a mechanism far less dramatic than some headlines suggest. Probably no artificial intelligence will walk into a hotel tomorrow, occupy an office and announce that it is taking responsibility for the budget, the team and the guest experience. The change will be quieter. While some professionals will continue to use these tools as a curiosity, others will learn to think, prepare decisions and mobilise their teams with far greater capability. The comparison between the two will eventually happen, even if nobody formally calls for it.

I have tested enough tools and observed enough ways of using them to confirm that access barely explains the difference. Two people can have the same system and achieve radically different results. One asks it to draft emails, summarise documents and prepare presentations. The other uses it to challenge assumptions, uncover connections between problems, prepare scenarios, examine risks and arrive at a meeting with questions they would not previously have asked. Both use artificial intelligence, but only one is genuinely expanding their leadership capability.

This distinction matters because roles are not protected by retaining their names. They are protected when they continue to deliver value that the organisation recognises and needs. A hotel may still have General Management, Revenue Management, Marketing, Human Resources and departmental leaders while profoundly changing what it expects from each of those roles. The organisation chart may look identical for years, yet the professional contract behind each box may no longer be the same.

Hotel management retains elements that are difficult to reduce to a tool. It requires reading the atmosphere, interpreting silences, understanding team tensions, sustaining uncomfortable conversations, accepting consequences and making decisions with incomplete information. It also requires walking the hotel, listening to the guest and identifying what has not yet appeared in any report. Artificial intelligence does not remove that reality. What it does is reduce the differentiating value of many tasks surrounding those responsibilities which, until recently, consumed a substantial part of the day.

That is why I am not especially concerned that a machine will literally replace a Hospitality professional. I am concerned about something more immediate and likely: that a professional able to combine hospitality judgement, artificial intelligence, business vision and human leadership will establish a level of performance that makes an earlier way of managing obsolete. Competition will no longer be limited to who has more experience, greater authority or more years in the industry. It will also depend on who learns faster, frames problems better and turns available capability into better decisions.

The role may still exist and no longer justify your position

One of the most common mistakes is to ask which jobs will disappear. The question is understandable, although it does not go far enough. In Hospitality, as in any complex business, a role is not an indivisible unit. It consists of tasks, decisions, relationships, knowledge, routines, responsibilities and outcomes. Some of those elements can be automated, others can be accelerated, and a few must clearly remain under human responsibility.

Hotel management traditionally included a great deal of information-gathering work. People had to request data, chase reports, organise incidents, compare periods, prepare a presentation and reconstruct what had happened before they could think about what should be done. That processing capability had value because it was scarce and costly. When gathering, organising, synthesising and cross-checking information becomes more accessible, the organisation begins to value something else: the quality of the question, the interpretation of context and the ability to execute a response.

This does not mean that preparing a budget, analysing a profit and loss statement or producing a forecast has ceased to matter. It means that professional merit can no longer rest primarily on producing the document. If one person takes two days to prepare an analysis and another obtains a reasonable initial diagnosis in two hours, verifies the data, identifies the important uncertainties and uses the remaining time to work with the team, the difference is not who presses more keys. It is who turns information into responsible action sooner.

This evolution confirms an idea we have already observed when analysing why roles change before their titles do. Workplace transformation rarely waits for Human Resources to update a job description. It begins when a new performance expectation emerges and someone demonstrates that they can meet it. Comparisons, promotions, structural reviews and uncomfortable questions about certain positions follow afterwards.

I call the distance that begins to open up between two professionals with apparently equivalent responsibilities, when one of them systematically expands their ability to analyse, prepare and learn, the leadership performance gap. That gap is not measured only in hours saved. It becomes visible in far more important matters:

• The speed of understanding a complex situation. An augmented leader can gather dispersed information, structure it and identify contradictions before the conversation gets lost in opinions. They do not automatically make decisions faster, but they reach the point at which it is worth deciding sooner.
• The breadth of scenarios considered. Faced with declining profitability, they do not limit themselves to reviewing costs or rates. They can simultaneously explore demand mix, productivity, commercial promise, reputation, acquisition, maintenance and customer behaviour, preventing each department from explaining the problem solely from its own perspective.
• The quality of working documents. Reports cease to be retrospective files and become decision-making instruments. They include assumptions, alternatives, risks and matters still requiring verification, rather than concealing uncertainty behind an especially polished presentation.
• The ability to maintain several time horizons. While addressing this week’s operation, they can prepare decisions for the next quarter and identify implications for future positioning. Urgency does not disappear, but it no longer monopolises all leadership attention.
• The transfer of knowledge to the team. What is learned does not remain locked inside a private conversation with a tool. It becomes better questions, clearer procedures, shared criteria and professional development for others.
• The discipline of follow-up. Decisions can be recorded along with their rationale, conditions and expected outcomes. This makes it possible to assess later whether they worked and avoids the very hotel-industry habit of celebrating an idea on Monday and discreetly forgetting it before Friday.

The sum of these improvements raises the expected standard. An owner who receives clearer analyses eventually loses patience with reports that merely describe what they already know. A team that works with understandable priorities begins to notice when a meeting lacks preparation. A committee that learns to compare scenarios stops accepting the first available proposal. Artificial intelligence has not eliminated any role in these examples, but it has changed the level of contribution that seems reasonable to expect.

Here an important contradiction emerges. The easier it becomes to produce texts, analyses and presentations, the easier it also becomes to simulate competence. We can prepare an impeccably written strategic plan in minutes that does not understand the hotel, a protocol that nobody can execute, or a leadership speech that falls apart as soon as the first incident arises. Formal quality rises easily; management quality does not.

I have read documents created with artificial intelligence that seemed brilliant until I asked a basic question: what specific decision will change tomorrow because of this work? The silence that follows is usually revealing. The tool had produced order, vocabulary and structure, but nobody had made the effort to connect it to actual demand, team capacity, margin, the asset or the guest experience. In Hospitality, a recommendation that ignores these constraints may sound intelligent and still be completely useless.

That is why I consider it dangerous to equate digital competence with the ability to write sophisticated prompts. Knowing how to converse with artificial intelligence will be useful, but it does not in itself constitute a new management profession. Value emerges when you know what context to provide, what information is missing, which response should be challenged and which part of the result should never be transferred directly into operations.

The real change lies in leadership leverage. Within the same working day, a person can prepare a negotiation better, examine more scenarios, understand a variance sooner and devote more time to conversations requiring human presence. They do not necessarily work more hours. They gain greater useful capacity from every hour, provided they do not use the efficiency gained to fill the calendar with new meetings and more documents that nobody requested.

This is also why redesigning management work is more important than automating tasks indiscriminately. If a tool frees up three hours a week and those hours end up absorbed by emails, minor approvals and additional presentations, we have not increased management capacity. We have merely accelerated bureaucracy.

The advantage appears when recovered time is reinvested in what the hotel needs and rarely manages to protect: developing managers, observing service delivery, reviewing the value proposition, hotel strategic planning, commercial relationships, process improvement and conversations with people who should not receive leadership in the form of an email.

We must also recognise that this change may make the professional market tougher. Until now, certain management limitations remained partly hidden by workload. It was possible to justify a lack of analysis because there was no time, weak follow-up because operations consumed the day, or an absence of planning because data arrived late. As some of these frictions diminish, organisations may begin to distinguish more clearly between a resource constraint and a judgement constraint.

It will not always be a fair comparison. Some hotels will have better systems, more reliable data, training, budget and corporate support. Others will continue operating amid incompatible spreadsheets and inherited procedures, which is a very hotel-industry way of practising administrative archaeology. Yet even within the same organisation, differences will emerge between those waiting for someone to hand them a solution and those learning to build a better way of working with the resources available.

Artificial intelligence also does not automatically make someone competent if they do not know the business. It can help explain RevPAR, but it does not replace an understanding of commercial mix. It can propose a response to a review, but it does not know on its own which operational failure caused it. It can prepare options for staffing levels, but it cannot perceive the effect of deploying less capacity precisely when service creates value. Hotel management still requires knowledge of the ground.

I have argued on other occasions that the smart hotel is the one that makes the best decisions. The same logic can be applied to people. The augmented professional is not the one who generates more content or mentions AI in every meeting. They are the one who uses new capabilities to raise the quality of their judgement without hiding behind the tool when the time comes to answer for the outcome.

The management capabilities now determining who progresses

If the bar is changing, it is worth specifying which capabilities are becoming decisive. I do not believe every professional must become a technical specialist. Nor do I consider it sensible to ask every hotel leader to understand the internal architecture of models. I do believe it is necessary to develop sufficient literacy to understand what they can contribute, where they fail and how to incorporate them into work without diluting accountability.

The first capability is problem framing. Many poor decisions begin before the analysis, when we accept a badly framed question. Asking how to reduce labour costs leads to a different conversation from asking what capacity the hotel needs to fulfil its promise profitably. Asking how to increase occupancy is not the same as asking which demand truly contributes to the result. Artificial intelligence can generate numerous answers, but it will hardly correct a misguided ambition if nobody allows it to challenge it.

The second is context-building. A general-purpose tool does not inherently know your hotel’s constraints, priorities, standards and contradictions. You must explain the segment, the demand period, available capacity, guest type, positioning, policies and financial limits. The more important the decision, the less reasonable it is to rely on a generic query.

The third capability is professional verification. A convincing response may contain errors, debatable assumptions or recommendations incompatible with operations. Verification does not consist only of confirming figures. It also requires asking whether internal sources are reliable, whether the reasoning omits a cost, whether the proposal creates a precedent or whether the alternative shifts work to another department. Hospitality knowledge becomes more valuable, not less, because it is what enables us to recognise a plausible but incorrect answer.

The fourth is decision-oriented synthesis. Management does not need to produce more information, but to turn it into understandable options. Good synthesis distinguishes facts, interpretations, assumptions and preferences. It explains what we know, what we believe, what we cannot confirm and what waiting would cost. This discipline improves both a hotel revenue management meeting and a reputation review, a supplier negotiation or an investment decision.

The fifth is translation across functions. Hotels lose a considerable amount of energy because each department speaks its own language. Revenue speaks of demand, Finance of margin, Operations of capacity, Marketing of conversion and the guest, quite logically, speaks about whether they can have breakfast at a particular time. A competent leader must connect these perspectives and show how a commercial decision affects service, the team, cost and the brand promise.

The sixth is orchestrating human and assisted work. Not all tasks require the same intervention. Some can be automated, others require review, and certain decisions should remain outside any autonomous execution. Establishing those boundaries requires clear authority, controls and traceability. That is why an organisation adopting these capabilities also needs an authority matrix for hotel AI, especially when recommendations may affect pricing, staffing, guest communication or commercial commitments.

The seventh capability is non-delegable accountability. The phrase “the system recommended it” should not close any relevant discussion. If a decision harms the guest, erodes margin or undermines the team, someone must explain why it was accepted. The tool may participate in preparation; accountability remains with the person and organisation that decide to use it.

The eighth is learning speed. Those who use AI only to save time achieve a limited improvement. Those who use it to test their own ideas, review past decisions, prepare questions and discover gaps can accelerate their professional development. The difference between the two uses is profound. The first produces efficiency. The second builds capability.

These capabilities can be observed in very everyday situations. Faced with a budget variance, a superficial use consists of asking for a summary of the financial report. A management use consists of separating structural from temporary causes, exploring relationships between revenue and operational capacity, formulating scenarios, identifying missing data and preparing a conversation with specific accountable leaders. The tool helps, but the value lies in the process design.

Something similar happens in hotel marketing. It is easy to generate campaigns, copy and content calendars. It is harder to decide which segment deserves investment, which promise operations can sustain, how to measure the profitability of acquired guests and which message fits the positioning. Creative production becomes cheaper, while commercial judgement gains importance.

In Revenue Management, automation is already challenging part of the traditional work. The change does not condemn the discipline, but it does require it to evolve from rate changes towards commercial strategy, total profitability, demand interpretation and business coordination. That tension explains why certain ways of practising Revenue Management are losing value. The same process can affect any management function that mistakes activity for contribution.

In hotel guest experience, AI can analyse comments and group incidents. Management capability begins afterwards. It is necessary to determine which issue deserves priority, which department must intervene, which policy is causing it and which improvement can be implemented without creating new friction elsewhere in the guest journey. Analysing thousands of comments is of little use if breakfast still starts late every morning.

To turn these ideas into a manageable practice, I propose an Augmented Management Competence Test. It is not intended to assess how much you know about technology. It seeks to establish whether your way of working is producing a different capability. You can apply it to a significant decision made in recent weeks:

• Did you define the problem before asking for answers? Write down the initial question and review whether it was focused on the real objective or assumed a solution was valid in advance.
• Did you provide enough hotel context? Check whether you included operational constraints, financial impact, the guest promise, team capacity and the time horizon.
• Did you separate confirmed information from assumptions? Identify which data came from the hotel, which elements were estimates and which claims required verification.
• Did you consider genuinely different alternatives? Three variations of the same idea do not constitute three scenarios. Look for options involving different decisions, trade-offs and risks.
• Did you explain what could go wrong? Every professional recommendation should include failure conditions, early warning signs and a reversal mechanism where possible.
• Did the decision improve the team’s capability? Review whether people understood the criteria and could resolve a similar situation without relying on the same management intervention.
• Did you record the outcome in order to learn? Define when you will revisit the decision, which indicator you will monitor and what conclusion you will draw if the effect does not match expectations.

If these questions cannot be answered clearly, you may have used artificial intelligence, but you have not yet transformed your professional competence. You have accelerated a task. That may be useful and even advisable, but it should not be confused with management evolution.

I also suggest reviewing the calendar. The agenda reveals better than any speech what an organisation considers valuable. If the time freed up remains occupied by manual controls, meetings without decisions, minor authorisations and information reconstruction, the new capability has not changed the model. The hotel continues to pay management talent to perform work that should have been simplified, delegated or eliminated.

A sensible implementation can begin with four moves. First, identify five recurring cognitive tasks that consume time without requiring all your judgement. Then choose only one or two and design an assisted way of carrying them out. Next, establish how you will verify quality and what information must never be incorporated without controls. Finally, decide in advance where you will reinvest the time recovered.

This final decision is essential. If you recover two hours of preparation time each week, you can devote them to supporting a manager, observing part of the service, reviewing a recurring cause of complaints or working on a hotel strategy that still cannot find space in the agenda. Management productivity only improves the business when freed capacity is directed towards higher-value responsibilities.

The team must be part of this transformation. Concealing the use of tools creates suspicion and hinders shared learning. Imposing them without explanation causes another understandable form of resistance. I prefer to speak transparently about what we use, what for, under which limits and with what review. It is also worth making clear that nobody will be assessed solely on producing more content. Speed without judgement can multiply mistakes with admirable efficiency.

There is another precaution I consider decisive: we must not allow constant assistance to weaken the ability to think without it. A professional who always accepts the first proposed structure, the first hypothesis or the first argument may gain speed while losing intellectual autonomy. That is why I continue to defend spaces for independent analysis, screen-free conversations and direct contact with operations. Artificial intelligence should expand thinking, not replace the effort of understanding.

Leadership in the hotel industry will continue to depend on deeply human elements. A system can help prepare a difficult conversation, but it cannot assume the emotional impact of having it. It can identify turnover patterns, but it cannot build trust on its own. It can organise arguments for a negotiation, but it cannot fully interpret the relationship, the history and the moment. The hotel still needs people capable of being present when a decision carries weight.

What does change is the level of preparation with which we arrive at those moments. Having more humanity should not be used as an excuse for poorer analysis. Defending experience does not justify working with disorganised information. Championing intuition does not mean giving up on testing it. The best hotel management will combine human presence and augmented capability without turning either into an excuse.

My advice is not to begin by asking which tool you should buy. Choose a recurring decision that currently forces you to gather information, compare options and prepare a judgement. Redesign that process, measure how much time you recover and assess whether the decision improves. A credible management transformation begins with an observable difference, not with a collection of applications.

Protect your judgement capital in particular. Read, listen, visit operations, speak with other professionals and review past decisions. Use artificial intelligence to challenge yourself, not merely to validate yourself. If you always agree with what it returns, perhaps you have not found an extraordinary assistant; perhaps you have built an excessively polite mirror.

Artificial intelligence will probably not knock on the door to ask for your role. The market, ownership, the team and results will compare your capability with that of professionals who are already learning to use it. The best protection will not be resistance or a display of technological enthusiasm. It will be demonstrating that, with you, the hotel understands sooner, decides better, learns faster and remains deeply human.