AI in Hotel Revenue Management: Lessons From Centaur Chess

AI in hotel revenue management is moving faster than most commercial teams expected, and the way you respond in the next few years will shape your career. The best response is older than you might think. Here you will learn what a 1990s chess story teaches about working alongside AI, and three habits you can start this quarter.

The Day the Machine Won

In May 1997, in New York, Garry Kasparov sat across the board from IBM’s Deep Blue and lost the match. He was 34 and widely regarded as the strongest chess player who had ever lived, and the headlines wrote themselves: man beaten by machine, the end of an era.

Most people in his position would have spent the following months resenting the computer. Kasparov asked a different question instead, which was what would happen if the human and the machine sat on the same side of the board.

The following year he launched what he called Advanced Chess, later known as centaur chess. In these games, a human player worked with a chess engine, consulting it, overriding it, and using its calculations to find moves they could not have seen alone. The results surprised almost everyone.

A human working with a machine consistently beat a machine playing alone. In a 2005 freestyle tournament, the winners were two amateur players using three ordinary computers, who beat teams that included grandmasters with more powerful hardware. The strongest player in the room was neither the best human nor the best machine; it was the team with the most disciplined way of working together.

For a period of years, the centaur was the strongest force on the board.

Why This Matters for Hotel Commercial Teams Today

Almost everything your commercial team does now has its own version of that 1997 moment. AI can read 10,000 guest reviews and pull out the main themes in seconds. It can adjust prices faster than a revenue manager can refresh a spreadsheet, and it is quickly becoming able to analyze data more thoroughly than any single analyst. Strategic analysis is the next layer it will reach.

Faced with this, many hotel teams react in one of two ways. The first reaction is denial, the belief that your skills are too specialized to be affected. The second is delay, the plan to deal with it when you have to. Both reactions are very human, and both are the opposite of what Kasparov did.

There is a third option, and it is the one worth taking: become the centaur now, while the opportunity is open.

Understanding the In-Between World

You can think of the next few years as an in-between world, a hybrid period in which AI is strong enough to do most of the analytical work but not yet strong enough to replace the judgment, the taste and the understanding of why a particular hotel behaves the way it does. Nobody knows exactly how long this period will last; it could be three years, or it could be five.

During that time, a professional who partners with AI will outperform AI working alone, and will comfortably outperform a professional working without it.

It is only fair to recognize the other side of the argument. AI working alone will keep improving, and the gap narrows every six months or so. Anyone who tells you that progress is overhyped is looking at one part of the field and missing the rest, so you should expect this window to close eventually.

That is exactly what makes it an opportunity. Teams that use this period well will come out of it far stronger than they went in, while teams that wait may find the advantage has already gone to someone else.

Three Habits to Build This Quarter

The three habits below come up again and again in conversations with hotel commercial teams. They are specific, they do not require new hires, and you can start them on Monday.

1. Partner with AI Every Day

Run every analytical task through AI this week, including pickup analysis, variance explanations, forecast reviews, comp set reads, and group displacement decisions. Use it alongside your own judgment rather than as a replacement for it.

Study the AI the way an apprentice studies a master, and also the way a master studies a tool. Keep a simple log of what it gets right, where it fails, and why it failed. Within a few weeks, you will have built a skill your team did not have before, which is knowing where AI in hotel revenue management can be relied on and where it still needs a human eye.

2. Expand Your Scope

If AI saves you ten hours a week, put those hours into growing your role. You might take on more properties, pick up a new area of responsibility, or finally tackle the project nobody has had time for in the past three quarters.

The aim is to show, through the work you deliver each week, that you, working with AI, can handle more than a whole team could without it. This is what makes you promotable in the centaur era.

3. Practice Overriding

Overriding is the hardest discipline and the one most teams practice least. Suppose the system gives you a rate recommendation that feels wrong because your comp set is going through a shift the model has not seen yet. You override it, and then you track whether you were right.

Over time, this builds what you could call exceptional judgment. Each override also teaches the system something about your market. You gradually become the player who knows when the engine is reading the position wrongly and when it has spotted something you missed.

Taken together, these three habits are less a transformation plan than a new way of working, and they fit comfortably within a single quarter.

Why Your Professional Identity Matters Most

The commercial professionals who struggle in the coming years will rarely struggle because AI was better at the job. More often, it will be because they found it hard to let go of the way they had defined themselves.

Consider two revenue managers. The first describes the role as “I am the person who runs the forecast.” That person is exposed, because the forecast is increasingly produced by a model that improves every few months. The second says, “I am the person who decides whether to trust the forecast and what to do about the gap.” That person is in a much stronger position, because the first role is one AI absorbs, while the second is one AI amplifies.

This is Kasparov’s lesson applied to your work. He did not stop being a chess player; he stopped being only a human chess player and became a different kind of player. The change in how he saw himself came first, and everything else followed.

What A Centaur Team Looks Like in Practice

In a hotel group, the centaur approach usually means giving your team an AI layer that reads the data, cross-checks segment performance, writes the first draft of the variance commentary, prepares counter-offers, and highlights sales opportunities nobody was actively pursuing. The AI does not replace your sales team. It points them towards business they might otherwise have missed and leaves them to close it. Kassandra, the AI agent from Juyo, is one example of a tool built to work this way.

Across properties where this approach works well, the pattern is consistent. Teams that treat the AI as a partner, override it when they know better, and use the time it gives back to extend their scope are outperforming teams that still run the week manually. Teams that carry on as if nothing has changed tend to fall behind without fully understanding why.

Free Report: Hospitality Analytics Trends Report 2026

Hotel leaders are navigating labor shortages, AI disruption, evolving guest expectations, and increasing data complexity. Based on industry research and real-world examples, this report outlines the major analytics trends shaping the hospitality industry in 2026 and their implications for hotel strategy.

Click here to download the report “Hospitality Analytics Trends Report.”

The window for building centaur skills is open now. Start this week by running one routine analysis with AI, tracking one override, and deciding how you will reinvest the time you save, so your team leads rather than follows.