Is elBulli an algorithm? (I)

Dish number 1846 was the last. Ferran Adrià decided not to invent any more. He wanted to move from the workshop to reflection and thought. He wanted to know what happened in the last 25 years of his life as he became a chef capable of transforming world cuisine.

Adrià is now sorting through all his knowledge and next October he and Fundación Telefónica will be presenting much of what was learned in the exhibition‘Ferran Adrià. Auditing the Creative Process’.

Meanwhile, the publicistsToni SegarraandJorge Martínez andthe mathematician Enrique Graciánare trying to discover the invisible (so far) score that led to elBulli. The findings will be applied to a guinea pig advertising agency created by Jorge and Toni, calledThe Table, to find out what a Bulliesque agency would be like.

The project starts from observation and interviews with several leading figures from elBulli. Then come hours of reflection and analysis. Conversations where everything is assembled and disassembled. Where there is light and some ideas are discarded.

It’s 1st April in Madrid. The three are sitting at a table. There is a notebook, a couple of mobile phones and two green stones. Yes. Two green stones.

Toni: We are trying to apply the research that Ferran Adrià did in the kitchen to the creation of the advertising campaign for the exhibition ‘Auditing the Creative Process’. In this process we want to clarify an unknown.We are starting with the idea that elBulli was to the traditional restaurant what a Bulliesque agency would be to the traditional advertising agency. In this equation we have to define the X factor of the Bulliesque agency.

Enrique: If we uncover the process that turned elBulli into a research laboratory, we will have a model applicable to any type of company. We would have the theory.The experience would be to apply it to an advertising agency and that agency would be like the first animal in the laboratory to which we administer the serum, to see whether it lives or dies.

To be able to understand these processes, we can use a scheme based on mathematical concepts, but which are often applied outside this area.  It is the theory ofblack boxes and white boxes.A black box is a device in which certain elements enter and others come out. We can understand any manufacturing process as a black box. The simplest example is this: two numbers enter – 2 and 3 – and 5 comes out. That means that an addition has taken place inside. To be a black box, it is important that what goes in and what comes out be different elements.It also implies that a transformation occurs inside, but above all that what happens inside does not require any decision-making.It reflects the idea of an automated process, meaning it is an algorithm.

An algorithm is a command.It is an operation; the one who performs it knows what to do. There are no doubts.Two numbers are entered and since I know the addition algorithm, I get the expected result.

Toni: That would explain the difference between handicraft and mass production.

Enrique: It’s the idea of not relying on one person to carry out a process. The goal is that if the instructions are correct, in the end a machine could perform the activity. And, in fact, a machine can calculate numbers. One example I often use is that of cooking, precisely because it is so clear.The algorithm would be the recipe.

At the entrance to the box you have a number of elements. Eggs, potatoes, salt, oil… Inside you have some instructions (which are the key to the box because they tell you what you have to do) and some tools (which are constantly inside it) and, in the end, what you get is a Spanish omelette.

Jorge: You have already talked about that to theelBullifoundation, team, but they do not feel very comfortable with the idea.

Jorge Segarra

Enrique: In cooking, there is a phenomenon similar to what occurs in music. What we hear is on a score. Can a machine play it? Yes. In fact, they already do. But there is also the interpretation, and that is where the artist emerges. The interpretation gives you a margin in obeying the algorithm. You can make a longer, shorter, stronger note… And in cooking, the chef, in some way, feels like the interpreter. This means that cooking has an artisanal touch, even though it may be governed by black boxes or have a well-defined algorithm… I don’t think these things are incompatible. It’s a conceptual issue.The chef knows he/she is an interpreter, but this does not mean that if we look at the chef as a company we cannot treat him/her as a black box.

In cooking, the chef, in some way, feels like the interpreter

Toni: You’re saying that the map of creativity they are working on at the elBullifoundation is like a circuit of black boxes.

Enrique: Yes. I gave an example with the recipe for white bean foam which, when you see it on paper, looks like a computer program. This is important, because, in a company, the first thing you have to check is that the circuit of black boxes works and can always be relied upon. The map of the culinary process at elBulli could lead to a system of black boxes because there are no holes. There are no doubts. You know exactly what to do. And not just in the kitchen. In the spaces where the diners sit down to eat as well. There is also a black box in there, where a product that is served at a restaurant goes in and a payment comes out. That is also a black box with its own algorithms.From a raw foodstuff until an individual pays a bill there is a circuit of black boxes, and that can be perfectly sketched out.

Toni: This theory argues that once those black boxes are perfectly built, you need to create white boxes.

Enrique: Yes. In black boxes, certain elements go in and others come out. In white boxes, information goes in and information comes out.The white box is a research process with a goal, and this is basic, because if there is no goal, there is nothing.

Toni: You mean, you have to be looking for something…

Enrique:   Sure.  In the case of stewed white beans we start with the beans, water, salt… Then we add the instructions (the recipe) and a number of tools (the hob, pot, etc.).  At the end, stewed white beans are produced.  But here comes the interesting part.  In elBulli they proposed creating something new: a white bean foam without milk, so it would not interfere with the taste.Until then they had always used a soda siphon and milk for foams.This time there was a new aim. They wanted to get that texture, but without milk. Here begins a process of research and a white box appears.A new tool was introduced that was not in the black box for the stewed white beans: the soda siphon.Afterwards, they experimented until they discovered the new key element to replace milk: gelatine.They see that it is possible to emulsify the product properly without altering the flavour of the bean.The interesting thing is that what comes out of a white box goes into a black box.This is where the instructions for cooking white bean foam have come from.Now there is a black box for white bean foam.This means that a person who has not been in the white box for white bean foam (the research process) is still capable of replicating the recipe, because it is now a series of instructions.

Toni: So what elBulli does is to build a series of recipes. Rather than affecting the restaurant business, elBulli influences cuisine as a whole.They created 1846 recipes that can be made by any restaurant in the world.

Each recipe comes from a white box

Enrique: Yes.White boxes create knowledge. elBulli, at some point, decided to generate knowledge.


Toni: In short, we can say that, until that point, restaurants were based on black boxes (replicating recipes), and then there was this moment when Ferran considered that white boxes were needed and built a research machine.

Jorge: The interesting thing is that he does it by asking questions.

Enrique: Sure.Like any good research centre.

Jorge: It’s what most companies don’t do.Most companies are black boxes…

Enrique: What’s more, at elBulli the circuit of white boxes kept going. After setting up the black box for white bean foam, he said he wanted a hot foam. The research process began again, in a suitable environment, where a number of people who already had the knowledge started to do research the way they would in a physics and chemistry laboratory. They investigated ways to heat the foam.They did something as strange and as risky as…

Toni: Now we’ve come to that moment in the elBulli story when everything could have disappeared.The moment Oriol put the soda siphon into a double boiler and only a miracle stopped it from exploding.

Enrique: Indeed. The soda siphon contains nitrogen. Putting it in a double boiler is extremely risky.In a laboratory with protocols he would not have been allowed to do it.

Jorge: These things are never down to chance, are they?

Enrique: No. Not at all. Those of us outside the box can consider them to be random. We say: “Oh, what a coincidence! What a stroke of luck that was!”But it’s never random for those inside, because there was a purpose and a goal.

Toni: The typical case is that of penicillin. They always say: “It was sheer luck.” But that man (Fleming) was searching.Millions of people had probably discovered penicillin before that, but they never realised it.

Enrique: What matters is the scenario in which it occurs. We’ve all had penicillin moments many times in the kitchen at home, but we’ve thrown the result away because we thought it was just rubbish. It’s not random when it emerges in a research laboratory. It is what he was looking for, and he found it.Chance is a bit mythologised.

Toni: That would lead us to conclude that the intelligence is in the place.It lies in the setting.

Enrique: This is one of the most important messages we can convey. Especially when considering creative environments. There is a tendency to say that what matters is the creative capacity and intelligence of the subject. That would mean that a company would have to find such subjects, but in reality what it needs to do is provide the setting for that to occur.The creativity and intelligence is in the setting.

Toni: I mean, you need talent, but unless you place it in the appropriate environment it won’t produce results.

Enrique: Sure.And less so when they are research projects with a specific goal.

But let’s close the circuit of black boxes and white boxes at elBulli. From the bean foam white box, a black box emerges. In turn, a new white box was born to find hot foam… From one bit of research others emerge, and the circuit of black boxes is completed. Thus arises a culinary culture.

We have a research area, a goal to be pursued and a number of elements that we are going to work with.If we can take away a more or less tangible analysis from this, we can set up the Bulliesque agency with the same method.

We want to try to universalise the elBulli method

Toni: What we are proposing when we try to universalise the elBulli method is for any company to be able to use this method of research, especially if it wants to succeed in this changing environment in which we live.That is what we call innovation, and we have not yet fully defined it…

Enrique: That’s right. Companies are doomed to innovate. If you look at yourself in the mirror, you know you have a creative person in front of you, because you have proved that you are. Ferran Adrià knows he is looking at a creative person, but if an employer is suffering a crisis and is only months away from shutting down the company, and you say to them: “Your solution is to be innovative and creative,” you are not helping them at all because they do not even know what it is to be creative.You need to give them a method.

Jorge: That’s one of the findings we are discovering.We are beginning to associate research with creativity.

Read the second part of this three-way conversation here: Is elBulli an algorithm? (II)