GENERATIVE ART

Generative art is a mode of artistic creation that consists in developing an algorithm that will generate an artwork. 

While the practice of designing algorithms capable of generating artworks began in the early 1960s, most of these experiments have so far focused on producing visual art, and few choreographers have attempted to design systems capable of choreographing.

The most notable choreographic artist to have tackled this is probably Merce Cunningham, who developed a procedural decision-making system with John Cage in the 1960s, which he used throughout his career to shape certain aspects of his work. 

However, with today's technology and our combined expertise, we believe we can bring the project of generative choreography to a whole new level of complexity and completion. 

Here are some examples of the geometrical patterns that our algorithm is already able to generate — and that we already have used the compose the piece for the 36 P.A.R.T.S. students.

NOT AI 

It is essential to note that the algorithm we are developing is not an artificial intelligence (AI) algorithm. It is purely a set of compositional rules that we are defining ourselves and translating into code. 

It does not draw inspiration from external data to generate choreographies. Instead, every choreographic work it produces is the direct fruit of our own choreographic thinking.

The fact that our algorithm is not trained on external data is crucial for two reasons : 

• Developing this algorithm does not consume the large amounts of energy that are typically associated with developing AI algorithms. In fact, it will be remarkably clean. 

• It does not appropriate the work of other choreographers, but follow its intrinsic logic.

Fundamentally, our algorithm consists in a set of carefully crafted mathematical rules able to generate highly complex geometric patterns.

Excerpts of Kaleidoscopic Dance, by Busby Berkley, and Calico Mingling by Lucinda Childs.

HISTORICAL PRECEDENTS 

Our work firmly inscribes itself in a lineage of choreographers concerned with writing the movement of dancers in space. 

The words of William Forsythe come to mind: “[choreography] is about organizing bodies in space,” and they resonate with those of Anne Teresa De Keersmaeker: “choreography is the practice of writing movement in space and time.” 

These choreographers’ works — with both of whom we have established personal conversations about the project — have been influential on our thinking, alongside with the musicals of Busby Berkeley and the pieces of modern and postmodern choreographers such as Merce Cunningham and Lucinda Childs. 

We also ground ourselves in the long-standing classical tradition of choreographing complex group movements for the corps-de-ballet, from the origins of baroque dances to the ballets of Marius Petipa and George Balanchine. 

Excerpts of Symphony in C, by Georges Balanchine, initially called Le Palais de Cristal and created in 1947 for the Paris Opera Ballet, and of Rain, by Anne Teresa De Keersmaeker, which entered the repertory of the Paris Opera Ballet in 2011.

THE INNOVATION OF THIS PROJECT

The computational approach to choreography that we are developing enables us to generate complex group movements, which would be nearly impossible to achieve using traditional techniques.

We are effectively building a library of hundreds — and soon thousands — of unique ways in which dancers can move through space in relationship to each other.

Once our algorithm complete, we will be able to choose from this immense repertory of translational patterns to compose highly elaborate choreographies. 

At the bottom of this page, you can see an example of a category of translational patterns — generated from one simple mathematical rule — that we have used to compose the piece that we have created for the 36 students of P.A.R.T.S.

We believe this technological innovation will open a whole new realm of possibilities for the choreography of large ensembles and we want to propose to the Opéra de Paris to be the first ballet company for whom we create a ballet using this technology. 

OUR PROPOSAL TO THE OPERA 

In 2024-2025, we will keep on developing the algorithm and bring it to the highest level of completion and quality we possibly can.

In 2025-2026, once the algorithm ready, we will use it  to generate a series of short choreographies which we will perform in contemporary dance and contemporary art venues with a group of contemporary dancers we hire. This will constitute the first real test of the capabilities of our algorithm. 

In addition to performing these choreographies with our “own company”, we will algorithmically generate a video of each of these and mint them as NFT’s on the Ethereum blockchain. These will be auctioned at the launch of the project. This will constitute a test of the business model we have imagined.

In 2026-2027, we want to collaborate with a renowned ballet institution to make an evening-long piece for their company. We think the Opéra de Paris would be the perfect partner for whom to create a first large scale piece using this ground-breaking technology. 

As we want this project to not only be viable for the Opéra, but also profitable, we have imagined the following.

• The evening-long piece we would create for the dancers of the ballet would be fractionalized in a series of short choreographic moments — lasting between a few seconds and one minute.

• Prior to the premiere of the work, we would produce a digital artefact representing each of these choreographic moments. The exact form of this digital representation remains to be defined : it could be a video, a score, or an animated score. 

• At the premiere of the piece, we would mint these digital artworks as NFTs and auction them. The auction could, for example, take place at the occasion of a gala organised by the Opéra in which our piece premieres.

In the spirit of helping the Opéra raise funds for its activities, we propose dividing the proceeds equally between the Opéra and the artists — 50% to the Opéra and 50% to Cassiel and Tyler.

Over the last couple of years, Tyler has sold dozens of works for millions of dollars and has built a wide collector base.

We also firmly believe that the groundbreaking nature of this project will spark significant interest from both the digital and traditional art worlds, ensuring substantial financial success.

PROCESS EFFICIENCY & CREATION TIME 

Finally, the method we are developing will allow us to work extremely efficiently with the dancers of the ballet and to create a highly complex piece in a record time.

For the creation with the P.A.R.T.S. students, we generated the entire piece in advance and taught it to the dancers in less than 5 days. 

To achieve this, we designed individual scores which gave to each dancer the exact set of instructions they had to perform. 

Also, an important feature of the algorithm that we are engineering is that all patterns it produces, while looking very complex, are actually fairly simple to execute from the perspective of the individual dancers.

We believe that with a refined process and dancers as skilled as those of the Opéra de Paris, we will be able to create a highly complex hour-long piece in a very short amount of time — possibly close to a month.

Here is an example of the score we created for one of the 36 P.A.R.T.S. students. 

IN SUMMARY

We believe the systematic and meticulous approach we are taking to bringing the potential of generative choreography to fruition will open the door to unprecedented aesthetic possibilities and we want to propose to the Ballet de l'Opéra de Paris to collaborate on this historical premiere.

We would be extremely pleased to have the occasion to meet you to discuss further a possible collaboration.

We thank you for your attention and look forward to hearing from you!

Best greetings,

Cassiel Gaube & Tyler Hobbs

A sketch from the process of designing the category of translational patterns we used to compose the piece for the P.A.R.T.S. students.