Moli\u00e8re is to the French what Shakespeare is to the English: the last word in historical literature, drama, wit and satire.<\/p>\n
Now, more than 350 years after his death, the 17th century dramatist has been revived after scholars at the Sorbonne University in Paris used artificial intelligence<\/a> to help write an experimental play in his style.<\/p>\n L\u2019Astrologue ou les Faux Pr\u00e9sages (The Astrologer, or False Omens), a three-act comedy, made its debut at the Royal Opera<\/a> at the Ch\u00e2teau de Versailles last week.<\/p>\n The two-hour play tells the story of a wealthy bourgeois Parisian who, under the instruction of a charlatan astrologer called Pseudoramus, insists his daughter Lucile marry a debt-ridden and elderly wigmaker.<\/p>\n While the theme could well have been dreamed up by Moli\u00e8re, the dialogue, music, costumes and scenery were all created with the help of a French AI tool called Le Chat (The Cat).<\/p>\n A group of researchers at the Sorbonne worked on the project, called Moli\u00e8re Ex Machina, for two and a half years. The team included a three-person group of artists and researchers called Obvious.<\/p>\n One critic described the AI imitation as \u2018striking, almost disconcerting\u2019 and said the dialogue was \u2018entirely believable\u2019. Photograph: St\u00e9phanie Lecocq\/Reuters<\/p>\n The production involved what they described as \u201cintellectual ping pong\u201d of about 20,000 exchanges between researchers, classical literature scholars, linguists, historians and Le Chat. As the team fed more information into the AI assistant, each word and scene it came up with went through numerous rewrites as the researchers explained to the AI assistant why certain passages did not work and asked it to try again.<\/p>\n \u201cThe process was long and demanding,\u201d said the play\u2019s director, Micka\u00ebl Bouffard, the head of the Th\u00e9\u00e2tre Moli\u00e8re Sorbonne. He added that Le Chat\u2019s first draft ran to only eight pages that were \u201cnot very interesting\u201d and as a result \u201cscenes had to be revised multiple times\u201d.<\/p>\n \u201cAI has a superpower: the ability to store everything Moli\u00e8re wrote and everything Moli\u00e8re read,\u201d Bouffard told France<\/a> Info. \u201cWe human beings can\u2019t do that.\u201d<\/p>\n The theme of astrology, which featured in at least one original Moli\u00e8re play, and the play\u2019s title were suggested by the AI and touched on current preoccupations regarding the use of the technology<\/a>. \u201cAstrology allows us to discuss manipulation, false beliefs and disinformation, which are particularly topical subjects,\u201d said Pierre-Marie Chauvin, an associate professor at the Sorbonne.<\/p>\n There are plans to perform the play across France and to take it abroad. Photograph: St\u00e9phanie Lecocq\/Reuters<\/p>\n Moli\u00e8re, who died in 1673, was so influential that French is often referred to as \u201cthe language of Moli\u00e8re\u201d.<\/p>\n AI remains one of the most sensitive issues in the entertainment industry and has generated intense debate. Using it to imitate Moli\u00e8re would have caused outrage in France were the project not being carried out by academic experts at the Th\u00e9\u00e2tre Moli\u00e8re, which specialises in accurately reconstructing 17th-century productions.<\/p>\n A report submitted<\/a> to the national assembly last year suggested generative AI was a \u201cmarvellous opportunity, a stimulating tool and a powerful driver of creativity\u201d. But it also said AI \u201cposes a threat to many professions in the cultural sector because it enables the production of content that may compete directly with human creations\u201d, adding: \u201cIt is necessary to strike a balance between different forms of creation.\u201d<\/p>\n Chauvin said L\u2019Astrologue had struck that balance. \u201cWe are demonstrating in concrete terms something that can be achieved in a novel way with AI. Not a play written by AI, but a play co-written with it,\u201d he said.<\/p>\n An audience of 100 people, including Catherine P\u00e9gard, the culture minister, saw the play during two performances last week. Afterwards, one audience member said: \u201cI think it\u2019s a success. The plot feels so real, the subject matter is so close to what we\u2019re used to hearing in these [Moli\u00e8re] plays.\u201d<\/p>\n Another theatre-goer was less impressed. \u201cA decent writer can do this without artificial intelligence,\u201d he said. \u201cI think we [humans] still have a bright future.\u201d<\/p>\n Christophe S\u00e9frin, the technology editor for 20 Minutes, attended one of the two performances. He described<\/a> the AI imitation as \u201cstriking, almost disconcerting\u201d and said the dialogue was \u201centirely believable\u201d.<\/p>\n