AI Translation: How to Train the “Horses of Enlightenment” | Books

‘TTranslators are the stage horses of the Enlightenment,” wrote the poet Alexander Pushkin in the margins of one of his manuscripts. Two centuries later, political scientist Steven Weber similarly compared translation to transportation: not of people and goods but of ideas and knowledge. Just as the world has traded horses for mechanical transportation, multilingual communication has also accelerated – and now, with the use of AI tools, translation can be done faster than ever.

But faster doesn’t always mean better: using AI comes with various risks. This week, the European Parliament adopted the Artificial Intelligence Act, the world’s first comprehensive AI law. This requires developers to be transparent about the data used to train their models and to comply with European copyright law.

In the UK, the Artificial Intelligence (Regulation) Bill is expected to have its second reading in Parliament next week. Literary translators, like other artists, continue to campaign for real recognition. It is essential that we are heard.

Since the adoption of neural networks in 2015, translation algorithms have greatly improved. Academic publishers have been using AI since 2018, when Massot éditions published a French version of Deep Learning by Ian Goodfellow, Yoshua Bengio and Aaron Courville, translated from English by DeepL. Commercial publishing quickly followed suit. The next frontier for technology is literary translation.

A unique creative practice, literary translation has always involved a myriad of choices. This word or that? Fidelity or freedom? Paper or screen? Dictionary or database? The advent of AI has introduced another dilemma: should a text be translated from scratch, or can it be run through AI translation software first?

The unreadable translations of pirated books flooding the Internet show that the process cannot currently be entirely outsourced to computers. Instead, publishers have started using AI-assisted translation. Some European presses work with Nuanxed, which employs humans to edit machine-translated books. The company aims to “maintain the quality of traditional translations”, offering savings to publishers and “market rates” to linguists.

The threat of being underestimated by machines is a constant concern for translators. In his recent investigation, the European Council of Associations of Literary Translators recommends that professionals avoid editing AI-generated texts or charging translation rates for this type of work. In the age of jet-powered information, translating by hand could become the new eco-friendly travel; The appeal of sustainably translated books would then increase, which would hopefully benefit their creators.

When asked about quality, even some AI advocates think it may be good enough for a boiler, but not for a book of poetry. Kristoffer Lind, CEO of Swedish publisher Lind & Co, told me that his company uses machine translation only for genres such as crime and romance. If the relevance of AI is often debated in economic terms, for Eva Ferri, publisher of Europa Editions UK and the Italian press Edizioni E/O, the choice is ethical. “Hiring a human being is the right thing to do,” she says, “even if the alternative is much more profitable.” »

Like many of my fellow translators, I have no interest in entrusting my literary projects to AI. Mark Polizzotti, who has translated more than 50 books from French, including works by Gustave Flaubert, André Breton and Raymond Roussel, spoke for most of us when he told me: “I do this work not so much to express as to discover what I think, where it will take me. By doing this work for me, AI would remove the aspects I enjoy most, and I would prefer to move at my analog pace and come out the other end with increased knowledge of both the work and myself .

ignore previous newsletter promotion

Meanwhile, Nichola Smalley, who translates Swedish and Norwegian fiction, had three deadlines to meet the afternoon we spoke. Yet, she says, “there’s no point in going faster and faster – that’s not the point of literature. »

Sometimes, however, it’s difficult to reach places underserved by technology without speeding things up. “I can imagine scenarios where AI helps translate from languages ​​for which there aren’t many translators available,” says Ferri. However, most AI tools “won’t have enough data” to translate from or to such languages, points out Anton Hur, a translator of Korean literature. Attempts to remedy this situation include a model developed by researchers at the University of Massachusetts. Designed for anyone to contribute to its training dataset, it should “help literary translators share more minority voices.”

How do we see our relationship with technology evolving? Tim Gutteridge, who translates literary and non-literary texts from Spanish, believes that in either case, using computer-assisted translation can be a “reasonable decision”, and if the tools include AI features that give us more choice and control over results, so much the better. Edwin Frank, editorial director of the New York Review Books Classics series, calls the work of human translators “crucial,” given that advances in AI are “absolutely inevitable.”

“I have no major opposition to the use of (such) tools,” he says, “any more than to the use of screwdrivers.”

When translating, we need all the tools available. Referencing texts created by humans and processed by machines could be a new way to consult existing translations. AI could be seen as another vehicle allowing us to navigate the multilingual expanses of knowledge. But this option should not prevent us from traveling at the speed we choose.

Anna Aslanyan is a journalist and translator and author of Dancing on Ropes: Translators and the Balance of History.

Bir yanıt yazın

E-posta adresiniz yayınlanmayacak. Gerekli alanlar * ile işaretlenmişlerdir