Dear translators, AI is ready to send you into retirement. Report Ft
The global AI-based translation market will grow from approximately $5 billion in 2023 to $12 billion by 2030 as the translation industry “serves every industry under the sun” and competition becomes fierce. The Financial Times article
Translation is an ancient art, perhaps as old as civilization itself. Some of the earliest surviving human texts, the clay tablets from Mesopotamia, include lists of equivalent words in several languages. The Financial Times writes.
Ancient roots, however, are no barrier to innovation, and the rise of “generative” artificial intelligence (AI), such as OpenAI's ChatGPT, has sparked a surge in translation technology. Data provider Statista predicts the global market for AI-based translation (particularly “natural language processing,” or NLP) will grow from about $5 billion in 2023 to about $12 billion by 2030 .
EVERYONE (OR ALMOST) NEEDS A TRANSLATOR
One company hoping to profit is Germany's DeepL, which was named winner in the IT and cybersecurity category of the FT 's Tech Champions 2023 survey. But there is no shortage of competitors, including not only other start-ups but also tech giants like Google and Meta, owner of Facebook, and enterprise software companies like Cisco and Microsoft.
“Everyone is in this together,” says Mark Beccue, research director at The Futurum Group, a technology research and consultancy firm. “Google has been very interested in [language translation] for a long time.”
The amount of content that companies and other organizations produce means that demand remains high. The translation industry “serves every industry under the sun, because every industry needs language translation,” says Florian Faes, CEO of Slator, a provider of language translation news and research.
Advances in artificial intelligence offer the possibility of saving money and time, with the intervention of human translators only when necessary, or according to the indications of quality prediction software, such as that created by the US start-up ModelFront. According to a study published last year by technology research firm Gartner, by 2025, 25% of human translators' work will consist of reviewing and editing machine-translated texts.
THE RISE OF DEEPL
DeepL has grown rapidly since it was launched in 2017 by Jaroslaw Kutylowski, a Polish computer scientist who is also the company's CEO.
Previously, Kutylowski was chief technology officer of predecessor company Linguee, a multilingual dictionary compiled through computer analysis of bilingual texts found online. He led a team that used Linguee's language data to help create what would become DeepL, an online translator based on neural networks (a type of AI system modeled after the human brain).
Today DeepL, as the privately held company has renamed itself, offers free and paid monthly subscription services and covers about 30 languages. It says it has more than 20,000 commercial customers, including publisher Elsevier and technology firm Fujitsu.
While it does not disclose its revenue, it announced in January that it had secured a $1 billion valuation from investors, giving it so-called “unicorn” status. “DeepL is probably the most successful technology company in the translation industry of the last 15 years,” says Faes.
The company claims to produce the best machine translation software in the world. This claim is based on blind tests, in which professional translators select the most accurate rendering of a text without knowing which company produced it.
“It all comes down to having the best technology,” Kutylowski says. The company's neural networks can analyze large amounts of data collected from the Internet to learn the most likely translation for each sentence.
NEURAL NETWORKS
But this also applies to DeepL's rivals, which also use neural networks. When asked what gives DeepL an edge, a company representative simply says it's a combination of the "architecture" of its neural networks, its training data and the human input behind its data models .
Linguee's original data was supplemented by drawing on other sources, such as company websites and social media, which, according to Kutyłowski, helps DeepL's technology master more casual idioms.
More nuanced translations will no doubt be well received by DeepL's customers. But will the company be able to maintain its growth in the face of competition from much larger tech companies?
NOT EVEN GOOGLE SCARES DEEPL
Kutylowski doesn't seem worried. DeepL will continue to focus on “academic-level” research in areas such as how neural networks should be “wired” for translation and how to obtain the best training data.
“We have been competing with Google Translate and therefore with Google since the advent of [our] company,” he says. “The landscape hasn't changed much for us.”
However, he acknowledges that the pace of change has increased recently. “There is a lot of development in AI and the speed of research has increased,” he says. “This obviously represents a challenge for our teams, who have to accelerate a little.”
DeepL has recently started to exit its core market. In January it released an AI writing tool that “improves written communication” in English and German. It is also looking to expand beyond its European base, especially into Asia and the United States.
One source of competition is the growing consumer market for machine translation. “People want to go to Europe [on vacation] and have a pocket translator,” says Bern Elliot, vice president and AI expert at Gartner.
WHAT ZUCKERBERG IS UP TO
In August, Meta released artificial intelligence software that it says can automatically translate from about 100 languages in both voice and text. The company said the technology, SeamlessM4T, is a step toward creating a universal translator, a real-life version of the eared Babel Fish in the sci-fi classic The Hitchhiker's Guide to the Galaxy.
Meta said it will make the data behind SeamlessM4T public, including 470,000 hours of speech and text, to allow researchers and developers to build on the technology.
Despite increasing competition in the machine translation industry, DeepL remains optimistic about its future, believing that its focus on translation quality will open up new opportunities.
“When you improve the quality of translation little by little, you unlock more and more use cases for machine translation,” says Kutylowski.
(Excerpt from the foreign press review edited by eprcomunicazione )
This is a machine translation from Italian language of a post published on Start Magazine at the URL https://www.startmag.it/innovazione/cari-traduttori-lia-e-pronta-a-mandarvi-in-pensione/ on Sat, 18 Nov 2023 06:16:13 +0000.