Artificial intelligence systems are now able to translate a large number of different languages, bringing the world closer to its old dream of removing language barriers among the inhabitants of the planet.
In the details, Meta, the parent company of Facebook, announced Wednesday that its artificial intelligence systems in this field are now able to fully translate 200 languages interchangeably, whatever the combination, after it was limited to 100 languages.
This announcement is an example of the race for automatic translation between the major Internet groups seeking to provide their services and products to the entire population of the world.
The artificial intelligence systems developed by Google, Microsoft and Meta are now able to translate languages for which there is little parallel data, which means that they are rarely translated into other languages.
It allows the translation of texts in languages with limited reach on the planet, such as Quechua, which is largely confined to the Eblero, and the language of the Fulani people, in West Africa, although no human has ever been involved in this task.
Francois Yvonne, a researcher at the National Center for Scientific Research in France, who specializes in linguistics, says that automatic translation is "particularly important for Facebook, which has to monitor hate messages" that are spreading all over the world and in all languages.
But the reliability of these tools remains to be determined. In this regard, Meta notes that its new system can provide performance "44% better compared to the previous model, which was able to translate 100 languages."
In some African and Indian languages, this improvement is more than 70 percent compared to the recently used translation systems, according to the world's leading social network.
But François Yvonne believes that the quality of automatic translations provided by Google or Facebook engines will undoubtedly remain unequal depending on the languages, and says: "Languages that are highly translated, such as European languages, will probably retain the preference."
In turn, the CEO of the leading French machine translation company, Systran, Vincent Godard, points out that the technology used by this group was initially the same as that approved by Meta and Google, but was enhanced by the work of real linguists to avoid errors.