If there is one domain where generative AI models are very good already is being a translator from one language to another, without excelling to the point of understanding the subtleties of a language all the time. But for normal usage, they can be decent translators, much better than something like Google Translate.
If that is the case, one could wonder, is it still important to learn foreign languages anymore, maybe except English, as the most widely spoken language in the world, both natively and non-natively?
My answer to this is yes. There are many reasons why someone would want to learn another language. If you move to a different country or work with colleagues from a different part of the world where a different language is native, it's much better to know it than not. You can't really go with your "AI assistant" on the street or in meetings to translate for you other than at the beginning. At some point, you'd rather want to understand things yourself and not through a 3rd party. Plus, imagine the AI translating a joke for you. And you laughing 2 minutes after everyone else, providing the AI got the joke and translated it correctly.
Plus, understanding a language is a way to better understand a culture. Of course, that's quite a few steps down the road.
The parallel can also be made between languages and programming languages, which are effectively very similar, that's why generative AI models are good with both of them.
I listened to a round table or something with 3 CEOs from the AI world at some point not so long ago, and they were asked at the end if young people should be guided to learn programming languages given the advancements in the AI field and here's what they said (I'm paraphrasing):
This morning I had this idea... Why don't I learn a new foreign language? My first choice was Spanish, a language widely spoken in the Western part of the world (Americas, in particular, plus Spain), which I understand to some degree since it has Latin roots like my native language. But I decided to try something completely different for a change.
Mandarin Chinese is spoken by over 1.1bn people, mostly native, which makes it the 2nd most spoken language on Earth after English, and the first one spoken natively. And with the explosive development in China, I don't think this is a bad choice of a dialect to understand, at least at a basic level. We'll see about advanced learning...
Here's a glimpse into what I learned so far...
Mid-1950s China introduced a sort of phonetic system using Roman letters for the Mandarin Chinese, to be easier learned by Westerners. It is called the Pinyin system.
In Mandarin, each character corresponds to a syllable, and each syllable (word) is formed this way:
Syllable = Initial (consonant) + Final (vowel) + Tone (of voice)
So far, the most interesting thing I found out compared to the languages I know to some degree or have some idea about is that what they call "tones" with which you pronounce syllables determine the meaning of some words.
Hope the AI knows what it's doing... I can only recognize the Mandarin character for "I" so far, and maybe the last one could be "name is", but can't figure out what the one in the middle means yet...
For example, in Pinyin, the Syllable formed by the initial "M" and final "a" (so one consonant and one vowel), can mean 5 different things, depending on the tone you use when pronouncing it:
However, in writing, the ambiguity is avoided, from what I could observe.
Maybe that's why many people say Chinese is difficult, other than the writing that may seem a mystery to us using the Roman alphabet. On the other hand, I heard that, in fact, Mandarin is quite easy to learn. Particularly because it has a much simpler grammar than other Western languages. We'll see...
