Time to stop studying languages?

General discussion about learning languages
User avatar
Uncle Roger
Orange Belt
Posts: 154
Joined: Tue Sep 05, 2017 8:53 am
Languages: Italian (Native), English (as good as you see me write it here?), Norwegian (C1?), French (B2), Swedish (B1?)
x 193

Re: Time to stop studying languages?

Postby Uncle Roger » Tue May 22, 2018 10:54 am

I believe these technologies might work in very selected and refined contexts. No ambiental noise. Clear, straightforward speech with no funnies, no charisma, no metaphors. Basically, very functional language use and between rather close languages.

But real relationships (whether professional or personal), real engagement, real connections... those will still continue to demand actual language learning. Plus I pity the fool who'd like to deprive themselves of the brain workout that learning a language is.
3 x
«If you want to get laid, go to college. If you want an education, go to the library.»
Frank Zappa

User avatar
rdearman
Site Admin
Posts: 7231
Joined: Thu May 14, 2015 4:18 pm
Location: United Kingdom
Languages: English (N)
Language Log: viewtopic.php?f=15&t=1836
x 23120
Contact:

Re: Time to stop studying languages?

Postby rdearman » Tue May 22, 2018 11:56 am

I for one welcome our robot overlords
0 x
: 0 / 150 Read 150 books in 2024

My YouTube Channel
The Autodidactic Podcast
My Author's Newsletter

I post on this forum with mobile devices, so excuse short msgs and typos.

User avatar
lavengro
Blue Belt
Posts: 728
Joined: Wed May 24, 2017 1:39 am
Location: Hiding in Vancouver. Tell no one.
Languages: English - finally getting a handle on this beast of a language. Also tinkering with a few other languages intermittently.
x 1994

Re: Time to stop studying languages?

Postby lavengro » Wed May 23, 2018 2:19 am

lavengro wrote:It is a solid lock that well within the lifetime of most of us, technology will come to pass that will allow one to plug headphones into an iPhone or whatever magic box is in vogue at the time and have it translate adequately in real time a foreign language speaker`s jibberish.

I am so certain of this that I would bet up to a gazillion dollars in support of this occurring, if I could find a bookie who would take the opposite end of this action (my own bookie won`t, and he has already bought his Lamborghini its own Maserati off of many of my foolish bets in the past - that shows how certain he also is that this will occur, and not terribly far off either).

From a BBC article posted today, the perspectives of a couple of punks at Stanford University on the issue:

"At Stanford University, in California, Wonkyum Lee, a South Korean computer scientist, is helping to develop translation and voice-recognition technology that will be so good that when you call a customer service helpline, you won't know whether you're talking to a human or a computer.

Christopher Manning, professor of machine learning, linguistics and computer science at the same institution, insists there is no reason why, in the very near future, computer translation technology can't be as good as, or better than, human translators."


http://www.bbc.com/news/world-44200901

I've checked - these guys won't take my bet, either.

I can readily accept that computer translation technology may at some not distant point rival in quality most human translation - our pending "robot overlords" are already better (sometimes much better) at spelling, punctuation and even homonym determination than most of the humans I see posting in reader comment sections on news sights. I mean, "sites" (thanks Skynet!).
0 x
This signature space now on loan to the mysterious and enigmatic Breakmaster Cylinder:
https://www.youtube.com/watch?v=hUGKldkiex4

Cavesa
Black Belt - 4th Dan
Posts: 4960
Joined: Mon Jul 20, 2015 9:46 am
Languages: Czech (N), French (C2) English (C1), Italian (C1), Spanish, German (C1)
x 17565

Re: Time to stop studying languages?

Postby Cavesa » Wed May 23, 2018 4:02 am

rdearman wrote:I for one welcome our robot overlords


Yay! The new international language is gonna be Python!
1 x

User avatar
adamwakoski
Posts: 7
Joined: Wed May 30, 2018 7:31 am
Location: Provo
Languages: English (C1), Spanish (B1), Polish (B2)
x 13

Re: Time to stop studying languages?

Postby adamwakoski » Wed May 30, 2018 12:55 pm

Oh no) I think there may be a situation when people will use devices for everything. That will make them too lazy causing the reverse process of general intelligence decline :roll:
1 x

User avatar
picklelanguage
Yellow Belt
Posts: 59
Joined: Wed Jul 13, 2016 8:31 pm
Languages: English (N), Spanish(B1-B2), German and French (I suck)
Language Log: https://forum.language-learners.org/vie ... =15&t=7984
x 110

Re: Time to stop studying languages?

Postby picklelanguage » Mon Jun 04, 2018 12:22 am

What I think some people get held up on is that they misunderstand what these devices and language translation technology are actually good for. We as language learners don't want to admit it but (at least for major languages), the age of machine translation being more accurate than human translation is fast approaching. That doesn't mean the machines are better at the languages. That doesn't mean the machines understand the languages. There is a difference between NLU and NLP in Artificial Intelligence. NLU (Natural Language Understanding) is one of those problems we haven't even scratched the surface of. Don't expect progress on that for a long, long time. NLP (Natural Language Processing) on the other hand is something that we are getting quite good at and are not really showing signs of stopping. NLP just looks at human language as data to be manipulated. It looks for patterns and relations. A NLP model doesn't need to know what words are actually equivalent from a linguistic standpoint, it just needs to be able perform better than a human (or another model) at the act of actually translating it.

I'd also like to address the idea of what has changed recently. There are two big changes that happened. One is the idea is the pure availability of data and processing power. The amount of available data and the ability to process that data has grown at blindingly fast speeds. Plus, we have found better ways to actually run the data. To give a picture, in 2012 Google spent $5B on this super complex 1000-CPU machine for running neural network models. In 2013, after it was found GPUs were much faster for these types of calculations, Stanford researchers built a $33K machine with 3 GPUs. It ran the same neural networks just as fast. Now we are building ASICs like TPUs (basically chips specifically made JUST for neural network models) that can speed things up even further.

The second huge change was the deep learning revolution. Not only was the previously stated computation boost allowing researchers to run complex networks that were previously computationally infeasible, new architectures were found that practically overnight brought the field forward years (like Convolution Neural Networks and computer vision - if you remember the sudden influx of computer vision and image generation stuff that happened 4-5 years ago, this is a big reason why). Also, say what you will about the buzzwords and overhype of neural networks that the mainstream media has brought, we it has brought is a lot of funding for more neural network research. It's hard to explain what makes neural networks so useful, especially in fields like NLP without either going too much into the technical details (or on the other side, using language that echos the pop-science singularity stuff), but long story short, deep neural networks are specifically equipped to learn those relationships between words that I mention earlier. It doesn't need to understand those words. It just needs to be able to learn the patterns well enough to pass a test better than a human can.

I guess the point of this post is to draw a distinction. Because of the media hype around artificial intelligence, it is easy to forget that there is actually academic research going on behind the scenes. Most of what the general public sees regarding AI is stuff advertised by big companies. They are trying to sell you something. People are pretty good at sniffing that out so they just reply (rightfully) "bullshit". But as somebody who is trying to go into the field and actively reading the research, I can tell you that there is a LOT going on and some of the things researchers were saying impossible or impractical about 5 years ago are what is being published in papers today. This product is obviously bullshit. But the idea that machine translation that is better than humans is something that is far off in the future is wrong.
2 x
Spanish - Half Challenge
: 0 / 50 Books 0/50
: 0 / 50 Films 0/50
German - Half Challenge
: 0 / 50 Books 0.0/50
: 0 / 50 Films 0.2/50

User avatar
picklelanguage
Yellow Belt
Posts: 59
Joined: Wed Jul 13, 2016 8:31 pm
Languages: English (N), Spanish(B1-B2), German and French (I suck)
Language Log: https://forum.language-learners.org/vie ... =15&t=7984
x 110

Re: Time to stop studying languages?

Postby picklelanguage » Mon Jun 04, 2018 12:45 am

rdearman wrote:
Cavesa wrote:And you are right that this is one of the areas where the machines will destroy jobs.

The reason new technology has not been the cause of mass unemployment is that new kit will only be used when it makes the productive process more profitable. Higher productivity frees up the resources to buy other goods and services. The rural workers that Thomas Hardy described in Tess of the D’Urbervilles found work in factories and offices. What’s more, it was better paid work, and so the upshot was an increase in living standards.
Business Today: sign up for a morning shot of financial news
Read more

Similarly, the age of robots will lead to more jobs. Kallum Pickering, analyst with Berenberg, says there is a big hole in the argument that artificial intelligence (AI) will lead to vast numbers of workers joining the dole queue.

“Producers will only automate if doing so is profitable. For profit to occur, producers need a market to sell to in the first place. Keeping this in mind helps to highlight the critical flaw of the argument: if robots replaced all workers, thereby creating mass unemployment, to whom would the producers sell? Because demand is infinite whereas supply is scarce, the displaced workers always have the opportunity to find fresh employment to produce something that satisfies demand elsewhere.”


There are a few flaws to that argument. First, it assumes that the effects of automation are immediate. They aren't. In fact, in the short term, it tends to be quite profitable. The problem is that a business will continue to do so as long as it benefits them. Unless they are incredibly forward thinking (and able to convince investors who tend to care more about profit in the short term), they will probably overautomate. And once you automate it is hard (and risky) to go back. In fact, it is basically a prisoner's dilemma.

Second, it sort of assumes some binary option - either lack of automation or complete joblessness. That's probably not what would happen. In fact, the economy is really good at preventing that from happening. Like they said, the market is really good at creating new jobs. The question is if those will be good, well-paying jobs. Like they said, the producers still need to sell. But the producers are now producing at a fraction of the cost. That means that they can make the same profit with much, much poorer buyers.

This is why I don't like the 100% doom and gloom idea of total and complete collapse of the economy when we talk about automation. That is almost a preferable option because it would likely cause a mass restructuring of society. The reality is potentially darker.
0 x
Spanish - Half Challenge
: 0 / 50 Books 0/50
: 0 / 50 Films 0/50
German - Half Challenge
: 0 / 50 Books 0.0/50
: 0 / 50 Films 0.2/50

User avatar
tastyonions
Black Belt - 1st Dan
Posts: 1570
Joined: Sat Jul 18, 2015 5:39 pm
Location: Dallas, TX
Languages: EN (N), FR, ES, DE, IT, PT, NL, EL
x 3853

Re: Time to stop studying languages?

Postby tastyonions » Mon Jun 04, 2018 6:11 pm

Another fatal weakness in the argument is that for profits and economic growth to continue only aggregate demand needs to stay the same, which can happen with a much smaller number of people with (higher-paying) jobs buying up the supply formerly bought by a larger number of people who are now mostly unemployed.
2 x

Ccaesar
Orange Belt
Posts: 216
Joined: Fri Apr 13, 2018 2:47 pm
Languages: Danish (N) English (B2-C1) German (B1-B2) Italian(B1), Japanese (beginner)
x 264

Re: Time to stop studying languages?

Postby Ccaesar » Mon Jun 04, 2018 7:42 pm

Hmm. Easy reply: Do you think that you should stop studying them? If no, then don't.
Second: I like to think of the development being like the calculator. Just because the computer can put 2+2 together doesn't mean that it does it correctly in the needed situation, neither will it perform a task without having been "told to". We still study maths despire having caculators, so I don't think you have to worry. :)

Just think of someone using a translation machine when ordering a "lattè" it will still come out in English as them ordering milk, which is not a wrong translation, it's just not the "4" that you want "2+ 2" to be in that case.
0 x
Hiragana practiced in hand : 48 / 48
Katakana practiced in hand : 48 / 48
Kanji : 50 / 2000
Assimil Japanese with Ease : 27 / 100

Cavesa
Black Belt - 4th Dan
Posts: 4960
Joined: Mon Jul 20, 2015 9:46 am
Languages: Czech (N), French (C2) English (C1), Italian (C1), Spanish, German (C1)
x 17565

Re: Time to stop studying languages?

Postby Cavesa » Tue Jun 05, 2018 12:33 am

I've been recently thinking about one more reason, why learning languages is far superior to the automated translation.

Most essays on this subject count with a "fact", that the language barrier is the only problem in communication. And that people would communicate much more directly and easily with just a bit better google translate or something like that. Or if everyone just learnt English.

From my own experience and observation: many people use (and abuse) the fact you cannot speak their language. Things get lost in translation not only because of a general unavoidability of some loss, it is sometimes the intention. Automated translation will not stop that, I have no doubts certain kinds of people will learn how to hide some stuff on the way. It will create new opportunities for this, actually. The best way to protect yourself and your interests is learning the language you are likely to need for more than a short vacation. Firstly, for the practical reasons. You'll understand all the conversation in the room, you'll understand the original versions of contracts (which are usually the ones that are binding, following the local laws), you'll be able to ask any questions and demand full answers instead of settling for bad ones, because "you should be grateful the other person is able to somehow communicate with you". And emotion is a part of this. People simply don't dare to do some stuff, when they know you understand them directly. They don't abuse their advantage.

It seems that people believe a technological babblefish will solve this, a thing will translate for you. But even if it was technically perfect, it could never put that piece of pressure on your foreign contacts. The good old "speaking to someone in their own language means speaking to their heart" (sorry, I can't remember the exact wording, but I think we all know that quote) is not just about people having a more pleasant time together. It is exactly about this. It obliges (perhaps subconsciously) the person in front of you to treat you better, without counting on your lower ability to defend yourself.
0 x


Return to “General Language Discussion”

Who is online

Users browsing this forum: No registered users and 2 guests