sudanglo:Machine translation already shows considerable intelligence. Look at the results from GT for these sentences translated into French

I like coffee
time flies like an arrow
fruit flies don't live very long

J'aime le café
Le temps file comme une flèche
les mouches des fruits ne vivent pas très longtemps

Examples like the 3 English sentences above are often quoted to show how machine translation 'can't work'.

But the current level of development is not the point. Where will we be in 10 years time?

The point is that given a choice between spending money to solve a problem and making some intellectual effort, many people will choose the former.

I see no sign that the Esperanto movement is taking seriously the near-future advances in this field.

The whole current propaganda for Esperanto pivots on the ease of learning Esperanto compared to national languages. But it is even easier to buy a hand held automatic translator.

Incidentally the future sales of such devices will be a measure of how real the public perceive 'the language problem' to be.

If you are taking the ad infinitum case then yes, machines may eventually be good enough to merit 95% certainty in translation accuracy. When exactly this would be is completely unpredictable, though. The only ways I can see it working is if we have translated nearly everything already and have a very big and complicated database that the machine searches, or if we have build a machine that actually learns like we do and thus translates from experience. I believe the only way to get consistent results would be with a learning machine that builds an internal model (which it understands to "make sense" or not) on which to perform the translation.

The phrases in the English/French translation are likely very common (and consistent) in both languages, but when you get some idioms that may be slightly different depending on the context or locale of the speaker there will again be problems. And forget about any misspelled words (accidentally, or even worse from ignorance of the correct spelling/word to use). I would also think that XXX-English and English-XXX translations will always have much more support than, say, a Spanish-Mandarin translator.

Hell, even Microsoft Word has trouble with my (correct) English about 80% of the time.

As for the effort, we already are forced to take several years of foreign language in school, so why not change this to Esperanto? In America I'd say an overwhelming majority of students take Spanish (with French, German, and the like not offered until High School sometimes), which is very similar to Esperanto in many ways. No foreign language teacher worth his/her salt would be unable to teach Esperanto at most levels after a few months of study.

I do recall hearing somewhere that there was an effort to use Esperanto as an intermediary language for machine translation...

What will happen when really good ones are widely available is that a lot of people will buy one. And then the stories will start to come out. Most of the stories will be of a humorous "I meant THIS, but it said THAT" sort. But there will also be stories where THAT gets someone arrested, hurt, or killed. There will probably be fewer of those stories, but it won't take much to erode confidence.

A 95% accuracy rate isn't good enough. A 99% accuracy rate isn't good enough, if that 1% is still enough to get you in potentially serious trouble.

Eventually people will figure out that the translators can't be trusted. Oh, they'll work most of the time. They'll even do a pretty good job. Most of the time. But sometimes they won't when it matters. Ultimately, people will be asked to trust that the machine is saying what they meant to say, without any convenient means of checking, and sometimes it won't. I'm not sure that I will ever trust machines to that extent for anything even slightly important.

I think people will buy them. I think people will use them. And I think many people will ultimately abandon them as more of a pain than anything else. They'll be good for ordering coffee on your trip. They won't be good for having a long dinner conversation with your foreign friend.

sudanglo:The point is that given a choice between spending money to solve a problem and making some intellectual effort, many people will choose the former.

But that's no surprise, that's natural. Thinking of water, it chooses a line of the least resistance. Intellectual effort is just done if it can simplify surviving.
The human body still uses the remnant of the time when it had to save energy for bad times. And the human brain needs a lot of energy. So watching a film/movie is much more relaxing than reading a book. The same for maps and navigation devices.

sudanglo:I see no sign that the Esperanto movement is taking seriously the near-future advances in this field.

The whole current propaganda for Esperanto pivots on the ease of learning Esperanto compared to national languages. But it is even easier to buy a hand held automatic translator.

As we have learnt from the past, languages are usually spread by wars and not by an intellectual effort and minorities always had been oppressed. The fact is the human being is mostly conducted by its emotion and not by its reason.
That's why no politician would propose Esperanto as a link language because he or she fears bad consequences. And we humans usually try to avoid bad consequences.
In spite of that, the very first humans left the protecting caves, build ships and set off for a journey into the unknown. Not due to fear, but due to curiosity. It is not for nothing that the Mars rover is called "Curiosity", too. And our history proofs that the human's curiosity is bigger than its fear. So we Esperantist can just hope that once there will be a politician whose curiosity is bigger than his or her fear of bad consequences, thinking of Esperanto.

If such devices will become so accurate that they'll allow a fluent conversation between two strangers, it would already be a good step in my opinion. All the world would not struggle anymore to learn English or another language. Because of that, languages would lose their economic interests to expands themselves and this could prepare a better environment to the rise of Esperanto.

The improvements in machine translation may well reduce the interest in learning English.

The key question is what will be the effect on the desire and motivation to learn Esperanto.

sudanglo:He improvements in machine translation may well reduce the interest in learning English.

The key question is what will be the effect on the desire and motivation to learn Esperanto.

I doubt it. I say this because as long as one language holds sway over commerce - whether through the impacts of one nation (Rome) or a group of nations related by a single language (British Empire giving way to the US) - that language will - rightly or wrongly - be seen as the vehicle for self-betterment and social standing. There will still be the perception of ignorance - and the judgments that go with it - for those that have to rely on a translator to speak whatever the dominant language may be - English or otherwise.

That said, and yes, I recognize the bleakness inherent in that statement, I'm still striving to learn Esperanto because 1) I like it, 2) I think it is useful and 3) its goals match up well with my desire to be able to communicate with everyone.

sudanglo:The key question is what will be the effect on the desire and motivation to learn Esperanto.

For people who don't care about languages, it will make them happy, because they'll have a way to order coffee in Denmark without straining their little brains. But those people weren't motivated to learn another language anyway. While they might seem like potential Esperanto learners, they really aren't.

Honestly, I don't think it will have any impact at all. The people for whom Esperanto has an appeal aren't going to suddenly stop being the way they are. They will continue to learn it at the same rate as today. The people for whom Esperanto has no appeal won't be impacted because it still won't appeal to them.

It does mean that Esperanto may never become the worldwide international language for ordering coffee. But that seems mostly a concern for the finvenkistoj. The raŭmistoj probably couldn't care less.

In case it isn't evident, I fall on the Raumist side of the equation.

It's a pity yet again that the interesting topics here are not discussed in Esperanto. Anyway, I can say that already the motivation to learn Esperanto for general population is low. Machine translation will have minimal impact on that. The current motivation will go much lower, if Esperanto loses its leading position among constructed auxiliary languages – and go up, if it gets political support, or support by school system(s) in some country/countries. Or if some major celebrity announces that he will answer to fan mail personally to everybody, but only in Esperanto.

My take is that machine translation will also have only a small impact on English's position. 1. Can you imagine education systems around the world dropping English because of automatic translation? 2. Can you imagine, that TV-stations around the world will stop showing American (USA) TV-series and movies and radio-stations stop broadcasting American (USA) music? 3. Can you imagine, that videogames sold in Sweden would switch to being in machine translated Swedish instead of in English like today?

In Finland, even the state-owned enterprises (and organisations) have imposed English on their employees. You don't get or keep a job if you don't speak English!!!

sudanglo:The improvements in machine translation may well reduce the interest in learning English.

There's also the effect, that you can read not only the English texts, but also the French and German newspapers through Google Translate. So automatic machine translation makes it even more important to learn English well, so you can get the gist of another languages texts through English. If you don't speak the world's, say, top ten languages, automatic machine translation will take much more time to develop.

ludomastro:

sudanglo:He improvements in machine translation may well reduce the interest in learning English.

The key question is what will be the effect on the desire and motivation to learn Esperanto.

I doubt it. I say this because as long as one language holds sway over commerce - whether through the impacts of one nation (Rome) or a group of nations related by a single language (British Empire giving way to the US) - that language will - rightly or wrongly - be seen as the vehicle for self-betterment and social standing. There will still be the perception of ignorance - and the judgments that go with it - for those that have to rely on a translator to speak whatever the dominant language may be - English or otherwise.

That said, and yes, I recognize the bleakness inherent in that statement, I'm still striving to learn Esperanto because 1) I like it, 2) I think it is useful and 3) its goals match up well with my desire to be able to communicate with everyone.

True but I think it would be just a cultural factor. Artists who would sell more copies would still produce products in English in order to target a bigger public. But, if these devices would be that efficient, I don't get why international and theoretically "neutral" organizations as the United Nations or the European Union or you name it would not push for Esperanto at that point. English would appear clearly biased to them in that situation.