Universal Translator

Our thoughts on the impending technologies that will bring us universal translation.

Category:

News & Politics

Tags:

• Google
• Translation
• Collaboration

[wbuist]

There's an interesting iPhone App that does elements of this - WorldLens

[wbuist]

There's a massive difference in two people moving from no understanding to having the means of general communications thorugh an imperfect translation. That's what's coming first. Nuance and understanding subtle meaning just isn't that important when there is no understanding at all. The question is not about perfection, but about whether a service that converted speech in one language to speech in another at minimal cost would have a big usage, even if it was only, say, 70% accurate. It would.

[jmhowitt]

Time flies like an arrow, fruit flies like a banana. Go on statistically anlyse that into Japanese. 20 or even 30 years ago there were going to be "thinking machines" AI and Universal translators. The are not. In 30 years time another William B who knows nothing about the underlying problems associated with both computers and language will be making the same optimistic, but essentially idiotic prediction based on a few Tomorrows World like programs on the Discovery Channel.

[wbuist]

@jmhowitt Time will of course tell.

I'm not sure on what basis you make the statements you do about my knowledge of language, translation and computing, or why you feel you need to. Wouldn't it be better to make your own case than make an ad hominem one?.

[jmhowitt]

Machine translation at it's best is 70% accurate. Even where a matching corpus of documentation exists. After that it drops off. The only people who believe it will ever work are the people like William who do not have a clue about language or about computers and believe the publicity material from like minded people selling snake oil.

[wbuist]

@jmhowitt I think the point here is not where we are - but where we will be.

Dr Hans Rosling, who is a Swedish professor of international health and an internationally known statistician presented the Joy of Stats programme recently on BBC 4 talked about the improving use of statistical bases for things like translation, and how this is dramatically changing the effectiveness and reach of translation by machine.

[ArtsWalesUK]

I find this concept fascinating! I do not agree with suggestions that there is no possibility of this being achievable in the very near future. Having watched some technologies develop at an exponential rate, nothing will surprise me about what might surprise me.

TED TV is a fantastic source of information about phenomenal discoveries, designs, ideas, and incredible possibilities and as far as I am concerned the more that we remain open to these changes the more likely they will exist for real

[UncleGizmo]

@nemisis1963

You are correct in observing that the video is “unprofessional”.

Your observation implies that you expect the content to be produced to a certain standard; however you obviously have little idea about the costs involved in producing professional quality video. It is clear to me that a video of this nature does not merit such extravagance.

[jmhowitt]

An ad hominem attack would be an attack on you yourself whereas i am only talking about the comments that you make in the video, so scratch that one. In the second place why am i commenting? Well you have placed yourself in the public space with a set of bald assertions and pious wishes without a shred of either evidence or analysis to support your arguments. So i felt it was right to respond to it. At the Ecademy you are used to not being contradicted and having the fanboy support but not here

[jmhowitt]

ANd remember the Discovery channel also makes programs endorsing the Da Vinci Code and the predictions of Nostrodamus.

[entrepologist]

Why for he having collar like the John Travolters?