re: [gutvol-d] google and the translation thing
michael said:
Methodologies are continually being upset by those who find some other way to do an expensive [time or money] function for an infinitessimal amount of the original.
well, i think we're both on the same page... except you're reading it and i'm writing it... :+) (in other words, if one is interested in the actual upsetting of methodologies, then one pays some attention to them. otherwise, one waits until they play out.) google is upsetting the methodologies here. and you are counting that machine translation will become up to snuff sooner or later, and you're not interesting in the interim period. both positions are equally fine to hold...
If not, then why talk in such generalities
i don't think i'm talking about "generalities" at all. in the current case, google is the entity doing it -- or so it has been reported, whether true or not -- and keith is the entity saying "it can't be done"... (well, he said "not in the next 100 or so years".) and in the other case i've mentioned -- the "dispute" between me and my "detractors" on this listserve -- there are no "generalities" either. we spent 2 years going back and forth at each other, so the positions are well-staked-out in the archives if you're curious.
It only matters when you get to the point of ending the debating and actually doing something the outside world can see and work with.
right. except when you're playing poker, the object is to win as much money as possible with the hands you win, and to lose as little as possible with the ones that you lose. and that means you don't always show all of your cards right away... google ain't showing all their cards. and i ain't showing all mine either... but i'm starting to show _some_. so we're past the point where any of this matters any more, in the matter of me vs. this list, since pudding is being served... -bowerbird
On Sun, 12 Mar 2006 Bowerbird@aol.com wrote:
michael said:
Methodologies are continually being upset by those who find some other way to do an expensive [time or money] function for an infinitessimal amount of the original.
well, i think we're both on the same page... except you're reading it and i'm writing it... :+)
If you only wrote as much and as well as you think, we might be in a much better place.
(in other words, if one is interested in the actual upsetting of methodologies, then one pays some attention to them. otherwise, one waits until they play out.)
No, one need not know the methodologies in use to come up with better ones. Only the reverse engineering types rely on this sort of logic. Innovation, when it comes, usually comes from a source well outside current methodologies, the rest is just incrementalism.
google is upsetting the methodologies here.
I guess you weren't aware of Golden Bow and the others who preceded them. I've been talking about Machine Translation since before there even was a Google. Don't you remember? One of the most difficult things about speaking with you is your apparent lack of memory/attn. This is what is most likely to get you placed on the spam list.
and you are counting that machine translation will become up to snuff sooner or later, and you're not interesting in the interim period. both positions are equally fine to hold...
Again you seem to have not paid attention. . . . _I_ have been promoting the interim phases as good enough for people to work with, while YOU have been saying that only perfection is enough, or close to it. Again, you are just asking to be ignored by any of the people who actually TRY to follow your words.
If not, then why talk in such generalities
i don't think i'm talking about "generalities" at all.
in the current case, google is the entity doing it -- or so it has been reported, whether true or not -- and keith is the entity saying "it can't be done"... (well, he said "not in the next 100 or so years".)
I'm sticking with my original prediction: By the time we have put a sigifican dent in public domain books that are available, 10-20 million eBooks, the next big thing after OCR will be MT. . .AND. . .this will all start to take place in the public eye by 2020. Just in case you try to misinterpret that. . .14 years.
and in the other case i've mentioned -- the "dispute" between me and my "detractors" on this listserve -- there are no "generalities" either. we spent 2 years going back and forth at each other, so the positions are well-staked-out in the archives if you're curious.
Sadly to say, I have read the vast majority of your archived messages, and have no desire to again. Obviously not even YOU think they are worth quoting, or you would have.
It only matters when you get to the point of ending the debating and actually doing something the outside world can see and work with.
right.
If only you said what you meant, and meant what you said. Back to Alice.
except when you're playing poker, the object is to win as much money as possible with the hands you win, and to lose as little as possible with the ones that you lose.
This is NOT a GAME, and MONEY is NOT the OBJECT. Again I refer you to Jon Noring, you have more in common than you would like to think.
and that means you don't always show all of your cards right away...
As above, stop PLAYING, start WORKING. Remember your physics lessons? It's not WORK if you don't MOVE something and then KEEP IT THERE.
google ain't showing all their cards.
Sadly to say, I expected more of you than of Google.
and i ain't showing all mine either...
Sadly to say. . . . No one can see or build on your work. You might have been a giant for someone to stand on. "If I have seen further, it is because I have stood on the shoulders of giants." Newton
but i'm starting to show _some_.
Sorry, strip-tease is not acceptable.
so we're past the point where any of this matters any more, in the matter of me vs. this list, since pudding is being served...
And that is why you will likkely continue to be ignored.
-bowerbird
mh
Hi Again, I will come back in. Am 12.03.2006 um 10:45 schrieb Bowerbird@aol.com:
michael said:
Methodologies are continually being upset by those who find some other way to do an expensive [time or money] function for an infinitessimal amount of the original. My point was that it IS NOT A NEW METHOLOGY OR NEW METHOD!!
well, i think we're both on the same page... except you're reading it and i'm writing it... :+)
(in other words, if one is interested in the actual upsetting of methodologies, then one pays some attention to them. otherwise, one waits until they play out.)
google is upsetting the methodologies here. and you are counting that machine translation will become up to snuff sooner or later, and you're not interesting in the interim period. both positions are equally fine to hold...
Actually, there is a very excellent transltion system out there already. SYSTRANS. But, what is availibable to the public you can forget. It uses grammar models, lexica and a lot more vodoo. They claim 95-99% out of the box. What is its draw back. It needs a hell of a lot of computing power. Even works with voice. Do not ask me what it costs either.
If not, then why talk in such generalities
i don't think i'm talking about "generalities" at all.
in the current case, google is the entity doing it -- or so it has been reported, whether true or not -- and keith is the entity saying "it can't be done"... (well, he said "not in the next 100 or so years".)
As I have mentioned before the method is not new. It will give you acceptable results for the average joe. It will not work for PG.
and in the other case i've mentioned -- the "dispute" between me and my "detractors" on this listserve -- there are no "generalities" either. we spent 2 years going back and forth at each other, so the positions are well-staked-out in the archives if you're curious.
It only matters when you get to the point of ending the debating and actually doing something the outside world can see and work with.
right.
except when you're playing poker, the object is to win as much money as possible with the hands you win, and to lose as little as possible with the ones that you lose.
and that means you don't always show all of your cards right away...
google ain't showing all their cards.
and i ain't showing all mine either...
This reminds me of my first semester in CL. Where the great inovators said my method is better than yours. I can do this you can not. You can do that, but I can do this. Na na nah nah! But, as Micheal said, we shall see if google will revelutionize the world of MT. I doubt it very much. Of course I could ask for my money back from the unversity.
participants (3)
-
Bowerbird＠aol.com -
Keith J. Schultz -
Michael Hart