View 2001 : Growth in a Shrinking World

J

Thread Starter

Jim Pinto

Automation Listers : In the blink of an eye we have fast-forwarded to the brink of a New Year, a new century and a new millennium. In this new age, the people who win will be, not scared, but excited about the technology tools that emerge and the new effectiveness and opportunities for growth that this will bring in a shrinking world. My new article : Growth in a Shrinking World was published in the ISA magazine InTech, December 2000 See the InTech article at : http://www.isa.org/journals/intech/opinion/1,1163,281,00.html The full-length article was published in Industrial Controls Intelligence & Plant Systems Report, January 2001. It is on the web at : http://www.jimpinto.com/writings/view2001.html I'll appreciate your comments & feedback. Happy New Year ! Cheers: jim ----------/ Jim Pinto email : [email protected] web: www.JimPinto.com San Diego, CA., USA ----------/
 
F

Fred A. Putnam

Jim, You asked for feedback. You veered off into the blue sky imho with the prediction of effective language translation in every PDA. This was predicted as doable far back as Marvin Minsky's early days at the MIT AI lab. Marvin was so wrong about this, and I think you are too. Human language embodies all the subtlety of our humanity and culture, and to think you can cram all that into a PDA is imho ridiculous. I think when you go out on a long limb like this, you damage your credibility with your readers. Fondly, Fred 1/9 -- Fred A. Putnam LABTECH, Suite B09, 2 Dundee Park, Andover, MA, 01810 tel:978-470-0099 x231 cell phone:978-375-3115 FAX:978-470-3338 pager: e-mail page message to [email protected] e-mail: [email protected] web: labtech.com
 
Fred Putnam commented on my article, which is on the web at : http://www.jimpinto.com/writings/view2001.html: >You veered off into the blue sky imho with the prediction >of effective language translation in every PDA. This was predicted as doable far >back as Marvin Minsky's early days at the MIT AI lab. Marvin was so wrong about >this, and I think you are too. Human language embodies all the subtelty of our >humanity and culture, and to think you can cram all that into a PDA is imho >ridiculous. Jim Pinto responds : Yes, I know this seems far-fetched. Indeed, Marvin Minsky and others predicted intelligent robots and a lot of other things which have not come true yet. It will indeed happen - but they will probably be a couple of decades off..... With the advance of Moore's Law (doubling of computer power) an MP3 player, a digital camera and a PDA have more processing power and memory than a PC of a couple of decades ago - my camera has 64 megabytes of memory, my MP3 player 32 mb. My prediction for speech processing stands : Within a decade (or two) a PDA will have effective language translation. Mind you, sometimes even people who speak the same language do not communicate..... Fred warns : >I think when you go out on a long limb like this, you damage your >credibility with your readers. Frankly, I enjoy "going out on a limb"...... Cheers: jim ----------/ Jim Pinto email : [email protected] web: www.JimPinto.com San Diego, CA., USA ----------/
 
I may be following Jim into the blue sky, but things have changed significantly since Marvin Minsk's start at MIT. Now a group Carnegie Mellon has taken on the far more ambitious tasks than simple language translation. The last few years have seen significant advances in Natural Language Processing (aka speech recognition) making a Star Trek style Universal Translator not that far fetched. Carnegie Mellon's Diplomat project (http://www.lti.cs.cmu.edu/Research/Diplomat/) boasts having already deployed a Serbo-Croation <-> English spoken word translator, and I recall reading an article about the same project working with English <-> German To me, this language processing research feels remarkably akin to the user interface research from PARC in the 60's and 70's. In my eyes, Jim was conservative to estimate effective language translation would be a function of a PDA in 10 years. Personally, I expect we'll see it sooner - perhaps much sooner. While a PDA might not have the computing horsepower today, it will when your pocket is filled with the equivalent of a 500 Mhz Pentium, 256 meg of RAM, and a stylus. Yours, Jeff
 
B
Hi Fred: I will draw your attention to two instances- Ninety years ago the leading scientists of the day stated and quite clearly believed that they had one more fact to establish then all physics would then be an open book with no more solutions to be sought. Then along came Neils Bohr. When the personal computer was first brought out a top IBM executive thought it was just a passing fad. Never say never about anything. Bob Pawley
 
W
I have to agree with Fred. Even IBM's much ballyhooed new translation engine leaves a great deal to be desired. I spent some of my misspent youth transliterating technical documents from "English written by a Japanese Non-Native Speaker" into "American English." It was hard, hard work and I am still proud of the simple, economical speech I was able to produce. I don't believe we have either the computing power or the algorythmic expertise to be able to do effective translation in a common and garden-variety PC, let alone in any sort of PDA. Just take a look at the state of the web-based language translators and dictionaries. They are plumb pitiful. A friend of mine asked me for a simple translation of the title, "The Black-headed Fox Woman" into Italian. After consulting dictionaries, and web translators, as well as my own knowledge of the language I gave her six possible phrases, and asked Pino Zani (the first non-North American to be President-Elect/Secretary of ISA, for those of you who don't know) for a translation from a native speaker. He and his wife, Carla, are still thinking about it. Technical material is both simpler and harder, depending on the prevalence of "loan words" that relate to the matter under discussion. For example, it is easy to translate "the pilot took the elevator" into Japanese, because both "pilot" and "elevator" were taken directly into Japanese as loan words ("pairotu" and "erebeta," expressed in romaji). We can all think of other sentences that aren't so easy to translate. Translation is actually "transliteration." Best, Walt Boyes --------------------------------------------- Walt Boyes -- MarketingPractice Consultants [email protected] 21118 SE 278th Place - Maple Valley, WA 98038 253-709-5046 cell 425-432-8262 home office fax:801-749-7142 ICQ: 59435534
 
Regarding the prediction that language translation would be part of a PDA within about a decade or so, in my article : "Growth in a Shrinking World" : http://www.jimpinto.com/writings/view2001.html Walt Boyes [[email protected]] supported Fred Putnam's contention that I was straining credibility : >I don't believe we have either the computing power or the >algorythmic expertise to be able to do effective translation >in a common and garden-variety PC, let alone in any sort of PDA. >Just take a look at the state of the web-based language >translators and dictionaries. They are plumb pitiful. "Jeff Dean" <[email protected]> supported my prediction, and even called it conservative : >In my eyes, Jim was conservative to estimate effective language >translation would be a function of a PDA in 10 years. Personally, >I expect we'll see it sooner - perhaps much sooner. While a PDA >might not have the computing horsepower today, it will when >your pocket is filled with the equivalent of a 500 Mhz Pentium, >256 meg of RAM, and a stylus. Jim Pinto applauds appreciatively : Thank you, Jeff! Language translation - both text and audio - are part of significant practical work that is going on in the field of NLP - natural language processing. Indeed, I am reviewing business plans - with practical and significant revenue projections - from at least 2 or 3 such companies. Please don't expect a PDA-language-translator to be perfect. However, you will be amazed at some of the practical and real-life demonstations I have already seen! The past is full of predictions that did not come true - but then DID, when the technology and the time was ripe. This is a new decade, century and millennium! Keep your mind open to the possibilities! Cheers: jim ----------/ Jim Pinto email : [email protected] web: www.JimPinto.com San Diego, CA., USA ----------/
 
R

Ralph Mackiewicz

> > You veered off into the blue sky imho with the prediction of > > effective language translation in every PDA. This was > > predicted as doable far back as Marvin Minsky's early days at > > the MIT AI lab. Marvin was so wrong about this, and I think > > you are too. Human language embodies all the subtelty of our > > humanity and culture, and to think you can cram all that into a > > PDA is imho ridiculous. > > Yes, I know this seems far-fetched. Indeed, Marvin Minsky and > others predicted intelligent robots and a lot of other things > which have not come true yet. It will indeed happen - but they > will probably be a couple of decades off..... > My prediction for speech processing stands : Within a decade (or > two) a PDA will have effective language translation. Speech processing in a PDA is one of the better applications of speech/language technology that exist. There actually might be a market for it. Many of the predictions about speech/language processing were proved false because the applications that were envisioned at the time were insufficient to justify the advances (it was a false need). I remember claims from years ago from the emerging vendors of the time that speech input was going to quickly obsolete the keyboard. What nonsense. > Fred warns : > > > I think when you go out on a long limb like this, you damage > > your credibility with your readers. > > Frankly, I enjoy "going out on a limb"...... You can see a lot more from the limb than you can from the trunk. Uh-oh. Who's that guy with the saw!? Besides, the credibility of prognosticators is judged by how interesting their predictions are, not by how accurate they are. Anybody remember Paul Ehrlich? According to him we were all going to be dead from starvation by now. He received a MacArthur Genius award a few years ago. Regards, Ralph Mackiewicz SISCO, Inc.
 
V

Vladimir E. Zyubin

Hello List, A couple of words to support Fred's opinion. Bob, it seems to me to use a parallel is not quite correct in the case. There is a matter of facts that some problems have no solution. It's pitty, but all we can do is to show our humility. And it seems to me that AI problem (the problem IMO we speack about) is among those problems. Unfortunately, I can't agree with Jim P.'s statement that the problem is a problem of quantity of memory (RAM, SRAM, DRAM, EtcRAM :). It is, i'm sorry, a bit simplified point of view. Moreover, it is absolutely wrong statement, these two things aren't connected at all. The problem of AI is _sintactical_ nature of data for computer. I personally am not aware of the solution. Please point the source when the solution is given. Thanx a lot in advance and best wishes, Vlad Z. Ref. Wednesday, January 10, 2001, 11:31:42 PM, Jim Pinto <[email protected]> wrote: "...With the advance of Moore's Law (doubling of computer power) an MP3 player, a digital camera and a PDA have more processing power and memory than a PC of a couple of decades ago - my camera has 64 megabytes of memory, my MP3 player 32 mb. " P.S. "LMA> Never say never about anything." I dare to violate your advise: Computer (in its present form) never understand the salt of the statement... independently how many Mb/Gb it has. ;-)
 
Top