February 20, 2020
  • 2:51 pm BREAKING Obama Training Army & DHS To Kill Christians!
  • 12:51 pm Box of Toys Military Toys Toy Guns for Kids Toy Weapons Army
  • 10:51 am Who Can Keep The Soldier Alive The Longest? (PROS Vs. AMATEURS Ghost Recon)
  • 8:51 am Military Kids Get a Christmas Surprise While Dad is Away
  • 7:51 am Chinese soldiers assassinate Japanese experts to seize classified information!Jagged soldier 12
How Cold War Computers Failed At Russian – MT #2 – with CompChomp


Last time on CompChomp:
Cold War computers were showing real promise at translating Russian into English. Georgetown
University staged a successful tech demo and the press was impressed. Times were good for
machine translation. Okay, now that we’re back, the money starts
rolling in because if the public supports something, politicians can’t lose by throwing
money at it, but people started taking sides. Round 1, fight!
In this corner, you had your linguistically minded perfectionists. They wanted to feed
computers all the detailed grammatical rules about every language. And in this corner,
there were your brute force crowd. Brute forcers were all about the evidence. You set computers
loose on real sentences and they should be able to find the patterns on their own.
Unfortunately, progress from both sides just wasn’t living up to the bold claims made after
Georgetown and the public, they started to notice. In 1962, Harper’s ran an article titled
“The Trouble with Translation”, and with a title like that you know it was going to be
fair and objective. The author recalled a demonstration of machine translation where
the computer was given the scriptural nugget, “The spirit is willing but the flesh is weak.”
The output, we’re told, is laughable: the liquor is holding out all right but the meat
has spoiled. I just want to pause for a moment. The demo reported there didn’t happen, but
this story in Harper’s really illustrates how far machine translation had fallen. A
few years earlier, those exact same reporters were just heaping praise on it and now it
was the butt of their magazine jokes. The scientists stepped in for some damage
control. Do you remember Bar Hillel, our first full-time machine translation man? He argued
that our expectations were just unreasonable. Natural language could be ambiguous and computers
cannot handle ambiguity. We shouldn’t expect them to translate languages for us. We should
just get them to help us translate for ourselves. Translating languages was way more complex
than even the experts could’ve guessed – it was literally turning out to be harder than
rocket science. Now, in the face of souring public opinion,
the US government – because, again, they only like throwing money at popular things – they
commissioned the ALPAC. Josh, is that a llama? Wait, no, sorry. It’s not a llama. It’s actually
an elite group of seven experts tasked with deciding the usefulness or uselessness of
machine translation and computational linguistics as a whole. That’s… quite a task!
Their verdict? Oh, we’ll… we’ll get into that. But first there’s something you need
to understand. Throughout the history of Artificial Intelligence,
people get sold on promises far beyond their wildest dreams and it happens so often that
there’s a term for it: the AI Hype Cycle. And when things go wrong, a chill sets in.
A chill with lasting effects. It’s called an AI Winter. (shivers) And no amount of jackets
can protect you from that chill. So, the verdict. The 1966 ALPAC report concluded
that “we do not have useful machine translation and there is no immediate or predictable prospect
of useful machine translation.” Ouch! Computers, it turned out, were twice as expensive
as human translators and they were doing way, way worse. The officials were convinced and
machine translation was defunded. The first AI winter set in.
Hmm. Computers and language obviously had a rough
start, but I promise it gets better. Eventually. Subscribe, go eat your borscht and, uh, write
some code. Chomp!

Tony wyaad

RELATED ARTICLES

36 COMMENTS

  1. CalculusFist Posted on March 6, 2016 at 7:16 pm

    Oh okay, I see what that person was saying about how you can tell she's reading her lines. You could fix that by using a cheat teleprompter rig.
    https://www.youtube.com/watch?v=hZaQWkRGtw8

    Reply
  2. Christian Jiang Posted on March 6, 2016 at 8:42 pm

    These videos are super interesting!!

    Reply
  3. Yogev Bocher Posted on March 6, 2016 at 10:13 pm

    I hate borshtsh.

    Reply
  4. BARBATVS 89 Posted on March 19, 2016 at 3:44 am

    "Artificial Intelligence" is a misnomer since something lacking awareness isn't intelligent.

    Reply
  5. sion8 Posted on March 19, 2016 at 4:33 am

    I think I prefer "the liquor is holding out all right but the meat has spoiled" to the 'actual' quote.

    Reply
  6. dani rich Posted on July 17, 2016 at 7:53 pm

    your videos are informing and interesting, just today i found your channel, and ill be staying tuned.
    i just wanted to comment that your voice is charming, and you are a fetching lady, but i would have felt much more at ease watching the videos with animation alone, like your partner does

    Reply
  7. joel Posted on July 22, 2016 at 9:00 pm

    look at the camera!

    Reply
  8. Nigel Hungerford-Symes Posted on August 17, 2016 at 11:48 am

    Kick ass channel, keep up the good work.

    Reply
  9. neo_pretor Posted on August 18, 2016 at 5:25 pm

    I like learning languages and accents. I've just discovered your interesting channel. Good research.

    Reply
  10. DrunkenCyclop Posted on August 22, 2016 at 9:13 pm

    Very interesting, I like those videos.

    BUT DAMMIT THE CAMERA IS OVER HERE.

    Reply
  11. marconatrix Posted on August 23, 2016 at 5:48 pm

    Is she autistic??

    Reply
  12. Константин Войнов Posted on August 25, 2016 at 3:19 pm

    Isaac Asimov was joking that machines should never translate natural languages.

    Reply
  13. Mirek Heikkila Posted on August 28, 2016 at 10:54 am

    Awww i love this presenter! So likable, animated and engaging! rawr! brings up fun concepts and brings up new ideas/problems as she presents them! very fun! tnx 😀

    Reply
  14. AWSMcube Posted on September 25, 2016 at 9:13 pm

    To this day, Google Translate still fails to translate with proper grammar

    Reply
  15. Einar Hjörleifsson Posted on November 21, 2016 at 6:47 pm

    нет

    Reply
  16. fuge74 Posted on December 11, 2016 at 2:41 am

    umm… so what is with the off camera focuses. it feels a little out of place and weird, except for the one time where she was showing off the gear in her exposition. do y'all have some weird love affair with your camera and set? should I expect this to suddenly break into a woman on machine porno love affair with odd and out of place exposition and really, really weird camera angles and positions?????

    Reply
  17. Samir Posted on December 17, 2016 at 7:08 pm

    there's no T at the end of borsh

    Reply
  18. William Razgunas Posted on December 18, 2016 at 12:08 pm

    This is a semi-unrelated comment and relates in computer-generated figure or humanoid robot bearing a near-identical resemblance to a human. The uncanny valley. The AI hype cycle looks nearly the same except the uncanny valley goes in the other direction.

    Reply
  19. Shubham Bhushan Posted on December 27, 2016 at 4:58 pm

    I think I love her 😍

    Reply
  20. Prod. Hxrford Posted on January 5, 2017 at 5:45 pm

    So they didn't have any native Russian speakers?

    Reply
  21. ⵉⵜⵔⵓⵏⴰⵓⵜ Posted on January 18, 2017 at 12:23 am

    You look like Beth Johanssen, another of my geeky crushes.

    Reply
  22. Omega Nine Posted on February 27, 2017 at 11:45 am

    "borsch", not "borscht". there's no "т" in "борщ"

    Reply
  23. M Posted on March 25, 2017 at 1:27 am

    Cars driven by A.I., will lead to a scond AI winter.

    Reply
  24. Sans the Skeleton Posted on March 31, 2017 at 11:19 pm

    Borsch has no T at the end.

    Reply
  25. EebstertheGreat Posted on September 18, 2017 at 7:44 am

    ALPAC's conclusion was absolutely accurate. In the 1960s, the notion of machine translation was completely hopeless. Even now, five decades later, with billions of times the computing power, machine translation is only barely adequate in most situations, and frequently still terrible. And even that is a bit of a "cheat," as Robin would say (god I hope her name is Robin; online citations are lacking). The best modern translators like Google Translate rely on suggested translations and on analysis of large, constantly updated corpuses. So a relatively few enterprising users can easily change translations for specific terms, and people have been doing this for years. Obviously we have come a very long way, but in the 40's, people were presenting the problem as if it could be solved in a decade. The disappointment of the 50s, 60s, 70s, 80s, 90s, 00s, and even today are completely justified. This is why nobody should announce a discovery before it has been confirmed.

    Reply
  26. John Azhderian Posted on October 29, 2017 at 10:25 pm

    The computer is an inanimate object. The programmer must a facility in English and Russian to properly write a program to do a proper translation. Both languages are complicated and convoluted. A computer is not a conscious being capable of real thought. If you want to translate from Chinese to Russian the programmer must thoroughly understand both language to do a proper translation. This is true for all other languages. The programmers makes the computer work.

    Reply
  27. Андрей Третьяков Posted on November 30, 2017 at 10:45 pm

    It's not "borsht", it is "Борщ" 🙂

    Reply
  28. Ken Archer Posted on December 14, 2017 at 5:40 am

    I would like to point out that translation is going to be harder than rocket science because rocket science is actually fairly easy. You simply point out what you want the rocket to do and it does it. The difficulties arise when the rocket scientist attempts to figure out how to get the rocket to do what they want. But, pretty much everything is harder than rocket science.

    Reply
  29. Tom Kelly Posted on June 14, 2018 at 11:33 am

    click english CCs on movie dialogue,AI makes spaghettispeak

    Reply
  30. Catriona Posted on August 1, 2018 at 10:35 pm

    Computers actually do a lot better translating from languages with a complicated grammar into english than vice versa. Because in highly inflecting languages you have information "hardcoded" into the language that in English are only transported by context or a lot of extra words you have to add to explain the extra meaning.
    This is by the way the reason Latin texts are so incredibly short in comparison to other languages. They just don't need to make extra words.
    In Russian, for example, "I was born in Moscow" will always tell you the gender of the speaker because it is coded in the flexion of "born". You can of course somehow translate that into english. And vice very if you add extra words in the english sentence. But few automatic translators will correctly shorten the sentence "I am a woman and I was was born in London" into gramatically Russian. I has no gender information from the word born and you would have to have an algorithm that picks up the word "woman", UNDERSTAND that it is related to the word "born" and then translate and shorten it to the correct Russian sentence "I was born in London" (in Russian it would contain the information) or at least the redundant "I am a woman and I was born in London". Because the latter would still have no clue wich of the two versions of "born" to use.
    Neither does "born" convey numerus. "born in Moscow" could actually refer to more than one person and in Russian you would clearly see it from the conjugation of the word "bear". Again from English to Russian – no information is given. Only from context.
    So to translate English into languages with lots of grammar is the actual nightmare. You'd need an AI that actually understands context. There is your next winter.

    Reply
  31. Mr Ramdom Posted on September 10, 2018 at 11:52 pm

    Borscht IS UKRAINIAN! !!!!

    Reply
  32. shadysheep Posted on February 26, 2019 at 4:24 am

    just use google translate, lol

    Reply
  33. Richard Deese Posted on March 14, 2019 at 10:45 pm

    Okay. This IS an interesting topic. However, I've "computed" that it is NOT going to be worth it (to me, at least) to watch 10,000 3 minute videos (or whatever the thoroughly unreasonable number IS) to get ONE "story." Bye-bye. Next time, fund a WHOLE video at once. Now THERE's a concept! Imagine if THAT caught on! Rikki Tikki.

    Reply
  34. 終極神卡 Posted on April 9, 2019 at 10:30 pm

    Я не знаю русский

    Reply
LEAVE A COMMENT