Google Translation: Getting More Intelligent?

February 7, 2019

Translation has never been easier with AI and NLP tools. It is amazing for people who cannot speak foreign languages to communicate with the assistance of translation apps, like Google Translation. While there are many translation apps on the market, Google is by far the best free one. As with many of its products and services, Google spends countless hours perfecting its language algorithms. The Verge published “Google’s Head of Translation On Fighting Bias In Language And Why AI Loves Religious Texts.”

Macduff Hughes heads Google’s translation and in the interview discusses how Google has moved from translating word by word but entire sentences. The new and smarter translation method is called “neural machine translation,” it uses machine learning, and a lot of its data comes from religious texts. One problem Google Translation faces is gender biased language. In order for translation AI to learn, it needs to be fed a lot of accurate and diverse data. These data sources, however, reflect societal biases which the AI can learn and replicate, such as doctors are male and nurses are female. The goal is to overcome these limitations so people know there is more than one way to phrase something as well as explain the differences.

Google is addressing three big bias and nuance initiatives. The first is to expand full sentence gender translation to more languages, the second is improving document translation based on context, and the third is addressing gender neutral languages. On a funnier and conspiracy based note is in 2018, when people typed nonsense words into Translate it spat back religious information. The explanation is a logical way of teaching AI:

“Usually it’s because the language you’re translating to had a lot of religious text in the training data. For every language pair we have, we train using whatever we can find on the world wide web. So the typical behavior of these models is that if it gets gibberish in, it picks out something that’s common in the training data on the target side, and for many of these low-resource languages — where there’s not a lot of text translated on the web for us to draw on — what is produced often happens to be religious.”

Translation is becoming a tool to organize more of the world’s information, according to Hughes, because it allows more people to access stuff that was in a different language. The naysayers argue that Translation provides a very shallow translation and Hughes acknowledges that. However, Translation works for basic translation and someday AI might have the skills of a professional linguist. It is not perfect, but Google Translate gets you to the train station and the bathroom.

Whitney Grace, February 6, 2019

Google Translate: A Facelift but Some Post-Op Needed

January 17, 2019

Google Translate is the go to app to figure out what people are saying in a foreign language. One of the best features is that Google Translate scans the text and can translate it, albeit a garbled version. Google Translate has gotten a makeover and Mobile Syrup spills the details: “Google Translate Web Site Gets Material Design Refresh And New Features.”

Google Translation’s face lift is part of the Material Design change. The basic design remains the same, but the text and document buttons are rounder. There are also two new blue buttons above the input box, labeled “text” and “documents.” Text is to translate the current page, while documents translated uploaded files. There is a new “History,” “Saved,” and “Community” buttons. History contains a list of previous translations, while saved opens your own custom phrasebook of saved translations. Community is for language lovers, where they can verify and view other users’ translations. However, if you have to speak two-five languages to participate.

The newest change is:

“Perhaps one of the most significant changes is a new responsive design. Adjusting the size of the window will dynamically adjust Translate’s layout. For example, if you use a smaller window, Translate will take on a more ‘mobile friendly’ layout with a vertical-oriented system. When you expand your window out, it’ll adjust back to a horizontal layout.”

The changes to Google Translate are slight, but the app could use some enhancements when handling popular words; for example, “jajillas.”

Whitney Grace, January 17, 2018

Smart Software and Clever Humans

September 23, 2018

Online translation works pretty well. If you want 70 to 85 percent accuracy, you are home free. Most online translation systems handle routine communications like short blog posts written in declarative sentences and articles written in technical jargon just fine. Stick to mainstream languages, and the services work okay.

But if you want an online system to translate my pet phrases like HSSCM or azure chip consultant, you have to attend more closely. HSSCM refers to the way in which some Silicon Valley outfits run their companies. You know. Like a high school science club which decides that proms are for goofs and football players are not smart. The azure chip thing refers to consulting firms which lack the big time reputation of outfits like Bain, BCG, Booz, etc. (Now don’t get me wrong. The current incarnations of these blue chip outfits is far from stellar. Think questionable practices. Maybe criminal behavior.) The azure chip crowd means second string, maybe third string, knowledge work. Just my opinion, but online translation systems don’t get my drift. My references to Harrod’s Creek are geocoding nightmares when I reference squirrel hunting and bourbon in cereal. Savvy?

I was, therefore, not surprised when I read “AI Company Accused of Using Humans to Fake Its AI.” The main point seems to be:

[An[ interpreter accuses leading voice recognition company of ripping off his work and disguising it as the efforts of artificial intelligence.

There are rumors that some outfits use Amazon’s far from mechanical Turk or just use regular employees who can translate that which baffles the smart software.

The allegation from a former human disguised as smart software offered this information to Sixth Tone, a blog publishing the article:

In an open letter posted on Quora-like Q&A platform Zhihu, interpreter Bell Wang claimed he was one of a team of simultaneous interpreters who helped translate the 2018 International Forum on Innovation and Emerging Industries Development on Thursday. The forum claimed to use iFlytek’s automated interpretation service.

Trust me, you zippy millennials, smart software can be fast. It can be efficient. It can be less expensive than manual methods. But it can be wrong. Not just off base. Playing a different game with expensive Ronaldo types.

Why not run this blog post through Google Translate and check out the French or Spanish the system produces? Better yet, aim the system as a poor quality surveillance video or a VoIP call laden with insider talk between a cartel member and the Drug Llama?

Stephen E Arnold, September 23, 2018

Goggle Works to Understand Language

April 27, 2018

Back in the day, Google used to do this annoying thing that if you were searching in a foreign language it would translate the word into English. It was difficult to search for foreign language Web sites using Google, unless you went to one of the foreign language Web site endings. Google eventually fixed the “feature”, but the search engine giant is doing something new with searches and translation. Google Translate now tries to understand query text meanings and build responses using natural language. has more information on the new development in the article, “Google Translate For Yiddish? It Ain’t Work Bupkis.” After a brief history lesson about psychologist Frank Rosenblatt and how an AI finally beat world go champion Ke Kie, we finally get into the how of the new Google Translate. The new neural network is a great upgrade, but here are the bugs:

“Although Google Translate’s new approach sounds like a giant leap forward it creates all sorts of problems. To begin with every text must first be translated into English before it can be rendered into another language. Secondly, instead of warning that it doesn’t recognize a certain word the new system automatically inserts its own “creative” meanings in a second-rate imitation of human neural processing.”

In other words, Google Translate still does not have the human comprehension necessary to translate words and sentences accurately to the satisfaction of a person familiar with a language’s nuances. The human mind is still the better language tool and if you are translating using Google Translate, especially Yiddish, keep a dictionary on hand.

Whitney Grace, April 27, 2018

Real Time Translation: Chatbots Emulate Sci Fi

April 16, 2018

The language barrier is still one of the world’s major problems. Translation software, such as Google Translate is accurate, but it still makes mistakes that native speakers are needed to correct. Instantaneous translation is still a pipe dream, but the technology is improving with each new development. Mashable shares a current translation innovation and it belongs to Google: “Google Pixel Buds Vs. Professional Interpreters: Which Is More Accurate?”

Apple angered many devout users when it deleted the headphone jack on phones, instead replacing it with Bluetooth headphones called AirPods. They have the same minimalist sleek design as other Apple products, but Google’s Pixel Buds are far superior to them because of real time translation or so we are led to believe. Author Raymond Wong tested the Pixel Buds translation features at the United Nations to see how they faired against professional translators. He and his team tested French, Arabic, and Russian. The Pixel Buds did well with simple conversations, but certain words and phrases caused errors.

One hilarious example was when Google translated the Arabic for, “I want to eat salad” to “I want to eat power” in English. When it comes to real time translation, the experts are still the best because they can understand the context and other intricacies, such as tone, that comes with human language. The professional translators liked the technology, but it still needs work:

“Ayad and Ivanova both agreed that Pixel Buds and Google Translate are convenient technologies, but there’s still the friction of holding out a Pixel phone for the other person to talk into. And despite the Pixel Buds’ somewhat speedy translations, they both said it doesn’t compare to a professional conference interpreters, who can translate at least five times faster Google’s cloud.”

Keep working on those foreign language majors kids. Marketing noses in front of products that deliver in my view.

Whitney Grace, April 17, 2018

Google Translate Gets a Needs Improvement on Its Translation System

February 5, 2018

I read “The Shallowness of Google.” The critique is not from a trendy start up in Silicon Valley or an academic who flopped in a Google interview. The analysis is by Douglas Hofstadter. if the name does not ring a bell, this is the fellow who wrote Gödel, Escher, Bach, a quite fun read.

The main point of the write up is that Google’s implementation of its artificial intelligence and machine learning technology for Google Translate is bad.

Image result for alpha sled dog

Google wants to be perceived as the alpha dog in smart software. Do you want to take this canine’s kibble? Google can bite even thought it may not get the whole “idea” and “understanding” behind a reprimand.

Mr. Hofstadter writes:

Having ever more “big data” won’t bring you any closer to understanding, since understanding involves having ideas, and lack of ideas is the root of all the problems for machine translation today. So I would venture that bigger databases—even vastly bigger ones—won’t turn the trick.

The idea is that “understanding” is not baked into Google Translate. In addition to providing examples of screwing up translations from French, German, and Chinese, Google Translate does not look up information in Google Search. Mr. Hofstadter does.

He points out:

Google Translate can’t understand web pages, although it can translate them in the twinkling of an eye.

He correctly observes:

As long as the text in language B is somewhat comprehensible, many people feel perfectly satisfied with the end product. If they can “get the basic idea” of a passage in a language they don’t know, they’re happy.

Mr. Hofstadter touches upon two issues, which another informed critic might convert to a write up in the Atlantic:

  1. Google is simply delivering “good enough” services. The object is advertising, not outputting on point products and services for a tiny fraction of its user base
  2. Google’s hype about its smart software is only slightly less off-the-wall than the marketing of IBM Watson. The drum beat for smart software is necessary to attract young programmers who might otherwise defect to Amazon or other Google competitors and to further the illusion that Google’s technology is magical, maybe otherworldly and definitely the alpha dog in the machine learning Iditarod.

The write up is worth reading. However, I would not run it through Google Translate if you prefer to ingest the article in one of Google Translate’s supported languages.

And for a person going through the Google interview process, it is not a plus to suggest that Google’s technology might be little more than a C or possible an F. Rah rah is a better choice.

That’s why we love Google Translate here in Harrod’s Creek, but we have switched to Free since Google implemented a word limit.

Stephen E Arnold, February 5, 2018

Privacy Is Lost in Translation

October 30, 2017

Online translation tools are a wonder!  Instead of having to rely on limited dictionaries and grammars, online translation tools deliver real-time, nearly accurate translations of documents and other text.  It is usually good to double check the translation because sometimes the tools do make mistakes.  Translation tools, however, can make mistakes that lose privacy in translation.  Quartz tells an alarming story in, “If You Value Your Privacy, Be Careful With Online Translation Tools.”

Norwegian state oil company Statoil used to translate sensitive company documents.  One would think that would not be a problem, except stored the data in the cloud.  The sensitive documents included dismissal letters, contracts, workforce reduction plans, and more.  News traveled fast in Norway, resulting in the Oslo Stock Exchange blocking employees’ access to and Google Translate.

It was dubbed a massive privacy breach as private documents from other organizations and individuals were discovered. views the incident differently: sees things a little differently, however, saying it was straight with users about the fact that it was crowdsourcing human translations to improve on machine work. In a Sept. 6 blog post responding to the news reports, the company explained that in the past, they were using human volunteer translators to improve their algorithm, and during that time, had made documents submitted for translation public so that any human volunteers could easily access them. ‘As a precaution, there was a clear note on our homepage stating: ‘All translations will be sent to our community to improve accuracy.’ also offered to remove any documents upon request, but sensitive documents were still available when the Quartz article was written.  Vice president of Sales for Maria Burud pointed out that they offer a paid translation software intended for businesses to maintain their privacy.  Burud notes that that anything translate using a free web tool is bound to have privacy issues, but that there is a disclaimer on her company’s Web site.  It is up to the user to de-identify the information or watch what they post in a translation box.

In other words, watch what you translate and post online.  It will come back to haunt you.

Whitney Grace, October 30, 2017

Instantaneous Language Translation in Your Ear

June 21, 2017

A common technology concept in cartoons and science-fiction series is an ear device that acts as a universal translator.  The wearer would be able to understand and speak any language in the world.  The universal translator has long been one of the humanity’s pipe dream since the Tower of Babel and as technology improves we could be closer to inventing it.  The Daily Mail shares, “The Earpiece That Promises To Translate Language In Seconds: £140 Will Be Available Next Month.”

International travelers’ new best friend might be Lingmo International’s One2One translator that is built on IBM Watson’s artificial intelligence system.  Unlike other translation devices, it does not reply on WiFi or BlueTooth connectivity.  It supports eight languages: English, Japanese, French, Italian, Spanish, Brazilian, Portuguese, German, and Chinese (does that include Mandarin and Cantonese?).  If the One2One does not rely on the Internet, how will it translate languages?

Instead, it uses IBM Watson’s Natural Language Understanding and Language Translator APIs, which intuitively overcomes many of the contextual challenges associated with common languages, as well as understanding the nuances of local dialects…This allows it to translate what you’re saying, almost in real-time.

Lingomo might be relying on IBM Watson for its natural language API, they should also consider using Bitext, especially when it comes to sentimental analysis.  Some languages have words with multiple meanings that change based on a voice’s inflection and tone.

The ramifications for this device are endless.  Can you imagine traveling to a foreign country and being able to understand the native tongue?   It is the dream of billions, but it could also end some serious conflicts.

Whitney Grace, June 21, 2017

Maybe Trump Speak Pretty One Day

June 15, 2017

US President Donald Trump is not the most popular person in the world.  He is a cherished scapegoat for media outlets, US citizens, and other world leaders.  One favorite point of ridicule for people is his odd use of the English language.  Trump’s take on the English tongue is so confusing that translators are left scratching their heads says The Guardian in, “Trump In Translation: President’s Mangled Language Stumps Translators.”  For probably the first time in his presidency, Trump followed proper sentence structure and grammar when he withdrew the US from the Paris Accord.   While the world was in an uproar about the climate change deniers, translators were happy that they could translate his words easier.

Asian translators are especially worried about what comes out of Trump’s mouths.  Asian languages have different root languages than European ones; so direct translations of the colloquial expressions Trump favors are near impossible.

India problems translating Trump to Hindi:

‘Donald Trump is difficult to make sense of, even in English,’ said Anshuman Tiwari, editor of IndiaToday, a Hindi magazine. “His speech is unclear, and sometimes he contradicts himself or rambles or goes off on a tangent. Capturing all that confusion in writing, in Hindi, is not easy,’ he added. ‘To get around it, usually we avoid quoting Trump directly. We paraphrase what he has said because conveying those jumps in his speech, the way he talks, is very difficult. Instead, we summarise his ideas and convey his words in simple Hindi that will make sense to our readers.’

Indian translators also do Trump a favor by translating his words using the same level of the rhetoric of Indian politicians.  It makes him sound smarter than he appears to English-speakers.  Trump needs to learn to trust his speechwriters, but translators should learn they can rely on Bitext’s DLAP to supplement their work and improve local colloquialisms.

Whitney Grace, June 15, 2017


Online Translation Becomes a Joke

April 26, 2017

I am not much of a TV buff. I noted the article “Anne Hathaway Sang the Most Awkward Google Translations Beautifully on Jimmy Fallon.” I noted that I will survive was allegedly translated as “I will be punctual.” Close, right. The image below shows the original lyric and Google Translate’s version:


Online translation definitely loses none of the nuance and emotional impact. Ooops ooops. When that artificial intelligence controls autonomous drones, what could go wrong? Answer: Nothing. Perfect.

Stephen E Arnold, April 27, 2017

« Previous PageNext Page »

  • Archives

  • Recent Posts

  • Meta