For many couples, smartphones can be a source of friction. Yet, for David Duda and Hong Liang in New Haven, Connecticut, their mobile devices are absolutely vital – so much so that they keep eight external battery packs on hand. For David and Hong, losing phone power means losing their ability to connect.
David, 62, is an English speaker, while Hong, 57, communicates in Mandarin. Their everyday conversations, much like movie subtitles, depend entirely on Microsoft’s free Translator app, which provides real-time text translations of their spoken words.
Despite being married for three years, a recent December afternoon found them strolling hand-in-arm, resembling newlyweds. This closeness, however, was as much a practical necessity as it was a sign of affection. As one person spoke and guided their path, the other kept their gaze fixed on the phone, absorbing the translated conversation.
Even a simple joke from David required a brief pause; he’d hold his laughter until Hong had time to read the translated punchline.
Communicating this way demands complete focus. David and Hong can’t simply half-listen or multitask during a conversation. There are no casual shouts from another room. When they seek true understanding and deeper connection, they dedicate hours on the couch or in bed, patiently exchanging thoughts until both feel fully understood.
“The translator forces you to be fully present, reading and listening simultaneously,” David explained. “It demands more attention, which, in a marriage, is clearly a valuable thing.”
**[Video: David Duda and Hong Liang use Microsoft Translator to communicate while browsing the bookstore David owns in New Haven, Connecticut.]**
“He’s incredibly thoughtful, noticing every detail and anticipating my needs,” Hong shared in Mandarin, as conveyed by a human interpreter. She added a note on the machine translation’s quirks: *“Machine translation: He is very attentive. He thought about anything for me, so I was impressed by a lot of his details.”*
Recent years have seen remarkable advancements in automated translation, making it a viable tool not just for globetrotters and businesses expanding into new markets, but also for individuals like David and Hong, whose deep connection defies linguistic barriers.
In fact, during the interview with Hong, I relied on Apple AirPods for real-time translation alongside Microsoft Translator. While I grasped much of her message, the inherent delay made the conversation tricky. Significant misunderstandings only came to light later, after a professional interpreter meticulously reviewed our dialogue.
“It’s truly remarkable that people can now communicate with a far broader spectrum of individuals than ever before,” noted Lera Boroditsky, a cognitive science professor at U.C. San Diego. She acknowledged, “While these translations will inherently be imperfect, the ability to connect at all outweighs the alternative.”
The concept of a universal translator, a staple of science fiction from “Star Trek” and beyond, has always captured our imagination. Today’s real-world translation tools from Microsoft and others, capable of handling around 100 of the world’s 7,000 languages, are bringing that dream closer to reality, allowing billions to at least navigate basic communication, if not achieve perfect understanding.
Sustaining a deeply intimate relationship purely through apps, however, might seem less plausible. As French author André Maurois once penned, “A happy marriage is a long conversation that always seems too short.” For David and Hong, their enduring conversation began at a fortunate moment in time, when such technological aids were available, and they possessed the patience and dedication required to master them.
“For us, it’s actually quite enjoyable,” David mused, adding, “If people weren’t genuinely in love, this process would undoubtedly be far more frustrating.”
Their Unique Love Story: A Journey in Translation
**[Image: Microsoft’s free translation app allows Mr. Duda and Ms. Liang to communicate fluidly, although the translations aren’t always accurate.]**
**[Image: Apple AirPods can now translate a dozen languages, including Chinese. Mr. Duda and Ms. Liang have been testing them out.]**
David and Hong’s story began in Xian, China, during the autumn of 2019.
David joined his brother on a business trip to explore Xian’s historical marvels like the Terracotta soldiers and Buddhist temples. It was there that his brother’s business partner introduced him to her friend, Hong, who would drive him around. By the week’s end, David was completely captivated.
“She’s the most cheerful, joyful person you could ever encounter,” he recounted. Sensing a strong connection, David attempted to kiss Hong goodbye before his departure, but she, taken aback, offered her cheek instead.
Nevertheless, Hong was intrigued. After David returned to his bookstore in New Haven, they continued to connect through WeChat, the popular Chinese messaging app. Their conversations were text-based, with Hong meticulously copying and pasting messages into an English-Chinese translator.
Just a few months after their initial meeting, the world plunged into the COVID-19 pandemic. Both David and Hong, previously divorced with adult children, found themselves in isolation. They began exchanging messages daily, delving into their pasts, discussing their previous marriages, and sharing details about their families. As pandemic lockdowns forced businesses to close, David’s work shifted to preparing books for online orders, leading him to work nights – a schedule that serendipitously aligned with Hong’s waking hours.
“For the next two years, our phones became our primary means of getting to know each other deeply,” David shared.
**[Image: David Duda and Hong Liang’s translation technology is so essential that they own eight external battery packs.]**
In September 2022, as China eased its pandemic travel restrictions, Hong made the momentous decision to book a one-way flight to the United States. A wave of nervousness washed over her: had she truly understood everything David had conveyed through their translated messages?
David greeted Hong at the airport, holding a sign that proudly declared, in Chinese characters, “Love of my life.”
“I was profoundly moved,” Hong recounted, “seeing him do something so heartfelt in front of everyone at the airport.” She again noted the machine translation’s simpler take: *“Machine translation: I think this is a very, very romantic thing, and in front of so many people at the airport.”*
Lost in Translation: A Near-Death Misunderstanding
Automated language translation has captivated computer scientists since the 1950s, when Georgetown University researchers designed a system capable of translating a few hundred Russian words into English. However, effective translation goes far beyond mere word-for-word substitution. Languages differ significantly in grammar, structure, and the contextual meanings of words. Furthermore, idioms, like ‘it’s a piece of cake,’ pose a distinct challenge, as they defy literal translation.
These intricate complexities spurred machine learning researchers to action, prompting the fundamental question: Could computers be trained to truly understand and interpret human language?
The early 2000s saw Google and Microsoft introduce web-based translation services employing statistical methods. A pivotal breakthrough, however, arrived a decade later when Google researchers unveiled a novel technique for processing vast datasets. This innovation heralded a new era for artificial intelligence, paving the way for advanced chatbots like ChatGPT. The core of this technique centered on language processing, with the original research demonstrating its efficacy in translating English into French and German.
Since then, automated systems have achieved remarkable accuracy in text translation. Nevertheless, using AI to translate spoken words in real-time presents a greater challenge and yields less reliable results.
While messaging on WeChat, David and Hong occasionally forgot their linguistic differences. However, face-to-face communication introduced new complexities.
Their initial attempts involved a $600 handheld translator and earbuds offering spoken translations. Both proved impractical, requiring Wi-Fi and rendered useless by ambient noise. Ultimately, they found their solution in Microsoft’s Translator app, which converts their speech into text. Interestingly, they haven’t yet explored generative AI chatbots for translation, despite platforms like ChatGPT and Claude promoting translation as a key feature.
Despite their intimate conversations being recorded and stored by a massive corporation, David expressed no concerns. He even gave Microsoft permission to analyze his audio clips, hoping to contribute to the technology’s ongoing improvement.
**[Video: David Duda and Hong Liang use Microsoft Translator while shopping for groceries.]**
Clearly, Translator still has room for refinement. Wallace Chen, a professor of Chinese-English translation and interpretation at Middlebury College, reviewed my interview with the couple. He noted that while the AI handled simple exchanges adequately, it struggled significantly with more extended conversations.
As an example, Hong shared that shortly after arriving in the U.S. and being greeted by David at the airport, she contracted COVID-19 and felt so unwell that she believed she was dying.
However, the app’s translation on my screen was entirely different: it reported her saying she got a “new crown” and thought she was going to die. Puzzled, I mistakenly asked if she meant David’s romantic airport gesture was so overwhelming it made her feel she could die.
“No, no, no,” she insisted, repeating her words, but the misunderstanding persisted.
A month later, upon reviewing the video, Dr. Chen clarified the confusion. The app had literally translated the Chinese term for COVID-19, ‘novel corona,’ as “new crown.” Hong’s intention was to convey that David’s devoted care during her illness had profoundly deepened her affection for him, hence her mentioning it.
Further communication breakdowns occurred when Hong’s responses were lengthy. The speech-to-text transcription often couldn’t keep pace, omitting words or entire sentences, akin to a garbled phone call.
Chris Wendt, a former Microsoft group program manager who dedicated two decades to Translator, emphasized that “the person speaking needs to verify what they said.”
This critical step involves focusing on the app’s display, rather than the person you’re addressing, to ensure your words are accurately captured and translated.
This revelation came as a surprise to the couple. “Why didn’t anyone tell us this two years ago?” David exclaimed.
Converting speech into text is inherently difficult, even before language translation is factored in. Philipp Koehn, a renowned machine translation pioneer at Johns Hopkins University, explained that microphone quality, background noise, and simultaneous speakers can all severely hinder accurate speech recognition.
Moreover, spoken language is far more nuanced than written text. It encompasses diverse accents and dialects, subtle hesitations that convey meaning, and layers of emotion and tone that are often lost in transcription. People speak at varying speeds, sometimes ramble, or leave thoughts incomplete.
**[Image: Pinned up at their home is a sheet to help Ms. Liang with the pronunciation of English letters.]**
Dr. Chen cautions that automated translators should be used with extreme care “in situations where the precision of every single word is paramount.”
David and Hong are well aware of the app’s frequent misinterpretations. They even have a playful expression for these moments: *bù bù hǎo*, pronounced “boo boo how,” a twist on a Chinese phrase meaning “not good.” When such errors occur, they patiently rephrase, resort to body language, or find an illustrative image online. Ultimately, words, whether human or machine-translated, can sometimes fall short for all of us.
The Grand Vision for Translation Technology
Microsoft declined to comment on the record. However, I posed a question to Chris Wendt, the former program manager: when his team developed Translator, did they ever envision it as a tool capable of sustaining a marriage?
“That was precisely the goal,” he affirmed. “We genuinely believed it was achievable.”
Wendt and his team harbored grand ambitions: to foster global connection by dismantling language barriers. Their vision gained traction in 2014 when Microsoft showcased simultaneous translation during a Skype call between two schoolgirls, one in Seattle and the other in Mexico City. The sheer amazement on the girls’ faces, and those of their classmates, as a computer-generated voice instantly translated their words into Spanish or English, was palpable.
Since that time, the translation industry has surged to an estimated $31 billion, according to market intelligence firm Slator. In the past year alone, platforms like YouTube and Instagram have rolled out tools enabling creators to automatically translate and dub their videos into various languages. Amazon is experimenting with “A.I.-aided dubbing” for foreign films into English and Spanish, and also provides AI translation for self-published Kindle books. Reddit has translated its entire site content into 30 languages, and numerous video call services now offer real-time translation for subscribers.
Wendt drew a parallel between AI translation and GPS technology, which underpins mobile navigation apps.
“With GPS, the fear of getting lost vanishes,” remarked Wendt, who personally uses the Translator app when traveling. “It empowers you to venture into situations you would otherwise avoid.”
Just two months after Hong’s arrival, David knew he wanted her to stay. During a Thanksgiving dinner with his extended family, he romantically knelt down and attempted a proposal in Chinese.
Though Hong couldn’t understand his exact words, the profound gesture of his kneeling was unmistakable. They married just one month later.
Their married life unfolds much like any other couple’s: they enjoy dining out, cycling, beach strolls, and watching TV shows with Chinese subtitles (they’re fans of “Ted Lasso”). Yet, in one significant aspect, they are quite unique: in three years of marriage, they haven’t had a single fight. A typical heated argument, often characterized by a swift volley of words, would be almost impossible to conduct through a translation app.
“Perhaps the secret to a truly lasting marriage is speaking different languages,” David quipped playfully.
Both David and Hong believe their previous experiences with divorce and their maturity have significantly aided their communication. Having witnessed marriages fail, they now possess a deeper understanding of what makes a relationship thrive. They are also incredibly attuned to non-verbal cues; Hong noted that David’s facial expressions and gestures convey more about his emotions than the Translator app ever could.
Per Urlaub, a professor and director of global languages at M.I.T., acknowledged that “these translation apps struggle with nuances like metaphors, humor, tonal shifts, and cultural references.” However, he concluded that for David and Hong, the apps are undeniably “meaningful and functional.”
“Technology is enabling intimacy in circumstances where it would otherwise be impossible,” he stated, concluding, “In their case, the success speaks for itself.”
Beyond Rom-Com Tropes: Real-Life Love and Language Learning
**[Image: Three years into their marriage, Mr. Duda and Ms. Liang have learned about 200 words in each other’s language. Technology, they have found, offers them a better way to communicate.]**
For anyone familiar with the rom-com “Love Actually,” the expected narrative for a couple with palpable chemistry and a language barrier involves both diligently studying to achieve fluency.
While David attempts to grasp basic Mandarin using children’s books and apps like Duolingo, and Hong, a masseuse and acupuncturist, took an English class upon arriving in New Haven and also uses language learning apps, they both admit to knowing only about 200 words in each other’s languages. David confessed feeling too old to achieve fluency in a new language.
David and Hong believe their enduring infatuation is what truly makes their unique situation work. David candidly admitted that should their passion ever wane, the constant effort of using translation apps could very well become maddening.
Hong’s gaze remained fixed on her phone, diligently reading David’s translated words before she looked up and offered her own thoughts.
“In essence, it aligns with a Chinese saying,” she remarked, as translated by my AirPods. “We believe two people come together because they are still unknown to each other, eager to uncover each other’s mysteries.”
She jokingly suggested that once she truly knows him, her interest might diminish. She then watched David on his phone, patiently waiting for him to finish reading her witty remark before both burst into laughter.