language translation today Apps are like self-driving cars. Incredibly useful, promising, nearing perfection, and almost entirely mechanically powered. The very fact that this technology exists is amazing.
Still, machine translation can sometimes be clumsy, if not unwieldy.
Consider a recent conversation I had with my neighbor, Andre, who immigrated from Russia last year. Andre, who speaks little to no English, relies almost entirely on Google Translate, the most popular voice recognition translation app that was first released 10 years ago, to navigate his American Girlfriend Dream.
Through his phone, Andrew and I can have amazingly deep conversations about where he's from, what he thinks, how we can help each other, and what he wants. But more than once, Google Translate didn't convey what Andre was trying to express, and we were both forced to shrug, laugh, and continue the conversation.
But as computers get smarter, Google, Apple, Microsoft, and others want to completely eliminate the language barrier that Andre and I shared that day. But for that to happen, faster neural machine learning is needed, which “could take years,” one developer I spoke to admitted.
It's not that it's important to wait. In fact, many consumers are surprised to learn how good today's translation apps already are. For example, this video shows three of his Microsoft researchers using the company's live translation software to conduct a conversation in multiple languages. This video is 7 years old. But when I showed it to some friends, they reacted as if they had seen the future.
“The technology surrounding translation has come a long way in a very short period of time,” said a spokesperson for DeepL, an award-winning machine translation service that licenses its technology to companies such as Zendesk, Coursera, and Hitachi. says Erica Richter. “But this did not happen in parallel with consumer awareness.”
I am also a good example. I've been writing about technology for nearly 20 years, but it wasn't until I started researching this story after a fateful encounter with Andre that I realized how adept Google Translate, Apple Translator, Microsoft Translator, and Amazon Alexa are. did not. This technology is not yet able to provide the instant translations that we would expect from a live human translator. But turn-based speech-to-speech, text-to-speech, or photo-to-text translations are incredibly powerful.
And it's getting better every year. “Translation is one of the products he built entirely using artificial intelligence,” he said, a Google spokesperson. “In 2016, he introduced Google's Neural machine, which has significantly improved its accuracy in translating whole sentences rather than just phrases.”
At the same time, in half of the six apps I tried for this article, even basic greetings sometimes failed. For example, when you ask Siri and Microsoft Translator to translate “Olá, tudo bem?” From Portuguese to English, both answered “Hello, how are you?” correctly. Google Translate and Amazon Alexa, on the other hand, returned a more literal and awkward “Hello, are you okay?” or “Hello, are you okay?” Not a complete failure. But it's nuanced enough to cause hesitation and confusion in the listener.
