Apple, Google, and Meta are working hard to perfect a sci-fi gadget: a universal translator.

For decades, TV shows like Star Trek and novels like The Hitchhiker's Guide to the Galaxy have featured fictional universal translators that can seamlessly convert any language into English and vice versa.

Today, these once-futuristic devices are edging closer to reality.

At last month's iPhone launch event, Apple showcased a video depicting a scenario that many travelers dream of. In the video, an English-speaking tourist is buying flowers in a Spanish-speaking country. The florist speaks to the tourist in Spanish, but the tourist hears clear and coherent English.

"All red carnations are 50% off today," the tourist hears in English through their headphones, almost simultaneously with the shop assistant's speech.

The video is part of the marketing materials for Apple's latest AirPods Pro 3, but this feature is one of many similar offerings from tech companies, including Alphabet, Google's parent company, and Meta (which owns Facebook and Instagram). The launch of ChatGPT by OpenAI at the end of 2022 spurred technological advancements and ushered in a new era of generative AI. Nearly three years later, these advancements have given rise to real-time language translation capabilities.

For Apple, real-time translation is a key selling point for the AirPods Pro 3. The new $249 headphones allow users to instantly hear French, German, Portuguese, and Spanish translated into English. The real-time translation feature will also be available as an update for AirPods 4 and AirPods Pro 2.

When two people wearing AirPods are having a conversation, the dialogue can be translated bidirectionally in each user's headphones simultaneously. In Apple's video demonstration, it looks as if two people are conversing in different languages.

Analysts are excited, believing that this feature could mark a step forward in Apple's AI strategy. This translation feature requires a sufficiently new iPhone to run Apple's AI software suite, Apple Intelligence.

"If we can really use AirPods for real-time translation, this feature will actually encourage people to upgrade," said Gil Luria, an analyst at DA Davidson.

As AI can translate languages as quickly as humans speak, translation is becoming a major battleground in the tech industry.

But Apple isn't the only player.

A Crowded Market

In the past year, Google and Meta have also released hardware products with real-time translation capabilities.

The Google Pixel 10 has a feature that can translate what a speaker says into the other person's language during a call. This feature, called "Voice Translation," aims to preserve the speaker's intonation. "Voice Translation" was launched on users' phones via a software update.

In Google's live demonstration in August, the Voice Translation feature successfully translated a line from comedian Jimmy Fallon into Spanish, and it actually sounded like the comedian speaking. Apple's feature doesn't attempt to mimic the user's voice.

Meanwhile, Meta announced in May that its Ray-Ban Meta glasses will be able to use the device's speakers to translate what a person says in another language, and the other party in the conversation will be able to see the transcribed translation on the user's phone.

Meta is expected to unveil its next-generation smart glasses at a recent product keynote. These glasses will have a small display in one of the lenses. It's unclear whether Meta will introduce more translation features.

OpenAI demonstrated the intelligent voice assistant mode of ChatGPT in June, and its built-in smooth translation function is one of many features. ChatGPT is integrated with Apple's Siri but doesn't offer a voice mode. OpenAI plans to collaborate with former Apple design guru Jony Ive to launch new hardware products in the coming years.

The rise of real-time translation could also reshape the entire industry. A study released by Microsoft Research in August found that translators are the profession most threatened by AI, with 98% of translation work overlapping with what AI can do.

Dedicated Translators

In the past few years, many specially designed translation devices have entered the market. Leveraging global high-speed cellular services and improving online translation services, they produce puck-shaped devices or headphones with built-in translation functions that cost a few hundred dollars.

"I like what Apple is doing because it really shows how urgent this problem is," said Joe Miller, the U.S. general manager of Japan's Pocketalk. Pocketalk manufactures a $249 translation device that can translate conversations between two people speaking different languages into audio and text.

According to Canalys, Apple shipped about 18 million pairs of wireless headphones in the first quarter alone. So, the company's entry into this market will expose a broader customer base to the advancements in translation technology made in recent years.

Despite Apple's entry into the market, manufacturers of dedicated devices say that their focus on accuracy and language knowledge will provide better translations than those offered for free on new phones.

"We've actually hired linguists," said Aleksander Alski, the head of Vasco Electronics in the U.S. and Canada. The Polish company launched a translation headset called E1 in January and plans to introduce a model that can mimic the user's voice, similar to Google's feature. "We combine AI with human input, and that's why we can ensure a significant improvement in translation accuracy for all languages."

There's also the home-field advantage. Vasco Electronics' largest market is Europe, and Apple states on its website that its real-time translation feature is not available to EU users.

Some products launched by tech companies lack universality and currently only support a few languages. Apple's voice assistant feature currently supports only 5 languages, while Pocketalk supports 95.

Miller of Pocketalk believes that the potential of this technology goes far beyond tourists ordering a glass of wine in France. He says that the most powerful applications of this technology are in workplaces like schools and hospitals, where the privacy and security features required far exceed what Apple and Google can offer.

"This has nothing to do with luxury travel and trips," Miller said. "This is about the intersection of language and friction, and it needs to be discussed."

This article is from the WeChat official account "Edu Guide," and is published by 36Kr with authorization.