Lost, and gained, in digital real-time translation

Hindustan Times

25-05-2025

Alphabet, the parent company of Google, recently announced the launch of real-time translation (RTT) in Google Meet, its popular video communication service.
While text translation has been available for a while and conversation was possible on mobile phones — Apple Translate, for instance — by recording in one and replaying in another language, RTT when 100 people are conversing across the globe on an online communications platform is a leap. Add to it Google Beam, a device capable of rendering immersive 3D videos. The result would be a video conference in 3D with participants talking in their preferred language. The translation reaches listeners in the voice and tone of the speaker making it super realistic. Behind all this is Google's upgraded artificial intelligence-based large language model (LLM) Gemini, whose self-description if you ask it, is, 'I am a super-smart computer program that can understand and talk with you'.
Until OpenAI's bot ChatGPT burst into the public domain, speaking to computers was a skill to be acquired in advanced university courses. It took a few years for a student to have proficiency in Basic, C++, SQL, Java, Python, or any of the other languages required to give commands in logical sequences and unique syntax to computers to get tasks done. Artificial intelligence removed that qualification. LLMs have it in reverse. Instead of humans speaking computer code, the bots learnt human languages. Now, it is possible to chat with computers in English, Hindi, German, Chinese, or, if duly trained, in even less spoken languages. RTT takes it to a new level.
It could be a blessing for India which speaks, quite literally, in a thousand and more tongues. There are 22 officially recognised languages and close to 20,000 dialects. It is an uphill task when dialects and cultural contexts change every few miles, but it opens up numerous possibilities and opportunities. Ideas could flow freely. Talent spotting and hiring could become easier. RTT can also provide some unexpected political benefits too should we choose to look at it with an open mind.
India can protect and preserve its language diversity. Parliamentary debates could become richer when members speak in their preferred language. It's a stretch but language chauvinism could diminish; south Indians would not need to learn Hindi and Indians north of the Vindhyas could skip Tamil or Kannada lessons. It is quite another matter that the underlying reason for language conflicts is not necessarily the lack of comprehension but a contest between zealous regionalism and rasping nationalism. Yet, technology presents a chance to overcome them, at least in the online world.
On the flip side, it neutralises the advantage India has had so far as an English-speaking country. The IELTS, a test of English proficiency, will lose its relevance. The competitive edge in IT services and outsourcing industries of knowing English would erode as AI-based RTT catches on. Translation as a vocation would eventually vanish. Work could dwindle for professionals such as dubbing artists.
The other, more sinister fallout of the technology is it will likely abet fraud. It would become easier to impersonate someone and victims rarely have recourse. Mumbai, for instance, reported over 2,000 cyber fraud cases in four years to April 2025 but only two have so far ended in convictions. Deep Internet and mobile phone penetration combined with the lack of awareness and skills makes matters worse.
India has 81 mobile connections for every 100 persons. Although it is a major global player in the tech sector, its population is not very high on information and communications technology skills. More than half of Indians are good at communicating through applications such as WhatsApp, but just a quarter of adult Indians know how to send emails. Only 1% know how to code. Yet, advances in AI tech offer Indians constrained by language and location to quickly bridge their skill gap. India has among the world's cheapest costs for communication. Combined with the breaking down of language barriers, it is an opportunity should we choose to see it for what it is.
Dinesh Narayanan is a Delhi-based journalist and author of The RSS and The Making of The Deep Nation. The views expressed are personal