At Google I/O, the tech giant announced a groundbreaking new translation service known as the ‘Universal Translator.’ This AI-powered system dubs and lip-syncs speakers in a different language while keeping the intended tone and style of the original message, making it a great tool to help people understand content in multiple languages. However, as with any powerful technology, there is potential for misuse, and Google was quick to point out the steps taken to prevent this.
The ‘Universal Translator’ utilizes deepfake-generating tech to achieve its results, and while it is immensely useful in certain realms – such as media production and online courses –Google must take precautions to avoid abuse. James Manyika, who heads up the company’s ‘Technology and Society’ department, deemed this a ‘tension between boldness and safety’, and thus the service is currently only available to authorized partners. This eliminates the risk of it being used for malicious purposes, such as creating false or misleading information. Google is also working to integrate watermarking into the models in order to eliminate the possibility of tampering or other devious activities.
However, the threats from malicious actors cannot be underestimated, and extra precaution needs to be taken. But despite the risks, the many benefits it offers of making content in multiple languages accessible cannot be ignored. James Manyika promised that ‘we’ll be seeing promising results in course completion rates’ when using this technology, and remarked that it was an ‘enormous step forward for learning comprehension’.
Google has shown its willingness to make products that are both safe and powerful, and thus the ‘Universal Translator’ is an exciting tool being handled responsibly. Google’s decision to talk transparency and be upfront about the risks it poses allows users to remain aware and vigilant, while still enjoying its new products.
