An Indian Politician Used AI to translate his speech into Another Language

While the newest technologies are being implemented into the daily lives of people to make life easier, it seems that politicians are thinking of ways to use it to meet their requirements when it comes to politics and elections.

Recently, Manoj Tiwari, an Indian politician, was in the news for having used artificial intelligence to make it seem like he was saying things that he did not actually say. He was seen speaking a native Indian language Haryanvi, a dialect of Hindi, that was different from the English version the video was shot in.

The firm responsible for this, The Ideaz Factory, explained that the video was created for the ‘positive’ campaigning of the local ruling party, Bharatiya Janata Party, to which Tiwari belongs. This was achieved using the technology of deepfake.

Deepfake is a combination of the words Deep learning and fake. It is the usage of deep learning technology in order to create fake videos. This deep learning is brought together by other powerful tools like machine learning and artificial intelligence. The machine learning employed involves training neural network architectures such as the autoencoders and generative adversarial networks.

This technology is capable of making a person in the video do things they were originally not doing, such as acting a certain way, or saying things, or being entirely replaced by someone else. This is achieved by using the said tools thereby manipulating the viewer.

Regarding how the video starring Tiwari was made, Sagar Vishnoi, who works with The Ideaz Factory elaborates: “We used a ‘lip-sync’ deepfake algorithm and trained it with speeches of Manoj Tiwari to translate audio sounds into basic mouth shapes.” This enabled Tiwari to extend his reach to the Haryanvi community and reach a voter base that he could not otherwise have done.

While deepfakes are not common in the political stream, this is not the first time this has happened. A video from 2018 of the then US President Barack Obama raised worries over how fake videos could be parading as genuine videos in the political sphere. Needless to say, this has gotten the political and technical analysts discussing what technology can do for a country’s politics and the various ways it can influence elections without actually meaning to.

Read – Emotion-Detecting Technology Should be Banned, Says AI Now

The US was quick to react with the US House Ethics Committee informing members that it would be considered a violation of House rules if any member posted such a deepfake video. California, on the other hand, has already passed a law making it illegal for deepfakes of politicians to be shared under 60 days from any election. Twitter, Facebook, and Reddit have also updated their policies to combat deepfakes on their respective platforms.

Deepfakes already has quite the reputation for being used in a slew of celebrity pornographic videos, fake news, hoaxes, and many such harmful videos. However, it is yet to be seen how the deepfakes will take over media consumption as we know it and how legislations tackle the morally grey areas that are sure to arise from such media.

Sreemitra Somanchi
A tech enthusiast with an itch to write. She is Interested in Consumer Technology in any form factor. She lives on the Google side of the world.

LEAVE A REPLY

Please enter your comment!
Please enter your name here