Punjabi singer-rapper Sidhu Moose Wala’s untimely death in a shootout on May 29 last year left millions of fans devastated. Moose Wala became a musical icon in a short span of time and released hundreds of songs that dominated the charts. Even after his demise, his songs have been released to huge fanfare. While his death at only 28 years of age left a dent in the Punjabi music industry, a man from the US has stunned music lovers by creating song covers of Moose Wala by generating his AI voice. Amarjit Singh, aka MRA, is a software engineer based in the USA, who has created many songs using Moose Wala’s vocals generated by the artificial intelligence software Stable Diffusion. From singing Rahat Fateh Ali Khan’s ‘Tumhe Dillagi’ to singing Diljit Dosanjh’s ‘Born to Shine’, the vocals sound just like Moose Wala and give goosebumps to the listeners. View this post on Instagram A post shared by Amarjit Singh (@djmrasingh) Moose Wala’s voice singing classic songs like Gurdas Maan’s ‘Challa’ or Surjit Bindrakhiya’s ‘Tera Yaar Bolda’ has no doubt caught the attention of fans. However, such a recreation of songs using a deceased singer’s voice raises questions of ethics as well. Amarjit has now decided not to release new songs using Moose Wala’s voice as he wants to respect his family’s wishes. View this post on Instagram A post shared by Amarjit Singh (@djmrasingh) What inspired him to create AI songs In a chat with The Indian Express, Amarjit shared what inspired him to create music generated by AI voices. “I have delved into AI research, particularly in the realm of finding ways to catch a glimpse of our departed loved ones. Losing my parents and sister has motivated me to explore this field and seek ways to connect with their memory through innovative technologies. In addition to my passion for software engineering, I enjoy creating music as a hobby,” he said. AI voice of Moose Wala Sharing about the response he received to Moose Wala’s voice, he said, people not only loved it but also expressed their profound admiration. “They recognised the remarkable capability of this technology, as it allowed us to effectively revive his voice. It is truly remarkable that we now have the means to preserve his legacy, ensuring that it continues to resonate with future generations,” he mentioned. Ethical side of creating AI music On the ethical side of creating songs using AI, Amarjit said that creating transformative art, such as a voice that somewhat resembles an artist, for research and tribute purposes is not unethical. “Many people create visual art and musical edits of their favourite celebrities as a way to pay tribute to them. However, presenting it as if the song is performed by a real artist and not AI would indeed be unethical. Additionally, using such technology for monetary purposes without permission would also be considered unethical,” he said. View this post on Instagram A post shared by Amarjit Singh (@djmrasingh) Why he vowed to stop creating new songs based on Moose Wala’s AI voice On what made him stop creating new songs based on Moose Wala’s voice, he said that though no one from the singer’s family reached out to him, he recently saw on Instagram their requests urging singers and music producers not to release Moose Wala’s songs without their permission. “Out of respect for his parent’s wishes, I have decided not to create anymore,” Amarjit said. Future of AI music Regarding the future of AI music, he said, “AI can be incredibly resourceful for small creators as it enables them to grow and complete projects within a limited time and budget. Singers can contribute their voices to support aspiring artists and lyricists, fostering collaboration and innovation within the music industry.” Amarjit added that AI can become a powerful tool for the families of departed artists to preserve and cherish their loved ones’ voices, providing solace and at the same time maintaining their artistic legacy. “Even ageing or injured artists can find renewed inspiration and artistic expression through AI. It enables them to continue creating and sharing their talent, despite physical limitations,” he further said. View this post on Instagram A post shared by Amarjit Singh (@djmrasingh) Difference between AI and real music in future Asked if he sees a bleak future in which it would be difficult to differentiate between what is real or what is AI generated, Amarjit said that people can already distinguish between official works and derivative creations. There are thousands of remixes made for songs, and individuals generally understand that these remixes are not directly produced by the original artist, he explained.
