A new update rolling out to Samsung‘s Bixby voice assistant could be both really cool and highly terrifying at the same time. Bixby is Samsung’s AI-based virtual assistant that’s baked into all its smartphones, including the flagship Galaxy S23 lineup. It was originally introduced with the Galaxy S8 and S8+ in 2017 as the successor to the S Voice virtual assistant that debuted on the Galaxy S3 in 2012. However, despite Samsung’s efforts to popularize it, Bixby lags behind Google Assistant, Siri, and Alexa.
In Feb. 2023, Samsung announced a new Bixby feature that will let people answer phone calls with a cloned version of their voice. Samsung says the new feature will help people reply to phone calls by typing out a response that Bixby will read out in the user’s voice. The feature enhances the ‘Bixby Text Call’ function, which already allows users with select Galaxy smartphones to answer phone calls by typing a message that Bixby converts to audio and reads out to the caller. With the latest Bixby feature drop, users can personalize the Bixby Text Call voice with the new Bixby Custom Voice Creator app.
AI Voice Cloning For Samsung Galaxy Smartphone Users
Users will have to record their voice for Bixby using the Bixby Custom Voice Creator to create an AI-generated clone that can then be used to reply to incoming calls. Samsung says that it is working on eventually making the cloned voice compatible with other Samsung apps as well. For now, the feature is only available in Korean, and to the users of select Samsung smartphones running One UI 5.0 and above, including the Galaxy S23, S23+, and S23 Ultra.
How well this feature will actually work remains to be seen, but it is a cool concept that can open up new possibilities. Even in its nascent stage, artificial intelligence can clone human voices almost flawlessly, with AI voice clones already being used in podcasts, video games, and other areas. With Samsung’s new feature, it could finally cross the confines of commercial usage and be available to the public, potentially allowing people to book tickets to shows, respond to spam calls, book restaurant reservations, and more, without actually speaking.
However, as this feature slowly expands and becomes more sophisticated, it could also create a terrifying situation, where people on the other end of the line may not even know whether they are talking to an actual person or a robot. In addition, 4Chan users are already using AI voice cloning tech to generate hate speech and harass people by imitating the voices of celebrities. Deepfake audio has also been used for fraud in the past, so there definitely have to be identifiers and limiters in place so that tech like Samsung‘s new AI voice cloning feature isn’t used for nefarious purposes.
Source: Samsung