If you have spent some time researching artificial intelligence, you will discover that the internet is heavy on this topic. Ever heard of AI voice cloning, well, this place where AI is gradually bringing its changes as well. AI audio reproduction is already being used in podcasts and video games. How long do you think it will take for the general public to access it directly? The answer may be faster than you think. Samsung today revealed a function for its smart assistant Bixby that allows users to clone their voice to answer phone calls. With this feature offered by Samsung if you get a call but are unable to respond verbally, you can type your response, and it will be read out loud in imitation of your voice.
At the moment, not many tests have been reported about this feature because now it can only be accessed in Korea as the Bixby Custom Voice Maker app for a few Samsung smartphones (Galaxy S23, S23+ and the new S23 Ultra). We are unable to give an in-depth knowledge of what it’s like but we are sure that the potential of this technology is huge. The warning states that the sound quality can be terrible and the reaction time may be too slow to be of value. With the capabilities of today’s AI technologies, which produce realistic versions of sounds from just a few minutes of audio, cloning voices to answer calls is within the reach of current technology.
It’s also not new to answer voice calls using a text interface. The functionality, known as Bixby Text Call on Samsung phones, debuted with the maker’s One UI 5 Android interface. It was previously available exclusively in Korean, but now an artificial voice is used in English (and only with One UI version 5.1). There is a similar feature offered by Google called Call Screen, which enables you to answer potential spam calls with an automated voice. But Google’s service only allows you to choose from a set of pre-written responses instead of manually entering your responses.
It is not difficult to envision these capabilities evolving into more advanced automated systems in the near future. After all, it wouldn’t be hard to hook up a text-to-speech audio version of yourself to a chat program like ChatGPT or (if you’re in a serious mood) Microsoft Bing. It’s not clear what Samsung means when it says that in the future, user-generated sounds will be “compatible with other Samsung apps other than phone calls.”
In the event that you feel frivolous, you can use such a bot simply to waste the spammer’s time or summarize the information received on your call. Making a talking version of itself and assigning tasks to it via a bot could make the long-promised notion in the tech industry that AI assistants would be able to handle our administrative tasks a reality. But it’s not just about avoiding conversations that one would rather not have. Technology can also help those who have difficulty speaking or those with speech impediments to communicate more easily.
On the other hand, it can also lead to a variety of problems. Of course, Google has promised similar capabilities with its Duplex AI voice calls. It was released in 2018 as a way to use AI voice to automatically book restaurant reservations. However, opinions were divided on the technology, with many denouncing it as unethical and noting that it increased the workload of those who received the calls. Google’s current ambitions for the technology are unknown. The company ended support for Duplex on the web in late 2022.
Just like with any new technology, there can be drawbacks. The first concern is privacy. The process of cloning your voice requires you to register your voice, and although Samsung guarantees the security of your data, there is always the possibility that it could be compromised. There are concerns about possible misuse of the technology, including the creation of hate speech and fraud.
For now, we’re keeping our eyes fixed on this technology and how it evolves. In the near future when Bixby’s voice features spread to the rest of the world when you pick up the phone, one of the questions on your mind will be whether the person on the other end of the phone is actually a human.