Apple's Innovation Gives Voice to the Voiceless

Apple's latest innovation, "Personal Voice," was released on International Day of Persons with Disabilities, marking an important step forward in voice technology.

Yesterday, Apple showcased its new feature, Personal Voice, through a short film and an e-book. The company has been praised for its accessibility features, which make it easier for individuals with visual, hearing, and motor impairments, as well as older adults, to use their devices. It has taken a step forward in artificial intelligence by introducing features such as VoiceOver, Guided Access, Door Detection, Live Listen, and Point and Speak for Magnifier.

Personal Voice was announced earlier this year and is designed to support different user groups based on feedback from individuals with disabilities. Although Apple has not openly discussed artificial intelligence, it is rapidly updating its features to integrate better technology.

Cloning voices for healthcare has been an ongoing process. Previously, patients who lost their voices due to various illnesses had to use an electrolarynx. This device needed to be placed on the patient's throat, and vibrations produced a robotic voice.

Companies that clone videos and images have also cloned voices, which are used not only in entertainment but also in healthcare. Companies such as ElevenLabs, Murf.ai, Resemble ai, and Respeecher have created voice and video clones.

Examples of Artificial Intelligence Applications

By utilizing existing features, the Personal Voice plugin enhances the user experience. Users need to read a series of randomly selected text prompts aloud to provide a sample of their voice. Acoustic analysis of the voice sample extracts acoustic features such as pitch, timbre, and intonation. A text-to-speech model is trained on the user's voice data and a large dataset of text-to-speech pairs. The model learns to associate acoustic features with corresponding text and generates synthesized speech that mimics the user's voice.

All of this is done on the user's phone, without any privacy risks - a feature for which Apple is well-known. The created voice can be used for phone calls, FaceTime, and other applications. This feature can be used in conjunction with Live Speech, which was announced around the same time. You input what you want to say, and your personal voice speaks it out loud for you.

Potential for Malicious Use

This feature, which will provide voices for many people, has also raised concerns about security and privacy, given the growing threat of deepfakes. The internet is filled with stories of individuals and companies being deceived by voice clones and having their bank accounts emptied. Is it really wise to voluntarily give your voice recordings to Apple?

In its announcement, the company ensures that all data processing is done locally on the device, reducing the risk of data leaks. Access to the generation and management of personal voices is protected through biometric locks such as FaceID or TouchID, requiring the device to be unlocked and preventing unauthorized access. Personal voices can be shared between devices linked to the same iCloud account and third-party applications, but there seems to be no way to transfer voices to other devices.

Regarding the possibility of implementing additional safeguards, such as tracking synthesized voices for detection, enhancing security can be considered. "Public detection measures would be a good way to go, although given the company's focus on privacy and security, I suspect it may already include this feature," wrote the author and security expert Matt Smallman on the topic.

Vinod Iyengar, an AI expert and product manager at Third AI, is not as optimistic. "Deepfakes will soon become rampant," he says. Voice cloning can be used to create seemingly authentic fake audio content, making it more difficult to distinguish between genuine and fake recordings.

This could be another gray area and a potential legal trouble in the near future.

Meanwhile, speculation about Apple's future direction is growing on social media, suggesting that these features indicate the integration of more advanced artificial intelligence into Apple's future products. Discussions about the possibility of Apple surprising users with new local AI tools indicate a trend towards shifting from cloud-based data processing to local data processing.