OpenAI Introduces Highly Anticipated Voice Assistant for ChatGPT: Safety Concerns Addressed

OpenAI is set to introduce a highly anticipated voice assistant for ChatGPT to a select group of users following a delay to address safety concerns. The AI company revealed that the voice functionality will be accessible to a small number of paid ChatGPT Plus subscribers initially. While the product will provide four preset voices, it will not have the ability to mimic the speech patterns of individuals, as per OpenAI’s specifications. Additionally, OpenAI has implemented new filters to identify and reject requests for generating copyrighted audio content. The voice feature played a significant role in OpenAI’s recent product launch event in May, showcasing GPT-40, an enhanced version of the GPT-4 model known for its improved real-time handling of text, audio, and images. Last year, the company introduced a basic feature allowing ChatGPT to respond to users, but the latest iteration promises enhanced speed and integration with advanced image recognition capabilities to transform the chatbot into a more interactive conversational partner. Originally planned for a June release to a limited group of ChatGPT Plus users, OpenAI postponed the launch by a month to ensure it meets the necessary standards. The company is focusing on enhancing the software’s ability to recognize and reject certain types of content while efficiently handling requests from a vast user base. OpenAI stated that a gradual rollout will enable them to monitor usage closely and enhance the model’s capabilities and safety based on real-world feedback. Furthermore, the company is still developing video and screen-sharing features demonstrated at the May event, with no confirmed launch date yet. Consequently, the initial version of the voice feature will offer a more restricted set of functionalities.