OpenAI has decided to delay the launch of the voice assistant features of its ChatGPT chatbot due to the need for additional security testing. Originally planned to be introduced to some paying subscribers in late June, the release has been postponed by a month. The new features will eventually be available to all paying users in the fall, depending on whether security and reliability standards are met.
In a recent demo, OpenAI showcased the new capabilities of ChatGPT, which generated excitement among users. The chatbot now has the ability to speak in synthetic voices and exhibit more advanced AI features using the GPT-4o system. This system can respond to a person’s tone of voice and facial expressions, engage in more complex conversations, and speak in expressive tones.
Amid the growing competition in generative artificial intelligence, various technology companies have encountered unexpected challenges with their AI projects. For instance, Google recently adjusted the frequency of AI-generated answers displayed in search results after some bizarre mistakes were made. Similarly, Microsoft made changes to its AI chatbot to address issues with its responses.
OpenAI emphasized the need for more time to enhance the new voice version of ChatGPT and improve its ability to identify and block inappropriate content. Many AI tools have faced criticism for producing false information, spreading biased or offensive content, or displaying inherent biases in their outputs. The complexity of interactions is heightened when designing a chatbot that can interpret and imitate emotions, potentially leading to errors.
The company stated that ChatGPT’s advanced voice mode aims to understand and respond with emotions and nonverbal cues, striving to create more natural conversations in real-time. OpenAI is committed to delivering these new experiences to the public with careful consideration for privacy and safety.