According to individuals with insider knowledge of the situation, OpenAI is currently fine - tuning its audio AI model as part of the preparations for launching AI - enabled personal devices. These upcoming devices are anticipated to center predominantly on audio - based interactions. At present, although ChatGPT is capable of providing voice responses, the foundational models powering its voice and text versions are distinct from one another. OpenAI's in - house researchers are of the opinion that the existing voice models lag behind the text models in terms of both the precision of answers and the swiftness of responses.
