OpenAI could debut a multimodal AI digital assistant soon
Briefly

The new model promises faster and more accurate interpretation of images and audio in comparison to existing models, aiding in customer service and educational settings.
The potential capabilities of the new model include recognizing caller intonation, assisting with math, and translating real-world signs, potentially surpassing GPT-4 Turbo in answering certain questions.
There are hints of a built-in ChatGPT ability for phone calls and evidence of OpenAI's preparations for real-time audio and video communication, separate from the upcoming GPT-5 model.
CEO Sam Altman refutes ties between the imminent unveiling and a model rumored to be better than GPT-4, with GPT-5 possibly seeing a release by year-end, impacting possible Google AI developments.
Read at The Verge
[
]
[
|
]