OpenAI is reportedly preparing to launch a new AI-powered digital assistant, according to a recent article by The Information. This new multimodal model would not only be able to communicate verbally, but also recognize objects in images and audio.
The article from The Information mentions that this development could be part of what OpenAI plans to reveal today Monday in a presentation. Apparently, the new model would offer a faster and more accurate interpretation of images and sounds compared to current models of transcription and text-to-speech conversion. This could have significant applications in customer service, as it could help agents better understand the intonation of customers’ voices and even detect sarcasm.
Although it has been suggested that this model could surpass GPT-4 Turbo in some aspects, its accuracy remains a concerning issue, as AI, nowadays, is still prone to making mistakes. Additionally, there have been speculations that OpenAI might also be working on a feature for ChatGPT that allows making phone calls, which could represent a significant advancement in digital assistant technology.
However, it is important to note that this development does not seem to be related to the alleged presentation of GPT-5, which according to OpenAI CEO Sam Altman, is not in the company’s immediate plans. Despite this, OpenAI’s announcement could have an impact on the I/O developers conference, as Google is also exploring similar AI applications in its products, such as its “Pixie” project, a multimodal assistant that could compete with OpenAI’s new model.
OpenAI plans to make the official announcement through a live stream on their website today Monday at 7 p.m. (Spanish peninsular time). With this new development, the company could be marking an important milestone in the field of AI and human-machine interaction.
