OpenAI has announced the launch of an update introducing a new flagship artificial intelligence model named GPT-4o. This update also includes improvements in desktop services and voice assistant capabilities.
During a presentation at the OpenAI offices, Chief Technology Officer Mira Murati emphasized the significance of the new model in advancing AI technology. The updated GPT-4 models are now accessible to free users, offering faster and more accurate AI capabilities that were previously restricted to paying customers.
“We are envisioning a future where interactions with machines are transformative,” stated Murati. “GPT-4o is a major shift in that direction.”
Among the updates revealed by OpenAI were enhancements to the quality and speed of ChatGPT’s international language capabilities, along with the ability to upload images, audio, and text documents for analysis by the model. The feature rollout will be phased to ensure safe use.
The event showcased live demonstrations of the model’s new voice features. OpenAI researchers interacted with the AI voice model, prompting it to generate bedtime stories about love and robots with varying emotions and vocal inflections. Additionally, the model was shown solving a mathematical formula using the phone’s camera function in ChatGPT’s voice mode.
During one demonstration, researchers asked the AI model to interpret facial expressions and emotions. The voice assistant from ChatGPT interpreted the researcher as “happy, cheerful, and possibly a bit excited with a big smile.”
“You seem to be in a great mood,” remarked ChatGPT in a pleasant female voice. “Care to share the reason behind that positivity?”
This announcement follows last week’s rumors about OpenAI unveiling a search product to rival Google. Reuters reported on the potential delay of the project’s announcement. Although CEO Sam Altman clarified that Monday’s announcement was unrelated to the search engine, he expressed enthusiasm on social media about the upcoming reveal.
While there were minor glitches and unexpected responses during the demo, most features worked as intended. Murati highlighted OpenAI’s commitment to ensuring the new voice capabilities are used responsibly. However, details on safety measures concerning facial recognition and voice generation were not elaborate.
After newsletter promotion
Reports suggest that OpenAI is close to finalizing a deal to integrate its generative AI capabilities into Apple’s iPhone operating system. Bloomberg noted Apple’s intentions to enhance its AI resources by incorporating advanced features, possibly surpassing Apple’s Siri voice assistant capabilities.
As OpenAI expands its product offerings and explores collaborations, it faces legal challenges from publishers and media outlets alleging copyright violations. Entities like the New York Times claim that OpenAI used their content to train AI models without proper compensation or consent, leading to legal action seeking substantial damages.
Source: www.theguardian.com
