OpenAI, supported by Microsoft, is developing a new multimodal AI model, potentially named GPT-5, which can process both text and visual inputs.

GPT-5 is anticipated to launch in mid-2024, but the official name and release date are not yet confirmed.

The model is currently in the training phase and will undergo extensive safety testing and 'red teaming' to address any issues prior to its public debut.

Improvements in GPT-5 aim to enhance the model's generalization abilities and minimize errors in output, commonly referred to as 'hallucinations'.

OpenAI's CEO Sam Altman has indicated the company's broader ambitions including a focus on robotics and the possible creation of a hardware startup for AI technologies.

Despite facing legal and controversy challenges, OpenAI is progressing in its efforts to advance the AI industry.