OpenAI has begun training a new flagship AI model in May 2024 to succeed GPT-4 that powers ChatGPT. It would like the new model to possess ‘the next level of capabilities’ as it strives to build AGI, a machine that can match a human brain. The new model, would be an engine for AI products (chatboats, digital assistants like Siri, search engines and image generators).

OpenAI is creating a new Safety and Security Committee to explore how it could handle the risks posed by the new model and future technologies.

The company would like to have a robust debate on capabilities and safety. OpenAI would like to accelerate the pace of research to move ahead of its rivals but is also conscious of the criticism that such a move could pose dangers for the humanity.

Experts differ on when the tech companies will reach AGI, but they have increased the power of AI technologies for more than a decade. There is a remarkable leap every two-three year. GPT-4 enables chatbot to answer questions, write emails, generate term papers and analyze data. GPT-4 was released in March 2023.

An updated version called GP-4o is not yet widely available. It has been released in May 2024. It can generate images and respond to questions and commands in a highly conversational voice It can learn skills by analyzing vast amounts of digital data — including sounds, photos, videos, Wikipedia articles, books and news stories