OpenAI does not currently teach GPT-5

Experts calling for a pause in AI development will be happy to hear that OpenAI is not currently training GPT-5.

OpenAI CEO Sam Altman spoke remotely at an MIT event and was interviewed about AI by computer scientist and podcaster Lex Friedman.

Altman confirmed that OpenAI is not currently developing a fifth version of its Generative Pre-trained Transformer model and is instead focusing on extending the capabilities of GPT-4, the latest version.

Altman was asked about an open letter urging developers to pause training AI models that exceed GPT-4 for six months. While he supports the idea of ensuring AI models are safe and consistent with human values, he believes the letter lacks technical nuance as to where the pause should be made.

"An earlier version of the letter claims that we are training GPT-5 right now. We are not, and we won't be doing that for some time. So in that sense it was silly." - Altman said.

"We're doing something on top of GPT-4, which I think has all sorts of security issues that we need to address."

GPT-4 is a significant improvement over its predecessor, GPT-3, which was released in 2020.

GPT-3 has 175 billion parameters, making it one of the largest language models in existence. OpenAI has not confirmed the exact number of GPT-4 parameters, but it is estimated to be around one trillion.

The OpenAI blog says that GPT-4 is "more creative and collaborative than ever before" and "can solve complex problems with greater precision, thanks to its broader shared knowledge and problem-solving abilities."

GPT-3.5 scored in the top 10 percent on the simulated bar exam. GPT-4, however, passed the exam among the top 10 percent.

OpenAI is one of the leading AI research labs in the world, and its GPT models have been used for a wide range of applications including language translation, chatbots, and content creation. However, the development of such large language models has raised concerns about their security and ethical implications.

Altman's comments indicate that OpenAI is aware of the problems associated with GPT models and is taking steps to address them.

While GPT-5 may not be on the horizon, the further development of GPT-4 and the creation of other models based on it will undoubtedly raise new questions about the safety and ethical implications of such AI models.