OpenAI, Meta gear up for new AI models capable of reasoning

[Photo by MK DB]

OpenAI Inc. and Meta Platforms Inc. are on the brink of releasing advanced artificial intelligence (AI) models capable of thinking and reasoning.

Brad Lightcap, chief operating officer at OpenAI, told The Financial Times on Tuesday that the next generation of GPT would show progress on solving “hard problems” such as reasoning.

“We’re going to start to see AI that can take on more complex tasks in a more sophisticated way,” he said in an interview.

GPT-5 is expected to significantly enhance its capacity for understanding and inference.

While current AI models are limited to sentence generation, the next-generation models will be able to deduce riddles and more. Moreover, prospects include capabilities for voice recognition, image generation, text generation, and even video creation.

Meta also plans to initially release two versions of the smaller Lama3 next week.

“We are hard at work in figuring out how to get these models not just to talk, but actually to reason, to plan … to have memory,” said Joelle Pineau, vice-president of AI research at Meta.