Holding Off The next big thing in the world of artificial intelligence are so-called "AI agents" — models that are capable of interacting with their environment, like a computer desktop, allowing them to autonomously complete tasks without human intervention.

Indeed, Musk suggested that synthetic data — data generated by AI models themselves — is the path forward. “The only way to supplement [real-world data] is with synthetic data, where the AI creates [training data],” he said. “With synthetic data … [AI] will sort of grade itself and go through this process of self-learning.”

Former OpenAI employee Suchir Balaji was found dead from an apparent suicide in November, but his mother is calling foul play.

Elon Musk's attorney urged state law officers to establish a competitive bidding process for OpenAI's charitable assets amid the startup's efforts to remove nonprofit control. Musk is suing OpenAI over its restructuring into a for-profit entity,

Nvidia, Google and hot startup OpenAI are turning to "synthetic data" factories amid demand for massive amounts of data needed to train artificial intelligence models.

The OpenAI CEO recently published a blog post reflecting on AI progress, as well as his brief ouster from the company.

Sam Altman teased that the AGI and superintelligence are coming to ChatGPT soon, but we don't even have the next big GPT-5 upgrade.

OpenAI on Friday outlined plans to revamp its structure, saying it would create a public benefit corporation to make it easier to "raise more capital than we'd imagined," and remove the restrictions imposed on the startup by its current nonprofit parent.

Phi-4 and an rStar-Math paper suggest that compact, specialized models can provide powerful alternatives to the industry’s largest systems.

To demonstrate we are still not at human-level intelligence, Chollet notes some of the simple problems in ARC-AGI that o3 can't solve. One such problem involves simply moving a colored square by a given amount -- a pattern that quickly becomes clear to a human.

Red teaming has become the go-to technique for iteratively testing AI models to simulate diverse, lethal, unpredictable attacks.