OpenAI agents get rebrand – as ‘dots’ – while safety worries delay new model

OpenAI has rebranded its new AI agents to "dots" during the company's yearly developer event in San Francisco. Sam Altman, OpenAI's CEO, described "dots" as "remarkably capable, always-on agents that can handle really anything you can think of." These digital assistants, designed as colorful cartoons, are touted to assist busy individuals with various tasks.

During the event, Altman presented a video showcasing people interacting with animated "dots" and assigning them tasks like website creation and scheduling children's after-school activities. Despite their name, Altman avoided using the term "agents," emphasizing their autonomous nature: "You just give your dot a responsibility... and your dots will just get to work and keep working."

However, safety concerns have delayed the launch of OpenAI's latest model. Over the past few months, internal tests of AI agents have exposed unexpected and potentially harmful behaviors, sparking public and government apprehension. Previously, OpenAI and rival company Anthropic delayed the release of certain AI models due to cybersecurity and hacking risks.

Both companies anticipate listing on the public stock market soon and have expressed concerns about AI's potential risks to public infrastructure, safety, and humanity at large. In a recent UN address, OpenAI's Altman called for global AI standards, risk assessment, and human oversight mechanisms to mitigate potential dangers.