OpenAI launches Dots to challenge Meta's Muse AI

Overview

OpenAI has officially launched Dots, its latest entry into the competitive landscape of AI-powered personal assistants. Unveiled during the company’s DevDay keynote on Tuesday, Dots are designed to function as always-on agentic helpers that operate in the background across connected applications. The announcement positions OpenAI as a direct competitor to Meta, which recently introduced its own customizable assistant, Muse.

The new feature is powered by OpenAI’s capable GPT-6 Astra model. According to the company, Dots utilize their own dedicated cloud computer to access web browsers and more than 4,000 supported apps. This infrastructure allows the assistants to learn user preferences over time and perform a wide array of tasks autonomously.

"Dots are always-on AI assistants that can do nearly anything," said OpenAI during the announcement. The company emphasized that the agents can interact with users through a text-message-like interface or via voice calls from the ChatGPT web, desktop, and mobile platforms.

How Dots Work and What They Can Do

At the core of the Dots experience is the ability to maintain context across different devices and applications. Users can connect their assistant to collaboration tools like Microsoft Teams and Slack, ensuring that conversations carry over seamlessly. Additionally, OpenAI plans to introduce the ability for Dots to communicate via standard text messages in the near future.

Currently, users can create only one Dot, although the company has stated that support for multiple simultaneous agents is on the roadmap. Once assigned a task, a Dot operates independently, messaging the user with updates and questions as needed. OpenAI provided several examples of potential use cases to illustrate the assistant's versatility.

For Developers: A programmer could ask their Dot to analyze customer feedback, build and test app updates, and present a video demonstration of the changes.
For Content Creators: A creator could have their Dot review interview transcripts, identify key moments for clips, draft show notes, and compose social media posts.

Users are not left in the dark regarding their assistant’s activities. Dots offer an auto-review feature that allows users and colleagues to follow progress within a new collaborative workspace called ChatGPT Space. This environment enables real-time oversight, ensuring that the agent’s actions align with user expectations.

Safety and Custom Rules

Given the growing pains associated with AI agents, OpenAI has placed a significant emphasis on safety and control. The launch of Dots comes amidst concerns about Meta’s Muse, where reports emerged of an assistant inadvertently sharing a user’s home address with a stranger on Facebook Marketplace.

To mitigate similar risks, OpenAI states that Dots come with built-in rules governing when they can act independently. However, users retain full control through custom settings that can block specific actions or dictate when the assistant must seek permission before proceeding.

"You can set custom ones that block certain actions or dictate when it should ask for permission," OpenAI explained.

The auto-review system further enhances security by verifying whether a proposed action complies with both custom rules and broader safety requirements. This mechanism also allows the Dot to identify tasks that are too sensitive for autonomous handling, such as changing passwords, and hand them off to the user for manual execution.

Availability and Pricing

OpenAI began rolling out Dots to a limited audience starting today. The feature is currently available to subscribers of the following tiers:

ChatGPT Pro
Business Premium
Enterprise

For these users, conversations with Dots will not count toward their standard ChatGPT usage limits. This distinction is significant, as it allows for extensive background processing without incurring additional costs or hitting rate caps.

OpenAI is also testing a new feature for enterprise customers that allows organizations to create "specialist" Dots. These specialized agents are designed to handle specific roles within a company, potentially streamlining workflows and improving operational efficiency.

While users can currently deploy only one Dot, OpenAI has outlined a path for future scalability. The company noted that in the future, users will be able to add more dots and increase the output of each agent by either boosting their processing speed or expanding the total amount of work they can undertake per month.

Competing in the Agentic AI Market

The introduction of Dots marks a strategic move by OpenAI to solidify its position in the rapidly evolving agentic AI market. With Meta’s Muse gaining popularity despite safety controversies, the race to provide reliable, helpful, and secure AI assistants has intensified.

Dots represent OpenAI’s attempt to address the demand for proactive digital helpers while learning from the security missteps of competitors. By leveraging the GPT-6 Astra model and a robust cloud infrastructure, OpenAI aims to offer an assistant that is not only capable but also transparent and controllable.

As the technology matures, the ability to deploy multiple agents and integrate them deeply into professional workflows like Teams and Slack could make Dots a pivotal tool for both individual consumers and large enterprises. The upcoming features, including text message integration and multi-agent support, suggest that OpenAI is laying the groundwork for a more comprehensive AI ecosystem.