Agents
OpenAI Launches Dots: Always-On Autonomous Agents
WIRED AIhigh confidence
Why it matters
- →Dots represent OpenAI's strategic pivot from reactive chatbots to proactive, always-on autonomous agents.
- →The GPT-6 Astra model powers these agents, indicating a specialized architecture for continuous task execution.
- →Restricting access to Pro subscribers highlights the high compute costs associated with running persistent AI assistants.
OpenAI launched Dots at its annual DevDay, introducing a new class of proactive AI agents powered by the GPT-6 Astra model. Designed to operate continuously, Dots can autonomously execute tasks such as drafting emails, planning trips, and managing schedules without constant user prompting. The service is currently exclusive to ChatGPT Pro subscribers ($100/month), positioning it as a premium alternative to competitors like Meta’s free Muse agent. While the technology demonstrates improved reliability in computer interaction compared to previous iterations, OpenAI faces challenges in establishing user trust regarding data access and proactive behavior.
Read originalThe story around this
Earlier coverage that leads up to this article, and what followed. Lines connect each piece to the closest one after it, converging here.
