OpenAI has launched ChatGPT dots (a personal cloud-based digital assistant), priced from $100 to $500 per month. This isn't another chatbot—it's the first consumer-grade AI that attempts to retain your long-term work context and proactively "catch errors" for you. Powered by GPT-6 Astra, it can open browsers, handle files, run software, and parallelize research, document drafting, and coding across multiple tasks.
What This Is
The key difference from regular ChatGPT is persistent memory—it carries forward the context, requirements, and preferences you share across sessions. One detail in the official demo deserves highlighting: an OpenAI employee was prepping slides, and at 6:56 AM, dots proactively flagged that the figure on page 26 didn't match pages 12 and 37; a calendar invite was missing an attachment and had an inconsistent title. This is the ability to cross-reference different materials, surface problems, and bring them back to you for confirmation—not a simple conversational response.
Industry View
We're seeing two schools of thought. Positive voices argue that dots pushes AI from "answering questions" to "continuously owning a task"—a substantive upgrade in product form. The pricing tiers (Pro plans at $100/$200/$500/month) show OpenAI testing the willingness-to-pay ceiling among high-value users.
But risks and questions remain. First, the $500/month price means this is a product "workers can't afford, only enterprises will consider"—running counter to the narrative of Agent (an intelligent system capable of autonomously executing multi-step tasks) democratization. Second, persistent memory means continuous uploads of personal documents, and data security and privacy boundaries remain uncleared publicly. Third, over-reliance on AI for cross-device management may cause human verification skills to atrophy—the hidden cost of shifting review responsibility from AI back to humans is widely underestimated.
Impact on Regular People
For enterprise IT: dots won't make procurement shortlists in the short term, but its "persistent memory + proactive comparison" design philosophy will pressure domestic tech giants to accelerate their competing products.
For working professionals: the practical use case is more "information organization + reminders" than "decision replacement." Starting at $100/month, it's suited for small-team leads or freelancers who treat AI as a long-term assistant and can decompose tasks with precision.
For the consumer market: the emergence of AI assistants at the thousand-yuan monthly tier shows AI products are stratifying—basic conversation free, deep service paid. If this model validates, it will reshape user expectations around "what AI is worth."