OpenAI's New AI Agent Wants to Run Your Life, But It's Rough
Could an AI assistant handle your shopping and chores? Here's what happened when one tried.
OpenAI has launched Dots, AI agents that can handle online tasks for you—like buying a couch or booking a flight. You chat with them like a friend, and they use a virtual browser to get things done. In a test, the agent got the user's name wrong, misheard mumbling as 'I love you,' and failed to solve captchas. It did create a detailed couch-buying guide with prices, measurements, and links, but the experience was far from smooth.
These agents are part of a new wave of AI that's supposed to feel personal and proactive. They can run in the background and message you with updates. OpenAI charges $100 per month for access, while Meta offers a similar tool for free. You can connect your Gmail for more personalized help, but that raises privacy concerns—giving an AI access to your email could lead to embarrassing or risky mistakes.
The couch test showed both promise and problems. The agent asked for dimensions and budget, then sent a three-page packet with four couch options, including return policies and photos. It even addressed both the user and his partner. But the awkward moments—like the mistaken 'I love you'—highlight how these agents still struggle with basic human interaction. OpenAI says it's working on making them more reliable and less emotionally confusing.
For now, Dots are a glimpse of the future: an AI that can shop, plan, and organize for you. But at $100 a month and with frequent errors, they're not ready for prime time. As the technology improves, these agents could change how we use the internet, but for now, you might want to keep your credit card to yourself.
- OpenAI's new AI agents, Dots, can perform online tasks like shopping but often make mistakes.
- The service costs $100 per month, while Meta offers a similar free tool called Muse.
- Giving agents access to your email or accounts raises privacy and security concerns.
Why It Matters
AI agents could save time on chores, but current flaws and high cost mean they're not ready for everyday use.