OpenAI promised users they would eventually be able to interact with the Dot agent through text commands after its launch last month. Now, nearly a month since the release, the company has begun promoting a preview feature among professional users in the United States that allows users to achieve this. According to OpenAI, with this new feature, users can send text, photos, and even voice memos to Dot. Additionally, Dot can proactively contact users to provide updates on tasks assigned to it.
To set up Dot's text feature, users need to open their profile in ChatGPT and select "Text." On mobile, if using an iPhone, users must ensure iMessage is enabled; for Android devices, RCS must be enabled. OpenAI stated that it partnered with Linq to provide messaging support and conducted "extensive" penetration testing on this feature to ensure its security. The company added on X: "Only you can text your Dot, and your Dot can only text you."
OpenAI Dot agent's text feature will enhance user interaction experience
Any information written into Dot's memory will be shared across ChatGPT and text, meaning context is coherent across different conversations. The market for personal assistants has suddenly heated up following Anthropic's launch of Claude Duet last week. In my testing, OpenAI's Dot appears to be relatively more mature. I believe that being able to communicate with them via text will be welcomed by some users. Currently, using Claude Duet requires joining a waitlist that is only open to users in the United States.
Anthropic stated that it plans to offer the agent to its monthly US$20 (approximately HK$156) professional plan subscribers, which is something Dot currently cannot achieve. They are not currently available to Plus tier users.