Ethan Mollick says personal AI agents like Dot lack task and memory transparency

Ethan Mollick says personal AI agents such as Dot are hard to use because users cannot easily see what the agent is working on, what it knows about them, or what it will do for them.

In an X post on October 30, 2026, he calls this a user experience problem that applies to personal agents in general, and he endorses advice about working with Dot. The post is the only report on this story, and it gives no details about Dot's features or any response from its maker.