Saad Ullah
⬤ Perplexity just dropped a major upgrade for its Comet AI assistant, rolling out a new Memory feature that seriously expands how the tool understands what you're working on. The update lets Comet pull information from past chats, active projects, browser tabs, and other personal data sources—basically giving it a much wider view of your workflow. The company says this makes interactions feel more natural and relevant, since Comet can now pick up on real-time context without you having to explain everything from scratch.
⬤ The Memory system works by recognizing what you've done before and folding that into current conversations. That means no more repeating yourself or restarting explanations every time you open a new chat. Comet can now surface info from documents you're working on, tabs you have open, or tasks you've handled recently, creating a much stronger contextual foundation. The tool also learns your preferences and strengths over time, so responses get more aligned with how you actually work.
⬤ This rollout puts Perplexity in a different league compared to most AI assistants that only look at what's happening in the current session. By tracking activity across projects and browser tabs, Comet can handle more complex, multi-step tasks without losing the thread. Early feedback has been positive, with users highlighting how Memory cuts down on repetitive prompts and makes research flows way smoother. Perplexity has hinted that this is just the beginning, with deeper integrations and features likely on the way.
⬤ Memory gives Perplexity a real edge in the increasingly crowded AI assistant space, where the ability to remember context and stay useful over time is becoming a major competitive factor. With cross-project awareness and tab tracking, Comet is now positioned as a solid tool for productivity, research, and content work. As AI assistants push harder into long-term retention and workflow support, Perplexity's latest move shows where the industry is heading—toward systems that stick with you across extended tasks and adapt as you go.
Saad Ullah
