Perplexity has introduced a new self-improving memory system called Brain for its AI Computer platform, a move the company says could reshape how users interact with artificial intelligence. The system, unveiled this week, is designed to remember past interactions and adapt its responses over time, making personalized experiences more seamless and workflow tasks more efficient.
What Brain Does
Brain acts as a persistent memory layer for Perplexity’s AI Computer. Instead of starting fresh with every query, the system retains context from previous conversations and actions. It learns from user behavior and automatically refines its understanding, aiming to reduce repetitive inputs and speed up complex tasks. The company describes it as a self-improving mechanism that gets smarter the more it's used.
The core promise of Brain is better personalization. By remembering a user's preferences, past projects, and frequent requests, the AI can tailor answers and suggestions without being told the same details twice. For professionals juggling multiple workflows — research, data analysis, content generation — that could mean fewer interruptions and faster results. Perplexity is positioning the system as a tool to optimize efficiency rather than just answer questions.
The AI Computer Platform
Brain is the latest addition to Perplexity’s AI Computer, a platform that already combines search, reasoning, and task automation. The platform allows users to run complex multi-step operations, and Brain adds a layer of continuity that was previously missing. While the company hasn't released detailed performance benchmarks or a public launch date for the memory feature, early demonstrations show it pulling context from hours-old sessions.
Perplexity has not yet announced when Brain will be available to all users or whether it will be limited to specific subscription tiers. The feature is currently being tested internally, and the company is expected to share more details on rollout and pricing in the coming weeks.
