Altman: OpenAI targets “infinite memory” upgrade by 2026
OpenAI chief executive Sam Altman said the next big leap in artificial intelligence will come from far stronger “memory,” not just better reasoning, as the company pushes to make its models more personal over time.
Altman described today’s memory tools as early-stage, saying OpenAI is still in a “GPT-2 era of memory,” and argued that future systems could remember a user’s preferences, context, and long-running goals at a depth humans simply cannot match. He said the direction OpenAI is excited about is “infinite, perfect memory,” where an AI can retain and use details across a user’s life to personalize help and recommendations.
He also linked that vision to a 2026 timeline, suggesting more advanced memory features are a near-term priority as OpenAI evolves ChatGPT beyond a chat window into a more continuous assistant.
The remarks come as major AI developers continue rolling out new model updates and product features, with personalization and longer-term context emerging as key areas of focus.
