OpenAI CEO Sam Altman says the next big advance in artificial intelligence will focus on memory, rather than improving reasoning skills. Current AI systems, including the latest models, are good at solving problems but cannot remember past interactions, meaning users often have to repeat information each time they use them.
Altman envisions AI that can retain long-term memory of user interactions, preferences, emails, documents, and conversations. This would allow AI to learn from past experiences, anticipate user needs, and provide more helpful suggestions without needing repeated instructions.
This shift to memory-based AI aims to create a personalized assistant that truly understands each user over time. Altman believes such AI could become available as early as 2026, moving beyond simple question-answering to proactive support in daily tasks.
At the same time, Altman acknowledges privacy and security concerns. He suggests that advanced encryption and safeguards will be necessary to protect sensitive user data.
With this focus, AI could evolve from a reactive tool into a trusted digital companion, offering smarter, more intuitive help tailored to each individual user’s life and work.
Also Read: B P Kanungo appointed IIFL finance chairman