AI Memory

OpenAI says V7 uses GPT-5.6 to turn scattered company files into source-linked context for agents, pointing to a new model for enterprise memory.

Google Research’s WikiSkill lets AI agents retain records of failures and successes, improving repeat-task performance without retraining the underlying models.

Meta researchers propose a selective memory agent for long-running AI tasks, reporting higher benchmark scores while highlighting costs, calibration, and open design questions.

AgenticSTS researchers say structured memory helped an AI agent beat Slay the Spire 2 while cutting token use, highlighting a practical path past context rot.

Micron Technology has commenced high-volume production of HBM4 memory chips for NVIDIA's next-generation Vera Rubin GPU, positioning the company at the center of an AI-driven memory shortage that analysts project could last until 2030, while Micron's revenue nearly tripled year-over-year.

Anthropic rolls out a free memory feature for all Claude users and introduces a ChatGPT data import tool, capitalizing on a wave of users switching from OpenAI following the Pentagon deal controversy.

Micron's market cap surges 313% as AI data centers consume 70% of memory chip production, with DRAM prices expected to rise 70-100% in 2026 amid persistent shortages.

OpenAI CEO Sam Altman believes the next significant leap in AI will be the development of 'infinite, perfect memory,' allowing AI to remember every detail of a user's life.