Lessons from Studying Every Memory System — Shlok Khemani, Independent
Shlok Khemani analyzed how different AI systems build memory of their users.
He started when ChatGPT incorrectly claimed he had traveled to Turkey — something he never did, only discussed as an alternative to Thailand. The issue wasn't the error itself but that the system didn't question the contradiction or verify his email records. ChatGPT uses a "running profile" that updates in the background with approximately 4,000 tokens of densely packed information that users cannot see without hacking in. Claude started differently, without a profile but with two search tools in old conversations, later adding a third. Khemani argues that memory systems fundamentally deal with computational cost: a large profile is expensive to maintain and costly each time it's loaded into the conversation, so different systems solve the same budget problem in different ways.
Vibekollen prepared this summary with AI from the original publication. The content belongs to AI Engineer.
More to read
Lumaril, which develops AI software for optimizing industrial production by connecting to existing control systems, raised a $66M Series A at a $400M valuation
Techmeme 3 h ago
DeepSeek open sources Super-Agent as a research prototype, a proactive coding agent it says can monitor group chats for coding issues and suggest solutions
Techmeme 4 h ago
Insurance software maker Guidewire agrees to acquire Hyperexponential, which makes AI-powered insurance underwriting and pricing software, for $1.7B in cash
Techmeme 4 h ago
One word for AI agents in 2027
Nate Herk 8 h ago