This one gives a lot of perspective over the fact that memory isn't just a plugin anymore. My take is, each approach has it's tradeoffs but …

This one gives a lot of perspective over the fact that memory isn't just a plugin anymore. My take is, each approach has it's tradeoffs but for me the mental model is, if the memory is more than or 80% dependent on inference, it's really just going to create untraceable behavior and failure points. By that I mean, presenting the right corpus to reason over. Separating each layer strictly from state awareness to reasoning is really important. This way, separating memory from the model providers would make sure the quality of the memory doesn't change and hence the reasoning over it.

Related

• a big part of agent harnesses is how they interact with context memory is just context its therefor impossible to separate harness from memo…
• For once I actually agree with the constructive argument here. There are more approaches that allow you to own your own destiny. Memory is j…
• This post is so so good. We are at this interesting point where we’ve started to really figure out ‘memory’ with these LLM-based systems. An…
• Directionally correct Open memory standards will need to emerge But it’s so early right now. We’re still just figuring out what best practic…
• This is the right frame. We’re currently designing our agent memory platform, and the hardest part isn’t storage — it’s deciding what to rem…
• The most important post I've read all week by @hwchase17 . Memory, particularly memory consistency is the biggest performance inhibitor to A…

Source: Harrison Chase (X) | 2026-04-12