In recent years, more and more people have become accustomed to using AI (LLM) to handle various problems and have gradually gotten used to treating AI as their "second brain." It can answer questions, write, code, analyze, and give advice. Often, its performance is even more stable and efficient than that of humans. Most of the time, we think it is omnipotent.
However, after deep usage, some strange issues begin to surface. For example, it can appear very intelligent in specific parts but lacks consistency overall. Sometimes it grasps the key points, while other times it gets stuck on details; sometimes its judgment is clear, and other times it seems to lose direction.
And, of course, there's the unavoidable fact that it often "fabricates" information.
These issues are inherent flaws. After all, for a probabilistic model, all its outputs are, philosophically speaking, fabricated—they are guesses. But if you are a deep, AI-first user, you'll find that these problems arise not just because of its flaws. The more fundamental issue is that it is not a true "silicon-based brain" and lacks the ability to think.
LLM-type AIs excel at processing information, but thinking is more than just processing information. The foundational prerequisite for thinking is the ability to distinguish the "importance and urgency" of matters. When humans make judgments, there is always an implicit prioritization—a pre-processing of known information to determine what to prioritize, what to ignore, and what to defer. This prioritization is implicit.
In AI, this layer is missing. That's why we see familiar phenomena: key premises gradually fade in long conversations, irrelevant details are repeatedly amplified, and the focus of each response drifts. This isn't just a misunderstanding; it's more like there's no stable priority system supporting the output.
Text is a tool for disseminating information through imagination. Therefore, if the sender lacks the ability to describe facts accurately, and the receiver lacks imagination or has an overactive imagination, the entire information transmission process becomes a disaster.
LLM is a system that relies solely on text for reconstruction. What it receives, from the very beginning, is compressed information. What it outputs is a high-probability restoration based on incomplete information.
In this process, humans use experience, feedback, and real-world constraints to correct their understanding. But AI lacks these; it can only continue making inferences within the textual space. So sometimes it appears very accurate, while other times it deviates completely. This is because it is always using limited information to imagine a world. Humans use imagination to compensate for the shortcomings of language, while AI has only imagination.
When OpenClaw was trending recently, I didn't jump on the bandwagon, even though I've been following it since its inception. I understand very well that its most striking feature, the one that caused a sensation, is at the human-computer interaction layer. A concrete character image and conversational interaction give OpenClaw the potential to provide services to ordinary people. But at the same time, relying entirely on conversational text interaction also magnifies the flaws of the underlying LLM core.
Current various Agents lack a reliable memory system.
Human judgment can gradually converge because memory is constantly at work. The decisions you've made, the pitfalls you've encountered, and the preferences you've formed are repeatedly invoked in subsequent judgments and gradually reinforced, eventually forming a relatively stable cognitive structure. This structure isn't perfect, but it gives your judgment continuity.
Most currently popular Agents rely on markdown or so-called RAG as their memory system, but the flaws are very obvious. I won't bother discussing RAG. The biggest problem with markdown is that it's just a storage format, not a cognitive structure. People can use it to record information, but they can't express relationships. Moreover, it's static; once written, the content hardly changes. A judgment that has been overturned and a conclusion that remains valid are essentially no different within the system. It won't be corrected or phased out; it just keeps accumulating.
This type of memory system only lets AI see more information; it doesn't enable AI to truly form judgments. What ultimately determines the quality of judgment is never the amount of information, but how information is filtered, organized, and forgotten.
An Agent lacking a reliable memory system is just a more user-friendly information manipulation tool; it doesn't solve the problem of how to use the tool well.
We can see that many people in research or AI-first insiders can use AI extremely effectively. They can provide AI with ample and rigorous contextual constraints, keeping the output results within a controllable range. This shows that, purely in terms of model capability, it already meets the needs of the vast majority of people. Therefore, what we should think about is how to make AI truly serve non-professional users.
Giving AI a brain isn't about stuffing it with more context; it should be about giving it a multi-dimensional, lifecycle-aware memory system.