It is interesting to compare this to LLMs - they also have the bounded context that you can see as the analogue to our working memory. It can contain enormously more bits of information than the 4 things the article says is the capacity of our working memory - but the 4 things can probably be much more complex internally - they are more like 4 pointers probably.
But at some level context engineering is very similar to what this article talks about.
This rings very true to me, and it's why I've been mildly obsessed for a decade plus with how to share mental models between people, and now LLMs, of any domain, be it technical, commercial, scientific or anything else. My inspiration was a book called Learning How To Learn by Novak, which TBH is so dry I'm not sure anyone I've recommended it to has actually finished it :) So then I point them to a talk here:
https://www.infoq.com/presentations/concept-map/
and an app to help render the shared mental model in plain text accessible to the LLM while providing visual interactivity to the humans here:
https://thinkingtools.software/concepticon/
I've learned that for both humans and models: system > willpower. The key is entirely in designing the environment.
For me personally, that means setting up 'attention getters' for the important things in life - 'totems' that force a context switch. For AI agents, it means well-designed CLI tools that help the agent orient itself in a task and pull exactly the 'context-for-the-job' it needs right then.
This is exactly what makes building modern GenAI decision-support systems so difficult. It's no longer just about finding the right software abstractions. You now have to account for the unknown cognitive construct of a completely different intelligence.
That is a good read, but as someone who tests in the first percentile for reading/listening, comprehension and retention, helps explain why many people who try to have discussions with me, shift there premise mid argument to fit there "evidence" and then cant remember where they started from, when reminded, and of course become agitated when challenged.
Right now I am puzzling over how to deal with a part time employee, who is addicted to this sort of disconected "style" of discourse, and am useing a disturb and observe approach,and as it seems to go unoticed, is informative in it's own right.
This reads to me as fully written by LLMs. Pangram agrees. Note the (alleged) author misHQ’s comments on this thread are getting downvoted as obvious slop.
Even if it were written by hand, it’s a very poor and frankly stupid essay about an interesting topic. “The model's attention is a fixed quantity, and it has to add up to one, so the more things you make it look at, the less of that attention any single earlier thing can keep.” This is borderline gibberish and it outright rejects the interesting question about LLMs and attention, namely that they have very different capacities from us. LLMs can read an entire OpenAPI schema in seconds and immediately construct valid requests from it. The article first points this out, and then switches to arguing that LLMs have similar limits to us. It’s completely incoherent.
I've been working like, almost this exact idea! https://github.com/hale-lang/papers/tree/main . The same capacity allocation bound algorithm appears naturally not only in human and llm/agent congnition, but in many natural systems as well.
To me this is optimization problem. How can we solve a problem if we don't understand it? Understanding takes a lot of effort exactly because our minds are wandering through useless context all the time, and not to mention interruption.
I formulate this problem as:
> Optimize for understanding
I know how to approach solving this exact problem. In fact, I've been doing exactly this since March 2026. We need to figure out how to isolate problems and context around them. And so my best bet right now is using graphs. Links can be easily added or removed between two nodes. And context is simply a group of links and/or linked nodes.
Now. What exactly is "understanding"?
To me, this is process when we look at some unpredictable, chaotic system and then creating structure from it. The chaotic system is an entangled, spaghetti-like graph. The ordered one is one we [hopefully] have in our brains, which allows us to act on it. I don't want to repeat entire article I wrote about this so if you want you can find it on my recent project (it's not ready for HN prime-time yet but I'll post demo soon).
But tl;dr we have "chaotic/unknown graph" and "structured/understood graph", and the bottleneck is moving nodes and links between these two.
The faster we understand the world around us, the better we understand why problems appear, and how to fix them. And once I realized this, to me it became clear where we need to move forward: to connect everything together in a way that makes understanding quick.
And fun fact, I already did this: I connected my article to yours.
16 comments
[ 2.0 ms ] story [ 41.8 ms ] threadBut at some level context engineering is very similar to what this article talks about.
For me personally, that means setting up 'attention getters' for the important things in life - 'totems' that force a context switch. For AI agents, it means well-designed CLI tools that help the agent orient itself in a task and pull exactly the 'context-for-the-job' it needs right then.
This is exactly what makes building modern GenAI decision-support systems so difficult. It's no longer just about finding the right software abstractions. You now have to account for the unknown cognitive construct of a completely different intelligence.
Tons of self-help books could be summarized in those 2 lines.
https://news.ycombinator.com/item?id=48706307
Even if it were written by hand, it’s a very poor and frankly stupid essay about an interesting topic. “The model's attention is a fixed quantity, and it has to add up to one, so the more things you make it look at, the less of that attention any single earlier thing can keep.” This is borderline gibberish and it outright rejects the interesting question about LLMs and attention, namely that they have very different capacities from us. LLMs can read an entire OpenAPI schema in seconds and immediately construct valid requests from it. The article first points this out, and then switches to arguing that LLMs have similar limits to us. It’s completely incoherent.
What I've heard is human short-term memory can hold seven things at once. Fortunately the mind is much more.
That's funny, isn't it the same for dogs?
To me this is optimization problem. How can we solve a problem if we don't understand it? Understanding takes a lot of effort exactly because our minds are wandering through useless context all the time, and not to mention interruption.
I formulate this problem as:
> Optimize for understanding
I know how to approach solving this exact problem. In fact, I've been doing exactly this since March 2026. We need to figure out how to isolate problems and context around them. And so my best bet right now is using graphs. Links can be easily added or removed between two nodes. And context is simply a group of links and/or linked nodes.
Now. What exactly is "understanding"?
To me, this is process when we look at some unpredictable, chaotic system and then creating structure from it. The chaotic system is an entangled, spaghetti-like graph. The ordered one is one we [hopefully] have in our brains, which allows us to act on it. I don't want to repeat entire article I wrote about this so if you want you can find it on my recent project (it's not ready for HN prime-time yet but I'll post demo soon).
But tl;dr we have "chaotic/unknown graph" and "structured/understood graph", and the bottleneck is moving nodes and links between these two.
The faster we understand the world around us, the better we understand why problems appear, and how to fix them. And once I realized this, to me it became clear where we need to move forward: to connect everything together in a way that makes understanding quick.
And fun fact, I already did this: I connected my article to yours.