The UX problem is elsewhere I think. Many users probably don't realize that the agent's context window is limited, and that clever compaction is happening regularly to make it seem infinite. But that necessarily means the agent has to forget stuff.

As a result, users will keep reusing the same coding or chat session again and again. While it would be better to start fresh for unrelated tasks.

▲

whstl 2 hours ago | parent | next [-]

I don't believe this is a context problem.

Claude Opus 4.7 has a very large context compared to itself, but IME it is the worst at following instructions, and completely disregards the (small) preferences prompt, even in the first or second message, even if the messages are just a few characters long.

IMO this is entirely a training problem.

▲

8note 2 minutes ago | parent | next [-]

if you look at claude code, it now says compaction is happening constantly, which is likely why

▲

apsurd 2 hours ago | parent | prev | next [-]

Isn't a large context window still a problem though? At the upper bound, the more you put in the more each sentence washes out within that window?

▲

whstl 2 hours ago | parent [-]

I’m not talking about large amounts of text, I’m talking about a couple sentences back and forth.

It disregards things like “no follow up questions”.

Haiku, for example doesn’t.

This bias is a very human thing, actually now that I think about it. You just disregarded the “even if the messages are just a few characters long”. :)

	▲	apsurd an hour ago \| parent [-]
		haha! yes i read too fast but i did read it and i took "message is small" to mean the message you want followed within the large context, not the entire context is just a small message. funny though it is a case in point: language is hard. and i get to hide behind being "preoccupied" . i wonder if llms have their own sense of preoccupation hmmm.

▲

stingraycharles 2 hours ago | parent | prev [-]

[dead]

▲

poly2it 2 hours ago | parent | prev [-]

The author of this post and the readers of this thread probably do understand context window limitations, but are frustrated nonetheless.

	▲	dust-jacket a few seconds ago \| parent [-]
		Well yeah. And there's little more frustrating than someone telling you not be frustrated because "that's just how it works". We get how it works. It's just irritating.