[kimixa]

Man I hope so - the context limit is hit really quickly in many of my use cases - and a compaction event inevitably means another round of corrections and fixes to the current task.

Though I'm wary about that being a magic bullet fix - already it can be pretty "selective" in what it actually seems to take into account documentation wise as the existing 200k context fills.

[humanfromearth9]

Hello,

I check context use percentage, and above ~70% I ask it to generate a prompt for continuation in a new chat session to avoid compaction.

It works fine, and saves me from using precious tokens for context compaction.

Maybe you should try it.

[pluralmonad]

How is generating a continuation prompt materially different from compaction? Do you manually scrutinize the context handoff prompt? I've done that before but if not I do not see how it is very different from compaction.

[robertfw]

I wonder if it's just: compact earlier, so there's less to compact, and more remaining context that can be used to create a more effective continuation

[nickstinemates]

Is this a case of doing it wrong, or you think accuracy is good enough with the amount of context you need to stuff it with often?

[romanovcode]

In my example the Figma MCP takes ~300k per medium sized section of the page and it would be cool to enable it reading it and implementing Figma designs straight. Currently I have to split it which makes it annoying.

[kimixa]

I mean the systems I work on have enough weird custom APIs and internal interfaces just getting them working seems to take a good chunk of the context. I've spent a long time trying to minimize every input document where I can, compact and terse references, and still keep hitting similar issues.

At this point I just think the "success" of many AI coding agents is extremely sector dependent.

Going forward I'd love to experiment with seeing if that's actually the problem, or just an easy explanation of failure. I'd like to play with more controls on context management than "slightly better models" - like being able to select/minimize/compact sections of context I feel would be relevant for the immediate task, to what "depth" of needed details, and those that aren't likely to be relevant so can be removed from consideration. Perhaps each chunk can be cached to save processing power. Who knows.

[IhateAI_2]

lmao what are you building that actually justify needing 1mm tokens on a task? People are spending all this money to do magic tricks on themselves.

[kimixa]

The opus context window is 200k tokens not 1mm.

But I kinda see your point - assuming from you're name you're not just a single purpose troll - I'm still not sold on the cost effectiveness of the current generation, and can't see a clear and obvious change to that for the next generation - especially as they're still loss leaders. Only if you play silly games like "ignoring the training costs" - IE the majority of the costs - do you get even close to the current subscription costs being sufficient.

My personal experience is that AI generally doesn't actually do what it is being sold for right now, at least in the contexts I'm involved with. Especially by somewhat breathless comments on the internet - like why are they even trying to persuade me in the first place? If they don't want to sell me anything, just shut up and keep the advantage for yourselves rather than replying with the 500th "You're Holding It Wrong" comment with no actionable suggestions. But I still want to know, and am willing to put the time, effort and $$$ in to ensure I'm not deluding myself in ignoring real benefits.