CjHuber

7 ms·

1.
by CjHuber 2mo ago
Every good thing is alike, every slop is sloppy in its own way
2.
by CjHuber 2mo ago
Given that codex itself now ships a proxy that wraps the subscription, it seems unlikely https://github.com/openai/codex/blob/main/codex-rs/responses...
3.
by CjHuber 2mo ago
> Probably because the general purpose subagents inherit the parent model only if you don't specify which model should be used
4.
OpenAI's API can now keep reasoning across turns instead of discarding it(drop-05a4352b-803.sophisticated-stay.workers.dev)
3 points by CjHuber 2mo ago|0 comments
5.
by CjHuber 2mo ago
> Instead of requiring developers to script every step or passing every tool response back through the model, Programmatic Tool Calling in the Responses API can filter large amounts of intermediate data, retain only what matters, and ada
6.
by CjHuber 2mo ago
but 5.5 was also no major version bump
7.
by CjHuber 2mo ago
> fit a lot more work inside the model's context window before you're forced into a slow and lossy compaction step This is something I never understood. Why the reasoning is not included until the context is full, then the reas
8.
by CjHuber 2mo ago
Well, obviously its true as you can see in the silbing comment. But still interesting how experiences differ. I wonder did you ever use codex in more demanding workflows, because that at least something like that is going on is obvious if y
9.
by CjHuber 2mo ago
Interesting to hear people like gpt-5.5. For me it feels smart only at one shot prompts, but if you try to build up session context before doing something it feels magnitudes inferior to Claude. I'm almost sure its because the think
10.
by CjHuber 2mo ago
> so you now have to wait in line for hours Interesting, I go there pretty regularly and while It is well visited, I not once encountered a queue there
11.
by CjHuber 2mo ago
I agree, adaptive thinking is a pest and without a minimum thinking budget especially Claude for me currently defaults to not think at all even on max effort. Sequential-thinking was really a step in the right direction, and works almost ex
12.
by CjHuber 3mo ago
I think this is a really good cause but I can't help but wonder about the contrast between the puppy analogy and then increasing the friction of installing it by making it a paid tool
13.
by CjHuber 3mo ago
What about GBrain
14.
by CjHuber 3mo ago
Yeah indeed it's very interesting. And the 60k initial context don't even contain the suggested change yet. For me if I don't do this the current models tend to fixate and local patches instead of tracing symbols and making a
15.
by CjHuber 3mo ago
I'm always a bit confused when people say things like this. 60k token is often more than the initial context I feed the model with. And I don't think I ever had a productive session that began under 150k tokens.
16.
by CjHuber 3mo ago
So isn't that just purely security by obscurity then? Would they not rather have someone publicly break it instead of selling a zero day?
17.
by CjHuber 3mo ago
Technically yes
18.
by CjHuber 4mo ago
I‘m glad you didn’t write value
19.
by CjHuber 5mo ago
Are you… explaining the effects of being acquired to Hashimoto?
20.
by CjHuber 5mo ago
It does have an built in documentation subagent it can invoke but that doesn’t help much if they don’t document their shenanigans
21.
by CjHuber 5mo ago
This is exactly what also confused me. I had the exact same prompt in Claude code as well, and the no option implies you can also keep the whole history. But clicking keep apparently only ever kept the user and assistant messages not the wh
22.
by CjHuber 5mo ago
No of course it’s unrealistic for them to hold the cache indefinitely and that’s not the point. You are keeping the session data yourself so you can continue even after cache expiry. The point I‘m making is that it made me very angry that w
23.
by CjHuber 5mo ago
I think it’s crazy that they do this, especially without any notice. I would not have renewed my subscription if I knew that they started doing this. Especially in the analysis part of my work I don‘t care about the actual text output itsel
24.
by CjHuber 5mo ago
I‘d much rather have it go slooower and check more. Why faster it’s way too fast
25.
by CjHuber 5mo ago
I honestly am very disappointed with this. I've only learned about CLAUDE_CODE_DISABLE_ADAPTIVE_THINKING and showThinkingSummaries: true from this post. I've been wondering for a while where the summaries went and am always hoping
26.
by CjHuber 7mo ago
It's terrible that people think like that, especially in Georgia where they are still not tied to the debt fueled pyramid scheme that is the EU. They still think of Europe as how it was 20 years+ ago, they always only look at the surfa
27.
by CjHuber 7mo ago
I think your point doesn't hold up really. Telling an LLM to summarize something losslessly will loose so much more nuance than updating the plan directly every time when some useful information is gained. That file is not even a plan
28.
by CjHuber 7mo ago
I mean I agree the last couple of messages in a rolling window are good to include, but that is not really most of what happens in compaction, right? > there's often tiny tidbits in your prompts that don't get written to plans.
29.
by CjHuber 7mo ago
I honestly still don't see the point of compaction. I mean it would be great if it did work, but I do my best do minimize any potential for hallucination and a lossy summary is the most counterproductive thing for that. If you have it
30.
CLI tool for live communication between agents in different TUIs and subagents(github.com)
2 points by CjHuber 7mo ago|0 comments

More ›