also poorly organized and formatted.

10ms? Good grief that’s slow.

Still faster than using a remote resolver: https://www.dnsperf.com/

But a remote resolver, such as the ones on the list, is not used to determine whether an item is contained in a set (0.0.0.0) but rather to retrieve the current IP of that domain.

The way Pi-Hole is used is to first determine whether a DNS should be blocked and only if it's not in that set, forward it to a remote resolver. I'd think the local resolver should be at least an order of magnitude less than the remote to justify it.

yeah but it doesn’t replace the remote resolver, in most setups PiHole is another layer on top of the remote resolver

my remote resolver roundtrips in 6ms. this project is a joke

Treating pleading the Fifth as evidence of guilt is exactly the kind of unfairness the Fifth Amendment is meant to protect against.

“first-order approximation” is pretty different from first-principles thinking in my opinion.

my gut reaction was to argue over semantics with you but no, you're right they are different. I don't think it's been the case in the overwhelming majority of my cases but you aren't wrong.

Arguably they already have, GLM-5.3 is insanely good at offensive security tasks, especially for the price.

what do you mean by reset at every turn? context stays until compaction. if you remove the reasoning tokens after every turn you will be constantly blowing cache which is far worse than filling up context.

That's not my understanding of how most agents work. This is what a chain of request/response looks like:

  Your Prompt 1: Prompt Content 1 -> cache-1
  LLM Response 1: Thinking Content 1 Response Content 1
  Your Prompt 2 (client side): prompt-1 + response-without-thinking-1 + Prompt Content 2
  Your Prompt 2 (server side): cache-1  + response-without-thinking-1 + Prompt Content 2 -> cache-2
  LLM Response 2: Thinking Content 2 Response Content 2
  Etc...

So reasoning gets dropped from context and you still get cache from the accumulating requests.

Edit:

I've realised I was incorrect, the thinking doesn't get passed back and forth but the latent snapshot does which result in using memory just the same.

Modern protocols loop back the reasoning tokens in raw textual form via an encrypted parameter. You can't see them (modulo the recent attack), but you do resubmit them.

Yeah I've done more research and that's what I meant in the edit you replied to.

But that's not the full reasoning token context, just a snapshot of the latent state at the end of it, no?

Have a look at the gemini ones they're pretty small.

There was an attack that convinced models to leak the contents of the reasoning tokens, and it came back as text that matched the length of the encrypted data very well. So it's probably still tokens.

Looping back latent state isn't that easy. The hidden data is the entire contents of the KV cache which can be massive. I don't think any provider is trying to loop the KV cache through the client, and neural compressions of the reasoning would be lossy / an advanced technique that is still firmly in the realm of research papers, as I understand.

The creator of Nix soft-forked it and started a company, that sort of inevitably leads to bad vibes I think

Can’t say I blame him, given the tone of some of the convos on the discourse site.

The fork occured before most of the drama.

ChatGPT has over a billion monthly active users.

I am an active user of Claude but I do not live my life through that thing. I try to get it to answer questions that it does not want to answer. The only questions it seems to answer without hesitation are technical in nature. In the entire existence of my account I have started 9 chats.

More interesting to me was that Alexa on Amazon Prime answered all my questions without hesitation or disclaimers but once I have my answers I close that tab.

This is definitely a 2026 type comment, but I'm very surprised that it's so "low." GPT is the most well known brand and it's the one that seemingly 100% of kids use to "assist" on their homework and other exercises from grade school to college. One would think that alone would already take you well over a billion.

This does not mean that all billion of them are using ChatGPT as their main interface with the internet.

I am monthly active user, absolutely do not experience web through ChatGPT alone and frankly, dont mind ChatGPT being cut off stuff. AI companies forcing themselves as a middle man is not a good thing.

Whoever using purely agentic worflow is not my concern. Just like people who read only physical book, learn only from podcasts or watch only netflix movies.

There is zero reason for the rest of us to worry about filling free data to companies that try to make themselves monopolies.

They probably should rewrite it in Rust, but also, I think the Bun guy wanted to rewrite it in Rust because he thinks Rust is a better choice for Bun, independent of Bun being used by Claude Code.