Nobody can save Anthropic from themselves.

I’ve been experimenting with Qwen 3.8 27B and I believe I can totally use it as my main coding model provided I have the hardware for the full context. I don’t need my model to be opus level. I need it to do the tasks I want it to do without being an overprotective nanny.

Here, you dropped this: /s

I love the /s

With Claude I start having to limit the context I give it about my problem in case it trips up the safeguards. :(

5090 is plenty for the Q4_K_M quantized version of 3.6 27B with reduced context size.

I run it on a 3090(24GB) and 64k context using GGUF format and llama-cpp. Double 3090 gives you 128k, quad 3090 gets you to full context - 256k.

Or you can run quantized context, there's some degradation but it fits in a lot less.

Reminds me of a discussion about how cyber implants would bring upon cyberpunk - a person who needs an implant asks themselves: What if I use my implant this way…

> I wonder why we have yet to see more systematic exploration of Math.

> 60 Claude subagents

I don’t find that to be particularly systematic because it’s still so haphazard. It’s like asking Claude to vibe code you a website.

I was mostly highlighting the sheer cost of systematically exploring it. 60 agents doing something very narrow with a domain expert navigating the slop is already a high cost but I can see it being done more often in the near future since it just wasn’t possible a few years ago.

Ah, yeah that's a fair point. I was thinking it'd be something the labs themselves, or other companies with billions of dollars in funding, would tackle. The labs seemingly have the most to gain since such a system could be used for recursive self improvement.

Why are you hoarding the 3090s T.T They've jumped from $700 to almost double on eBay.

You can still learn from thieves and crooks. Especially if they break down what they did and how they did it.

Just keep in mind you might be the mark of his current scheme