Hacker News: sheeshkebab

New comment by sheeshkebab in "Open source AI must win"

sheeshkebab — Sat, 13 Jun 2026 02:41:35 +0000

Qwen models are actually very competitive with frontier models, and you can run them on your local computer. Gotta have a decent graphics card and by that time the current cost of the rig may not justify it over paying $100/month for cloud model but it’s all out there.

New comment by sheeshkebab in "Statement on US government directive to suspend access to Fable 5 and Mythos 5"

sheeshkebab — Sat, 13 Jun 2026 02:10:11 +0000

Well, it was great while it lasted - I had fable build me a bunch of stuff this week that opus was just screwing up too much and could never finish. Good thing there are plenty of choices now even if US gov fucks up US AI.

New comment by sheeshkebab in "AWS Bedrock to require sharing data with Anthropic for Mythos and future models"

sheeshkebab — Wed, 10 Jun 2026 11:49:37 +0000

Pro/max subs are not as flexible as bedrock in api use and don’t seem to run the same models either - often times they are notably dumber (quantized I guess) than bedrock equivalent.

New comment by sheeshkebab in "Claude Fable 5"

sheeshkebab — Tue, 09 Jun 2026 23:55:05 +0000

I’ll ask it to write me some win32 ui crap when I get hands on it, it will need all its brainpower to get that idiocy right.

New comment by sheeshkebab in "MiMo-v2.5-Pro-UltraSpeed: 1T model with 1000 tokens per second"

sheeshkebab — Tue, 09 Jun 2026 01:37:35 +0000

Opus regularly bitches and wines to me how long something will take and that I should think before asking it to do it. But then it does it anyway in 15 minutes.

New comment by sheeshkebab in "How LLMs work"

sheeshkebab — Sat, 06 Jun 2026 03:21:58 +0000

considering they work with any architecture/configuration given enough compute, just more or less efficiently - then maybe it's fundamental, in the same sense as why electricity works...

New comment by sheeshkebab in "MAI-Code-1-Flash"

sheeshkebab — Wed, 03 Jun 2026 00:30:09 +0000

I like gpt oss - great model even if not too smart.. runs on my laptop at over 100ts has a certain tone that I like over all these qwens stuck up their asses.

New comment by sheeshkebab in "Liquid AI reveals 8B-A1B MoE trained on 38T"

sheeshkebab — Sat, 30 May 2026 15:40:21 +0000

27b is slow as molasses vs 35b on local stuff I have (m5 max). Mtp doesn’t make any difference either.

New comment by sheeshkebab in "Constraint Decay: The Fragility of LLM Agents in Back End Code Generation"

sheeshkebab — Sun, 24 May 2026 15:40:46 +0000

…but they reason well enough given enough context (using their matmuls).

New comment by sheeshkebab in "Electrobun 2.0 will be decoupled from Bun due to the Rust rewrite"

sheeshkebab — Sat, 23 May 2026 16:16:07 +0000

One of these days you’ll learn about “enterprise” code

New comment by sheeshkebab in "Electrobun 2.0 will be decoupled from Bun due to the Rust rewrite"

sheeshkebab — Sat, 23 May 2026 14:51:56 +0000

Mist of human written codebases are unusable for llm dev by that definition.

New comment by sheeshkebab in "Gemini 3.5 Flash"

sheeshkebab — Thu, 21 May 2026 02:53:52 +0000

It’s great laptop to mess around with llms, it won’t replace claude opus or even sonnet.

New comment by sheeshkebab in "Gemini 3.5 Flash"

sheeshkebab — Wed, 20 May 2026 00:25:38 +0000

You can only run heavily quantized models on all 3/4/5 rtx gpus (with 32gb or less vram) - and you probably want moe versions like Qwen 35b for this to run at speed somewhat comparable to Claude. It’s still not there to be honest but getting there. Personally I mess around with llama.cpp on m5 max with 128gb - it’s a decent setup to try various medium sized things, and runs llms surprisingly well without quantization, at least the moe models.

New comment by sheeshkebab in "OpenAI models coming to Amazon Bedrock: Interview with OpenAI and AWS CEOs"

sheeshkebab — Wed, 29 Apr 2026 01:25:06 +0000

The more Dario talks the less I want to have anything to do with his wares.

New comment by sheeshkebab in "DeepSeek v4"

sheeshkebab — Fri, 24 Apr 2026 12:52:37 +0000

Ask it if there was a Tiananmen square massacre. Then decide if you really want to be part of this murderous propaganda.

New comment by sheeshkebab in "SpaceX says it has agreement to acquire Cursor for $60B"

sheeshkebab — Wed, 22 Apr 2026 00:19:23 +0000

IntelliJ is a bit dated, and its plugins are too. I use IntelliJ all the time, in its various incarnations, but vscode is really up there now.

New comment by sheeshkebab in "Claude Opus 4.7"

sheeshkebab — Fri, 17 Apr 2026 03:00:09 +0000

So they nixed the fun part of working with the bot - reading its thinking output. Now this thing just plain unfun and often stupid.

So, yeah, good job anthropic. Big fuck you to you too.

New comment by sheeshkebab in "System Card: Claude Mythos Preview [pdf]"

sheeshkebab — Tue, 07 Apr 2026 23:41:44 +0000

Again, wake me up when it can do laundry.

New comment by sheeshkebab in "F-15E jet shot down over Iran"

sheeshkebab — Sat, 04 Apr 2026 04:18:10 +0000

It’s very likely their initial shock now wore off, and they got resupplied by putin and xi. We might start seeing much more damage going forward. US hasn’t fought a proxy war of this kind in many decades.

New comment by sheeshkebab in "F-15E jet shot down over Iran"

sheeshkebab — Sat, 04 Apr 2026 04:01:13 +0000

These traitors will eventually be all prosecuted. They are all traitors with putin connections, every one of them.