Hacker News: danbrooks

New comment by danbrooks in "Ask HN: What's your current AI agentic setup?"

danbrooks — Thu, 05 Nov 2026 09:55:47 +0000

Claude App + Opus 5.5

New comment by danbrooks in "Mistral Large 4"

danbrooks — Tue, 06 Oct 2026 21:00:20 +0000

That's one heck of a pelican!

New comment by danbrooks in "Getting the most out of Opus 5.5 in Claude and Claude Code"

danbrooks — Sat, 03 Oct 2026 21:06:08 +0000

Agreed on Opus 5.5 being a great model. It's the first one that I trust for long running (>1 hour) tasks.

New comment by danbrooks in "Jeff – Jev-compatible 0.8B decision models, trained at home, ~30 ms"

danbrooks — Mon, 28 Sep 2026 21:38:48 +0000

This type of project looks extremely useful. There was a lot of buzz around Jev, but having models that run locally and can be fine-tuned is extremely helpful.

New comment by danbrooks in "LLM Ass Bench"

danbrooks — Tue, 22 Sep 2026 22:05:41 +0000

Glad my work blocked this. But can we get a NSFW tag? If this is NSFW?

New comment by danbrooks in "A Faster Shortest Path Algorithm"

danbrooks — Tue, 22 Sep 2026 21:03:13 +0000

I'm confused.

This seems to imply that they used Claude Opus 5.5 before it was released on September 22nd.

New comment by danbrooks in "Claude Opus 5.5"

danbrooks — Tue, 22 Sep 2026 16:52:05 +0000

Many people knew this announcement was coming. The betting markets suggested a very high likelihood of Opus dropping today. I was anticipating this quite a bit!

New comment by danbrooks in "Disney+: New user agreement allows ads before movies in all subscriptions"

danbrooks — Mon, 21 Sep 2026 17:02:47 +0000

Not sure this is as nefarious as it might look at first glance.

Disney+ supports a variety of content types - some of which naturally have ads.

A big example is the Super Bowl (coming in February), which has always had ads.

New comment by danbrooks in "Bonsai 2 27B: Near-Lossless Compression in a 9x Smaller Footprint"

danbrooks — Thu, 17 Sep 2026 22:42:09 +0000

Nice! Does anyone know how this compares to the Unsloth quantizations of this model? https://unsloth.ai/docs/models/qwen3.8#run-qwen3.8-guide

New comment by danbrooks in "Show HN: How Stale Is Your AI? Release age and training cutoff for 20 models"

danbrooks — Wed, 16 Sep 2026 21:15:06 +0000

An aside - isn't it nice that have been fewer model releases the past week or so? A brief respite!

New comment by danbrooks in "Ask HN: What are you working on? (September 2026)"

danbrooks — Mon, 14 Sep 2026 05:06:29 +0000

Just gave it a look. Turns out to be a paid service after a bunch of clicks.

New comment by danbrooks in "Mamdani Bans AI in NYC Schools"

danbrooks — Wed, 02 Sep 2026 22:02:00 +0000

Seems like a well designed policy. Currently for one year and up to the middle school level.

New comment by danbrooks in "OCR It – pull text out of un-copyable documents for your LLM"

danbrooks — Mon, 24 Aug 2026 21:49:21 +0000

Most LLMs can do a pretty good job of this out of the box.

If you have a coding agent available, ask it to try transcribing a few of the PDFs.

New comment by danbrooks in "I were 17, I'd learn how to build LLMs from scratch"

danbrooks — Mon, 24 Aug 2026 20:25:53 +0000

I had the same inspiration!

I spent a days reviewing the lecture notes for CS336: Language Modeling from Scratch - and then trained a nanoGPT-esque model in PyTorch.

I'd recommend trying it for those who are curious. Computational bottlenecks become much more intuitive when you've looked at the overall process.

New comment by danbrooks in "Mistral Patent for “Code implemented tool calls”"

danbrooks — Mon, 10 Aug 2026 22:40:56 +0000

Typically, companies file these patents to protect themselves against patent trolls.

New comment by danbrooks in "Claude-real-video － any LLM can watch a video"

danbrooks — Fri, 03 Jul 2026 01:00:17 +0000

I’d imagine they could. I’d try Gemini 3.5 flash with high fps.

New comment by danbrooks in "How many of the 170k English words do you know?"

danbrooks — Sat, 20 Jun 2026 03:50:43 +0000

Super high scores for the community!

I got 83/100 suggesting 60,000.

My SAT reading was 760/800.

New comment by danbrooks in "MTG Bench: Testing how well LLMs can play Magic"

danbrooks — Fri, 12 Jun 2026 00:21:50 +0000

Very cool. I’ve been daydreaming about whether LLMs can be used to reason through gaming decisions.

New comment by danbrooks in "CS336: Language Modeling from Scratch"

danbrooks — Mon, 01 Jun 2026 17:20:43 +0000

I'd be interested in joining a discord server.

Would be great to have a community to discuss the material - even if folks can't commit to the full course.

New comment by danbrooks in "Justice Department announces a $1.7B fund to compensate Trump allies"

danbrooks — Mon, 18 May 2026 19:35:18 +0000

Theoretically, the fund is non-partisan. I'd be curious to see what happens if some folks claim to be politically targeted by the current DOJ.