Top all-time · #20743
Matthew Newton
Personal blog · matthewnewton.com/blog · on HN as mattnewtonaiprogramming
Matthew Newton's personal blog at matthewnewton.com covers ai and programming. Since 2025, 3 different Hacker News readers have linked it 6 times, most often “Hacking Diffusion into Qwen3 for the Arc Challenge”. No feed was found, so read it on the site itself.
Too few sampled comments (3) to say how often HN recommends it.?
Why HN links it
The HN comments that link this blog are Krea job listings from 'Who is hiring' threads. They don't discuss the writing, so the topics of the linked posts can't be determined from them.
Summarised by Claude from a sample of 3 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.
Thoughts on machine learning, startup engineering, and building AI systems at scale.
Most-linked posts?
- Hacking Diffusion into Qwen3 for the Arc Challenge3 people · 4 mentions · first linked 2025-08-05
What Hacker News says?
Krea | Fulltime | San Francisco, CA, USA | Onsite Krea ( https://www.krea.ai/ ) builds creative tooling and large-scale infrastructure for multimedia teams. Users range from hobbyists to world-renowned creatives: architects behind the World Trade Center or creative teams at Apple. We treat research, engineering, and creativity as first-class citizens. We're looking for engineers to join our research, product, and GT…
Comment linking /blog/arc-challenge-diffusion
on “Ask HN: Who is hiring? (September 2026)”Submitted Hacking Diffusion into Qwen3 for the Arc Challenge as a story · 126 points
HN discussionI recently wrote a short anecdote in a similar vein- in my testing, “agentic” retrieval where you simply pass an annotated list of files to an LLM and ask it which ones it wants to look at is probably better than traditional RAG for small datasets (few hundreds of docs). I found it was both much simpler and more accurate at the cost of marginally more time and tokens, compared to RAG on embedded chunks with a vector…
Comment linking /blog/replacing-rag
on “Show HN: I replaced vector databases with Git for AI memory (PoC)”Submitted You Might Not Need RAG as a story · 3 points
HN discussionKrea; Fulltime; Onsite, San Francisco, CA, USA. I'm the cofounder of Krea ( https://www.krea.ai/ ), a company building creative tooling and large-scale infrastructure for multimedia teams. Users range from hobbyists to world-renowned creatives: architects behind the World Trade Center or creative teams at Apple. We treat research, engineering, and creativity as first-class citizens. We're looking for engineers to jo…
Comment linking /blog/arc-challenge-diffusion
on “Ask HN: Who is hiring? (April 2026)”Krea; Fulltime; Onsite, San Francisco, CA, USA. I'm the cofounder of Krea ( https://www.krea.ai/ ). We build creative tools and large-scale infrastructure for multimedia teams. Users range from hobbyists to world-renowned creatives: architects behind the World Trade Center or creative teams at Apple. We treat research, engineering, and creativity as first-class citizens; we want engineers for our research, product,…
Comment linking /blog/arc-challenge-diffusion
on “Ask HN: Who is hiring? (May 2026)”