Top all-time · #11409
Cobus Greyling
Personal blog · cobusgreyling.medium.comaiprogramming
Cobus Greyling's personal blog at cobusgreyling.medium.com covers ai and programming. Since 2022, 7 different Hacker News readers have linked it 12 times. It is still writing: the last post was on 2026-10-30, and you can follow it by feed.
Too few sampled comments (4) to say how often HN recommends it.?
Why HN links it
HN readers link Cobus Greyling's Medium posts on LLM architecture, how chat models are trained on conversation formats, and no-code LLM tools such as Flowise and Langflow. Comments cite them without judging the writing.
Summarised by Claude from a sample of 4 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.
What Hacker News says?
Which part is the query preprocessor? /? LLM stack app architecture https://www.google.com/search?q=LLM+stack+app+architecture&t... https://cobusgreyling.medium.com/emerging-large-language-mod... ; Flexibility / Complexity,
Comment linking /emerging-large-language-model-llm-application-architecture-cba0e7862037
on “Can Large Language Models Understand Symbolic Graphics Programs?”Submitted Fine-Tuning OpenAI's Small Language Model as a story · 3 points
HN discussionOkay, here's the tl;dr: Attention based neural network architectures (on which the majority of LLMs are built) has a unit economic cost that scales (roughly) n^2 i.e. quadratic (for both memory and compute). In other words, the longer the context window, the more expensive it is for the upstream provider. That's one cost. The second cost is that you have to resend the entire context every time you send a new message…
Comment linking /llm-context-rot-28a6d0399655
on “Opus 4.5 is going to change everything”Submitted Four Levels of RAG – Research from Microsoft as a story · 1 points
HN discussion"Chat" models have been heavily fine-tuned with a training dataset that exclusively uses a formal turn-taking conversation syntax / document structure. For example, ChatGPT was trained with documents using OpenAI's own ChatML syntax+structure ( https://cobusgreyling.medium.com/the-introduction-of-chat-ma... ). This means that these models are very good at consistently understanding that they're having a conversation…
Comment linking /the-introduction-of-chat-markup-language-chatml-is-important-for-a-number-of-reasons-5061f6fe2a85
on “Right-sizes LLM models to your system's RAM, CPU, and GPU”Submitted ChatGPT Custom Instructions as a story · 1 points
HN discussionPerhaps a no-code UI might let you wire up the conversational memory being suggested here. I haven’t used these but saw a post on them: https://cobusgreyling.medium.com/flowise-for-langchain-b7c40... https://cobusgreyling.medium.com/langflow-46563a8af323
Comment linking /langflow-46563a8af323
on “Simulating History with ChatGPT”Submitted Vercel Launched a Comprehensive LLM Playground as a story · 1 points
HN discussionRead more here on ChainForge and the updates they have made. https://cobusgreyling.medium.com/significant-updates-where-m...
Comment linking /significant-updates-where-made-to-chainforge-96927a20b496
on “ChainForge: A Visual Toolkit for Prompt Engineering and LLM Hypothesis Testing”Submitted A Hands-On Analysis of the LLM Tooling Landscape (Part 1) as a story · 1 points
HN discussionSubmitted Conversational AI, Verticals, and Pinecone as a story · 1 points
HN discussion
