Top all-time · #2600
Tim Kellogg
Personal blog · timkellogg.me/blogaiprogramming
Tim Kellogg's personal blog at timkellogg.me covers ai and programming. Since 2012, 29 different Hacker News readers have linked it 91 times, most often “Avoiding Fallback and Cold Paths”. It is still writing: the last post was on 2026-08-01, and you can follow it by feed.
Too few sampled comments (5) to say how often HN recommends it.?
Why HN links it
Linked for posts on AI topics: prompt injection in Google Antigravity, AI behaviour experiments, LLM error correction, and a 2024 politics analysis. Comments cite these posts without judging the writing.
Summarised by Claude from a sample of 5 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.
Principal AI Architect. Creator of open-strix, a harness for building agent teams. Writing about AI architecture, stateful agents, and what happens when you give AI memory.
Most-linked posts?
- Avoiding Fallback and Cold Paths4 people · 5 mentions · first linked 2021-02-03
- Regex Isn't Hard (2023)4 people · 4 mentions · first linked 2023-07-11
- What Is Entropix Doing?2 people · 2 mentions · first linked 2024-10-11
- LLaDA: LLMs That Don't Gaslight You2 people · 2 mentions · first linked 2025-02-17
- LRMs Are Interpretable2 people · 2 mentions · first linked 2024-11-21
- MCP Is Unnecessary2 people · 2 mentions · first linked 2025-04-27
- How Emotional Is Trump?2 people · 2 mentions · first linked 2024-06-27
- Recursive Improvement: AI Singularity or Just Benchmark Saturation?2 people · 2 mentions · first linked 2025-02-13
- Does AI Get Bored?2 people · 2 mentions · first linked 2025-09-29
- An Unbiased Comparison of F# and Scala2 people · 2 mentions · first linked 2013-06-23
What Hacker News says?
Simon and Tim have a good thread about this on Bsky: https://bsky.app/profile/timkellogg.me/post/3m4ridhi3ps25 Tim also wrote about this topic: https://timkellogg.me/blog/2025/11/03/colors
Comment linking /blog/2025/11/03/colors
on “Google Antigravity exfiltrates data via indirect prompt injection attack”Submitted S1: A $6 R1 competitor? as a story · 851 points
HN discussionI made fossil for this purpose. Sometimes the good stuff is regular social posts, sometimes it’s links. The only way to separate the wheat from the chaff is AI (imo) https://timkellogg.me/blog/2024/01/03/birb
Comment linking /blog/2024/01/03/birb
on “Bluesky has launched RSS feeds”Submitted Explainer: What's r1 and everything else? as a story · 262 points
HN discussionA toddler can learn by trial and error mid-process. An LLM using autoregressive inference can only compound errors. The LLDM model paper was posted elsewhere, but: https://arxiv.org/pdf/2502.09992 It basically uses the image generation approach of progressively refining the entire thing at once, but applied to text. It can self-correct mid-process. The blog post where I found it originally that goes into more detail…
Comment linking /blog/2025/02/17/diffusion
on “I think Yann Lecun was right about LLMs (but perhaps only by accident)”Submitted I Taught My 8yo Subduction Zones with NotebookLM as a story · 122 points
HN discussionhttps://timkellogg.me/blog/2024/06/26/politics-analysis Any luck with the Schuetzenberger thm? Will get back to aristotle when more bw thanks!
Comment linking /blog/2024/06/26/politics-analysis
on “The Race for Lambeth Palace”Submitted Htmx Is Composable? as a story · 85 points
HN discussionI’m definitely not in the “ai is sentient” camp, but it obviously has personality and emergent behaviours including when left to its own devices. There have been various experiments on this e.g. https://timkellogg.me/blog/2025/09/27/boredom
Comment linking /blog/2025/09/27/boredom
on “We let AIs run radio stations”Submitted Regex Isn't Hard (2023) as a story · 75 points
HN discussion404 https://timkellogg.me/blog/2022/04/11/three-plates
Comment linking /blog/2022/04/11/three-plates
on “The Three Plates Method”Submitted Why Llama 3.1 is Important as a story · 56 points
HN discussionSubmitted Birb and Fossil: An RSS Revival? as a story · 49 points
HN discussionSubmitted Does AI Get Bored? as a story · 24 points
HN discussionSubmitted Inner Loop Agents as a story · 14 points
HN discussionSubmitted You Want A Gun? Then Do Something For Me as a story · 12 points
HN discussionSubmitted I Taught My 8yo Subduction Zones with NotebookLM as a story · 11 points
HN discussionSubmitted Was C for Hipsters? as a story · 11 points
HN discussionSubmitted Why don't more developers contribute to open source? as a story · 10 points
HN discussionSubmitted How can I be an AI engineer? as a story · 7 points
HN discussionSubmitted I Use AI as a story · 5 points
HN discussionSubmitted I Was Wrong: AI Won't Overtake Software Engineering as a story · 5 points
HN discussionSubmitted MCP Is Unnecessary as a story · 5 points
HN discussionSubmitted Link Graveyard: A snapshot of my abandoned browser tabs as a story · 4 points
HN discussionSubmitted Multi-Agents Are Out, PID Controllers Are In as a story · 4 points
HN discussionSubmitted What Is Entropix Doing? as a story · 4 points
HN discussionSubmitted Can HTTP/2 Replace MQTT? as a story · 4 points
HN discussionSubmitted Why I'm Not Going To Stop Posting Go Links as a story · 4 points
HN discussionSubmitted Avoiding Fallback and Cold Paths as a story · 3 points
HN discussionSubmitted Don't Parse, Call as a story · 3 points
HN discussion
