Top all-time · #8821
Jan Daniel Semrau
Personal blog · jdsemrau.substack.com · on HN as jsemrauaiprogramming
Jan Daniel Semrau writes Encyclopedia Autonomica, a personal blog at jdsemrau.substack.com about ai and programming. Since 2023, 9 different Hacker News readers have linked it 63 times, most often “Nemotron vs. Qwen – Game Theory and Agent Reasoning IV”. It is still writing: the last post was on 2026-10-05, and you can follow it by feed.
Too few sampled comments (10) to say how often HN recommends it.?
Why HN links it
HN readers link Jan Daniel Semrau's Substack for posts on AI agents, LLM reasoning, agentic user experience and game theory. One commenter calls a game theory and agent reasoning series fantastic.
Summarised by Claude from a sample of 10 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.
Trusted by readers at Nasdaq, Nvidia, and Stanford, Encyclopedia Autonomica is the go-to guide for anyone building or deploying autonomous agents. We cover what matters: how real-life AI agents reason, plan, remember, and act. Click to read Encyclopedia Autonomica, by Jan Daniel Semrau (MFin, CAIO),
Most-linked posts?
- /5 people · 9 mentions · first linked 2023-11-12
- Nemotron vs. Qwen – Game Theory and Agent Reasoning IV2 people · 4 mentions · first linked 2025-01-25
- Meta's Hyperagents and Self-Correcting Agents2 people · 2 mentions · first linked 2026-05-11
- How Should Agentic User Experience Work?2 people · 2 mentions · first linked 2025-10-27
- /p/pair-programming-superbill-with-codex2 people · 2 mentions · first linked 2026-03-09
- A Self-Improving Code World Model Agent2 people · 2 mentions · first linked 2026-07-06
- Discovering APIs with Knowledge Graphs2 people · 2 mentions · first linked 2025-12-01
What Hacker News says?
Seems to be https://jdsemrau.substack.com/ , also in their bio.
Comment linking /
on “Building Effective "Agents"”Submitted Cognitive Reasoning Agents and the Extended Information Filter as a story · 14 points
HN discussionYour comment reminded me of this article becasue of the Star Trek comparison. Chatting is inefficient, isn't it? [1] https://jdsemrau.substack.com/p/how-should-agentic-user-expe...
Comment linking /p/how-should-agentic-user-experience
on “GPT-5.1: A smarter, more conversational ChatGPT”Submitted Schelling Points and the Secret Society of Agents as a story · 3 points
HN discussionI wrote this a while ago about xTTSv2 mixed with Nvidia's Nemo. Maybe it kicks off your journey. https://jdsemrau.substack.com/p/teaching-your-agent-to-speak...
Comment linking /p/teaching-your-agent-to-speak?utm_source=publication-search
on “Generate audiobooks from E-books with Kokoro-82M”Submitted Recursive Language Models and Neurosymbolic Context Management as a story · 3 points
HN discussionI am playing around with LLMs and Game Theory. Currently, I let them play 5x5 board games in a tournament against Reinforcement Learning Agents and a Random Player. I am capturing the "thoughts" of the agent for behavioral data analysis. My background is Risk Management, I am trying to gather an understanding how such LLMs report their "decision" for applications in human/ai interactions for identifying and reportin…
Comment linking /
on “Ask HN: What are you working on? (October 2024)”Submitted The Hierarchical Reasoning Model as a story · 3 points
HN discussionI am working to better understand AI agents. What they are. their capabilities, and their and risks and write about it on my Substack Encyclopedia Autonomica: https://jdsemrau.substack.com On that note, I also curate a list of resources around AI Agents that fit my narrative: https://github.com/jsemrau/Agent-Repository
Comment linking /
on “Ask HN: What are you working on? (April 2025)”Submitted Anthropic's Model Context Protocol as a story · 3 points
HN discussionI think it can be quite interesting especially if you consider different character types (in Anthropic lingo this "personality"). The only problem right now is that using a proprietary LLM is incredibly expensive. Therefore having a local LLM might be the best option. Unfortunately, these are still not on the same level as their larger brethren. [1] https://jdsemrau.substack.com/p/evaluating-consciousness-and...
Comment linking /p/evaluating-consciousness-and-reasoning
on “Project Sid: Many-agent simulations toward AI civilization”Submitted Real-Time Market Data Forecasting with Transformer Models as a story · 3 points
HN discussionYou are on the right track in my opinion. The key is to encode the interface between the game and the agent so that the agent can make a straightforward choice. For example, by giving the agent the state of a nxn board as the world model, and then a finite set of choices, an agent is capable of playing the game robustly and explaining the decision to the game master. This gives the illusion that the agent reasons. I…
Comment linking /p/evaluating-consciousness-and-reasoning
on “Project Sid: Many-agent simulations toward AI civilization”Submitted Wax On, Weights Off: Knowledge Distillation Explained as a story · 2 points
HN discussionI am getting obsessed with knowing everything about autonomous AI Agents and write about my journey on my substack. https://jdsemrau.substack.com/
Comment linking /
on “Ask HN: What are you working on (September 2024)?”Submitted Meta's Hyperagents and Self-Correcting Agents as a story · 2 points
HN discussionI still haven't found what an actual working product in AI Agents could be and write about my journey into capabilities, frameworks, and restrictions here: https://jdsemrau.substack.com/ Initially I thought there is a use-case in finance, but the barriers of entry are incredibly small and the value add is not that large. Currently, there seems to be a lot of traction in code generation (Cursor, Lovable, et al), but…
Comment linking /
on “Ask HN: What are you working on? (May 2025)”Submitted Discovering APIs with Knowledge Graphs as a story · 2 points
HN discussionIt might appear so, but then you could validate it with a simple test. If the LLM would play a 4x4 Tic Tac Toe game, would the agent select the winning move 100% of all time or block a losing move 100% of the time? If these systems were capable of proper reasoning, then they would find the right choice in these obvious but constantly changing scenarios without being specifically trained for it. [1] https://jdsemrau.…
Comment linking /p/nemotron-vs-qwen-game-theory-and
on “The Case That A.I. Is Thinking”Submitted Discovering APIs with Knowledge Graphs as a story · 2 points
HN discussionThis looks like the post: https://jdsemrau.substack.com/p/finer-gliner-and-smolagents-...
Comment linking /p/finer-gliner-and-smolagents-in-finance
on “When ChatGPT broke the field of NLP: An oral history”Submitted How Should Agentic User Experience Work? as a story · 2 points
HN discussionI am still writing about Cognitive Agents on my Substack "Encyclopedia Autonomica" [1]. Last month I crossed 500 subscribers (~15 paying) on Substack and since today Google is properly indexing it. Throughout February I wrote largely about applications in FinTech like Market Making, Agents as Financial Decision Engines, Advanced Named Entity Recognition, and that I built Deep Search for Finance in 2023 and nobody ca…
Comment linking /
on “Ask HN: What are you working on? (February 2025)”Submitted Context Window Saturation in Reasoning Agents as a story · 2 points
HN discussionThe SEC has a well defined API with EDGAR. https://jdsemrau.substack.com/p/mem0-building-a-sec-10k-anal...
Comment linking /p/mem0-building-a-sec-10k-analyst-with
on “Extracting financial disclosure and police reports with OpenAI Structured Output”Submitted Cooperate or Collapse as a story · 2 points
HN discussionI have been doing some experiments with Agents, Reinforcement Learnings playing a 4x4 Tic Tac Toe game.[1]. Given my analysis of the "thought" process we are still really far from true understanding of such games. While in my game as well as OP"s, the rules are pre-trained and the models are good enough to reach a conclusion (which in itself is already impressive), it is still a long way. [1] https://jdsemrau.substa…
Comment linking /p/nemotron-vs-qwen-game-theory-and
on “OpenAI's o1 Playing Codenames”Submitted How to build a Spacecraft that flies itself as a story · 2 points
HN discussionWow. This is really beautiful? Can the code be customized to other spacecraft? I made a tracker for starlink satellites but think your visualization is much nicer. [1] https://jdsemrau.substack.com/p/code-clinic-tracking-starlin...
Comment linking /p/code-clinic-tracking-starlink-constellations
on “Show HN: Beautiful 3D ISS tracker with live video and near-realtime clouds”Submitted GTA6 – Deconstructing a System and Method for Virtual Character Locomotion as a story · 2 points
HN discussion
