Hacker News Reader

arthurddon Ask HN: What are you working on? (August 2026)

I'm working on an interactive wildlife & plant identification quiz powered by iNaturalist: https://arthurdick.github.io/local-biodiversity-trainer/

arthurddon The Cost of Winning:How RL Training on Poker Leads to Evil LLMs

I don't think poker is good for people either. The goal of poker is to win, taking money from other players by exploiting their weaknesses. It's a training ground to become ruthless, unempathetic, and deceptive.

arthurddon RBT unifies math, physics, and computation from absolute nothing

Humbled to see the new official timeline of physics: Newton, Maxwell, Einstein, and now... Grok in 2025 claiming a single bit-flip explains life, the universe, and everything. What a time to be alive.

arthurddon Do Communication Skills Matter?

Software development is rarely a solo activity. You will need to communicate with other developers, product managers, designers, and non-technical stakeholders.

arthurddon Ask HN: Do you still bookmark websites?

I still use traditional browser-based bookmarking, and sync with my Firefox account. I don't see the need to share my bookmarks.

arthurddon Why LLMs Can Think (as per new fundamental theory)

An LLM will generally go along with whatever you prompt it with. Try asking it to be skeptical: https://g.co/gemini/share/78b94bea5977

arthurddon Why LLMs Can Think (as per new fundamental theory)

The "Participatory Interface Theory" is not a scientific theory. It is a work of science fiction masquerading as physics. It uses the following techniques to create an illusion of credibility:

Poetic Language: Replaces rigorous explanation with evocative metaphors.

Jargon Salad: Mixes terms from physics, computer science, and philosophy to sound profound.

Reverse-Engineering: Builds a complex model designed to reproduce a known anomaly (MOND) and presents this as a success.

The "One Falsifiable Prediction" Gambit: Creates a single, specific (but logically weak) prediction to gain the appearance of being scientific, while ignoring a mountain of contradictory evidence.

While the vision it presents is beautiful and thought-provoking, it has no demonstrable connection to reality.

arthurddon Ask HN: What ad-hoc software do you use regularly?

I built my own recurring task manager, Perennial Task, to overcome the limitations I faced with TaskWarrior. Both are open source, so it's not something I was paying for.

arthurddon Remove YouTube links from Gemini responses

I think you could more simply disable the YouTube app in the Gemini settings.

arthurddon Ask HN: What are you working on? (July 2025)

Recently I've been working on a project I call term-to-svg. It's a command-line tool that converts terminal session recordings into animated SVG images.

https://github.com/arthurdick/term-to-svg

arthurddon Play Chess Against GPT-4 First GPT-4 Based (Working and Legal) Chess Web-Bot

Hi Matt, I can't find any contact info in your profile. You may want to review the OpenAI best practices for API key safety: https://help.openai.com/en/articles/5112595-best-practices-f...

arthurddon Play Chess Against GPT-4 First GPT-4 Based (Working and Legal) Chess Web-Bot

Good work! It took me 21 moves.

1. d4 d5 2. c4 Nf6 3. Nc3 e6 4. cxd5 exd5 5. e4 Nxe4 6. Bd3 Nxc3 7. bxc3 h6 8. Nf3 Be7 9. O-O O-O 10. Bb2 Nc6 11. Qb3 Na5 12. Qa4 Bd7 13. Qxa5 c5 14. Qxd8 Raxd8 15. Rfe1 c4 16. Rxe7 b6 17. Ne5 Be6 18. Bc2 Rfe8 19. Rxa7 Ra8 20. Rxa8 f6 21. Rxe8#

arthurddon Show HN: The Paradoxical Quest – A Branching Story Generated by GPT4

PlotBranch is a side project of mine generating branching stories using GPT4. You may run into a generation screen if you request a page nobody has viewed before.

"The Paradoxical Quest" is the most popular story, but there are more to choose from on the homepage if you are interested.

arthurddon Ask HN: Does your personal page get any interaction?

My site is mainly for my own entertainment, and I've only been contacted a handful of times with nearly 20 years running it.

It is probably helpful to present a competent image when job searching. It shows you can develop and run a site, and that you have interests of your own.

arthurddon Ask HN: How do you tell if online content is AI generated?

Since experimenting with the GPT models I do notice when posts and articles seem to have used generated content, but I can't identify what it is that gives me this feeling. Maybe it is just a feeling.

When using the models, I find the content generated frequently must be edited to be useful. Presumably if it doesn't wander off or repeat itself too much, there is a human in the loop guiding the generation. Take it as seriously as you would any anonymous post on the Internet.

Fact check important stuff by seeking reputable sources. Treat the rest as entertainment. Even if it is possible to detect, just because text is generated does not mean it's incorrect.

I imagine given the way text is generated using probabilities, it can in theory be detected when stretches of text are using high probability words in sequence. I'm not aware of any tools that do this at the moment though.

arthurddon Google Timer is gone

I've hacked together my own replacement, as a mini project on my blog:

https://arthurdick.com/blog/timer/

arthurddon Show HN: Draw on Paper

This is a vector drawing web app I've been working on in my spare time. It is open source and hosted on Github at https://github.com/arthurdick/drawonpaper