Hacker News: janoelze

New comment by janoelze in "Polymarket gamblers betting millions on war"

janoelze — Sat, 11 Apr 2026 14:56:32 +0000

it's an imperfect system, but UMA decisions can also be disputed again, up to two times iirc, which moves the dispute/decision to a different group of UMA whales. tbh, i think this is as good as it gets – unless you stay out of this completely.

New comment by janoelze in "Thought-Terminating Cliché"

janoelze — Mon, 09 Feb 2026 13:54:39 +0000

Don't overthink it.

New comment by janoelze in "Claude Chill: Fix Claude Code's flickering in terminal"

janoelze — Wed, 21 Jan 2026 01:17:34 +0000

amazing, thanks

New comment by janoelze in "Scaling long-running autonomous coding"

janoelze — Tue, 20 Jan 2026 03:18:26 +0000

Any views on the nature of "maintainability" shifting now? If a fleet of agents demonstrated the ability to bootstrap a project like that, would that be enough indication to you that orchestration would be able to carry the code base forward? I've seen fully llm'd codebases hit a certain critical weight where agents struggled to maintain coherent feature development, keeping patterns aligned, as well as spiralling into quick fixes.

New comment by janoelze in "A prediction market user made $436k betting on Maduro's downfall"

janoelze — Tue, 06 Jan 2026 04:43:36 +0000

"Correct" is hard to define with prediction markets as likelihood often increases/decreases in the hours before a market is actually called or confirmed. i can't speak for the methodology here, but i remember this data report from a user: https://dune.com/alexmccullough/polymarket-brier-score

New comment by janoelze in "Court report detailing ChatGPT's involvement with a recent murder suicide [pdf]"

janoelze — Wed, 31 Dec 2025 19:40:34 +0000

Thinking more about this, a right to inspect the full context window ensured by consumer protection laws wouldn't be a bad thing.

If there's one place to implement a PsyOp, context is it. Users should be allowed to see what influenced the message they're reading on top of the training data.

New comment by janoelze in "Court report detailing ChatGPT's involvement with a recent murder suicide [pdf]"

janoelze — Wed, 31 Dec 2025 19:25:02 +0000

I'm generally positive on LLMs, but became convinced that long term memory features that LLM chat providers implemented are just too hard to keep on track.

They create a "story drift" that is hard for users to escape. Many users don't – and shouldn't have to – understand the nature and common issues of context. I think in the case of the original story here the LLM was pretty much in full RPG mode.

I've turned off conversation memory months ago, in most cases i appreciate knowing i'm working with a fresh context window; i want to know what the model thinks, not what it guesses i'd like to hear. I think conversations with memory enabled should have a clear warning message on top.

New comment by janoelze in "Show HN: Vibe Kanban – Kanban board to manage your AI coding agents"

janoelze — Fri, 11 Jul 2025 21:17:15 +0000

it came to mind first, you're free to use whatever flavour of LLM f̶l̶o̶a̶t̶s̶ ̶y̶o̶u̶r̶ ̶b̶o̶a̶t̶ vibes your code.

New comment by janoelze in "Show HN: Vibe Kanban – Kanban board to manage your AI coding agents"

janoelze — Fri, 11 Jul 2025 21:14:16 +0000

yes, i was just doing/thinking the same, it was an interesting experience to sculpt a somewhat complex codebase to my needs in minutes.

New comment by janoelze in "Show HN: Vibe Kanban – Kanban board to manage your AI coding agents"

janoelze — Fri, 11 Jul 2025 21:06:41 +0000

fork, task claude to remove all github dependence, build.

New comment by janoelze in "Odyc.js – A tiny JavaScript library for narrative games"

janoelze — Sat, 07 Jun 2025 00:10:10 +0000

its a thoughtful idea – and been there too, just recently did a faithful html recreation of the 2005 MSN chat interface to surprise someone with teenage nostalgia. it was sweet.

New comment by janoelze in "Odyc.js – A tiny JavaScript library for narrative games"

janoelze — Sat, 07 Jun 2025 00:06:06 +0000

I recently looked into that for a project and was surprised that it's a built-in feature! :) https://medium.com/@amatewasu/how-to-record-a-canvas-element...

New comment by janoelze in "Writing C for Curl"

janoelze — Mon, 07 Apr 2025 09:27:52 +0000

This is remarkably clear writing — you sense how it was formed by thousands upon thousands of hours spent communicating, really cool.

New comment by janoelze in "From Prompt to Adventures:Creating Games with LLMs and Restate Durable Functions"

janoelze — Mon, 10 Mar 2025 12:59:29 +0000

I've experimented with a very similar flow back in the days of GPT3.5. It resulted in really fun story lines, but i found that the player choices benefitted from some steering: For each plot point i generated three choices, where [1] progressed the story, [2] stalled the plot and [3] resulted in a plot twist, just these additions to the choice prompts increased playability by a lot, and prevented plot stagnation and "feedback loops".

New comment by janoelze in "Thinkserver: My web-based coding environment"

janoelze — Thu, 13 Feb 2025 09:34:21 +0000

nice work! very similar to the scrappy codepen clone i built for myself, will copy a few things! I've integrated AI edits and use it suprisingly often as well to sketch out mini apps. https://endtime-instruments.org/scratch/

New comment by janoelze in "Show HN: Take the one minute Table Setting Efficiency Test (T.S.E.T.)"

janoelze — Mon, 13 Jan 2025 16:19:01 +0000

Show HN: Take the one minute Table Setting Efficiency Test (T.S.E.T.)

janoelze — Mon, 13 Jan 2025 16:16:52 +0000

The Bureau of Hospitality & Food Service is kindly requesting all Hacker News users to take the Table Setting Efficiency Test (T.S.E.T.) to ensure consideration in the Post-AI Resource Restructuring Program.

Comments URL: https://news.ycombinator.com/item?id=42685069

Points: 1

# Comments: 1

New comment by janoelze in "Show HN: News – A merged news feed of Hacker News and Reddit submissions"

janoelze — Sat, 04 Jan 2025 17:35:40 +0000

think this should help, but will look into smarter content clustering in the next days. https://endtime-instruments.org/news/?preset=science

New comment by janoelze in "Show HN: News – A merged news feed of Hacker News and Reddit submissions"

janoelze — Sat, 04 Jan 2025 15:45:58 +0000

not yet, but i'm just adding a few features here and there as we speak. will keep that on the list!

New comment by janoelze in "Show HN: News – A merged news feed of Hacker News and Reddit submissions"

janoelze — Sat, 04 Jan 2025 15:07:47 +0000

thanks, so are you!