Hacker News: yaodub

Show HN: Checkpoint! Airport security sim game built with Claude Fable

yaodub — Thu, 11 Jun 2026 17:24:31 +0000

Fable just came out and I wanted to see what it was actually capable of.

The game: an airport security sim where manage passengers and keep them from getting angry. You place rope barriers, open lanes, build out a concourse mall. Pixel people route through whatever you designed. Standing still feels much longer than walking, so a long snaking line keeps people calm and a formless crowd gets angry and leaves. You're designing how the wait feels, not how long it takes.

Whole thing came together in one context window. I'm impressed.

Comments URL: https://news.ycombinator.com/item?id=48493384

Points: 2

# Comments: 0

New comment by yaodub in "Workers are spending over 6 hours a week botsitting AI, fueling job frustration"

yaodub — Thu, 11 Jun 2026 14:22:18 +0000

Your coworkers haven't changed. What changed is that people can hand off work they never had to think through themselves. So you don't know what they checked and you don't know what you need to. You just have to read the whole thing.

New comment by yaodub in "Ask HN: Did Anthropic Just Win?"

yaodub — Wed, 10 Jun 2026 18:31:17 +0000

Sure, but you're paying the higher rate just to find out if that's true for your use case.

New comment by yaodub in "Ask HN: Did Anthropic Just Win?"

yaodub — Wed, 10 Jun 2026 18:28:28 +0000

On capability, yeah. On pricing, they just told most developers they're not the target anymore.

New comment by yaodub in "Claude Fable 5"

yaodub — Tue, 09 Jun 2026 17:38:04 +0000

SWE-Bench measures single tasks in isolation. In a real loop the model usually loses track of what I was trying to do long before code quality becomes the issue.

New comment by yaodub in "Claude Fable 5"

yaodub — Tue, 09 Jun 2026 17:33:09 +0000

Depends whether "unable to fully automate" means "needs occasional human checkpoints" or "slowly stops caring about your actual goal." Pretty different.

New comment by yaodub in "Ask HN: What are tools you have made for yourself since the advent of AI?"

yaodub — Mon, 08 Jun 2026 20:22:48 +0000

Built a quant system that reads earnings transcripts for what management is trying not to say. The model is surprisingly bad at this. Turns out management is too.

New comment by yaodub in "Running Python code in a sandbox with MicroPython and WASM"

yaodub — Sun, 07 Jun 2026 03:15:58 +0000

Interesting middle ground between full WASM lockdown and a bare environment. Did you end up needing to block anything else beyond the package manager?

New comment by yaodub in "My Agent Skill for Test-Driven Development"

yaodub — Sat, 06 Jun 2026 20:07:12 +0000

Same with legal questions, tbh. Spots the same issues, completely disagree on which ones matter. Maybe you just need a third model to choose between the outputs lol

New comment by yaodub in "My Agent Skill for Test-Driven Development"

yaodub — Sat, 06 Jun 2026 02:26:49 +0000

Have you found integrating outputs from different frontier labs consistently improves final results, or is it just kind of voodoo?