Hacker News: revv00

New comment by revv00 in "Using AI to write better code more slowly"

revv00 — Tue, 26 May 2026 09:14:33 +0000

i bet you've contributed a lot of training trajectories for those AI's.

New comment by revv00 in "Hacker News front page as a site"

revv00 — Mon, 25 May 2026 23:27:49 +0000

Cool, seems need some deduplication. Maybe when you turn to the new page, some items fall back.

New comment by revv00 in "Show HN: Tilde.run – Agent sandbox with a transactional, versioned filesystem"

revv00 — Thu, 07 May 2026 11:35:34 +0000

JuiceFS for posix but lack of versioning

New comment by revv00 in "The Art of Risk Management (2017)"

revv00 — Fri, 10 Apr 2026 12:36:21 +0000

Agreed. This maps directly to the white-box vs black-box testing distinction: either you own your priors and trace the full data lineage from training through validation, or you're relying on an opaque validation set of unknown provenance. And that's before factoring in the organizational politics.

New comment by revv00 in "Native Instant Space Switching on macOS"

revv00 — Thu, 09 Apr 2026 23:44:09 +0000

It doesn't work macbook pro(Intel Sonoma 14.4.1), any thoughts? app icon is a forbidden icon.

New comment by revv00 in "Native Instant Space Switching on macOS"

revv00 — Thu, 09 Apr 2026 23:34:52 +0000

Not a space user, command+tab solve most of my problem. But will give it a try.

New comment by revv00 in "Show HN: I built a local data lake for AI powered data engineering and analytics"

revv00 — Thu, 09 Apr 2026 22:58:50 +0000

Great work! What is the difference between writing a shell script to solve "cloud setup, ETL pipelines, orchestration, cost monitoring", and using a local app?

New comment by revv00 in "Improving Interactive In-Context Learning from Natural Language Feedback"

revv00 — Wed, 08 Apr 2026 05:48:31 +0000

Interesting paper, it reminds me how we human improving our math problem solving capability. But those efficient guys use few problems(and reasoning trajectories) to build their solving skills, how to build this effective human training process into LLM's?

Improving Interactive In-Context Learning from Natural Language Feedback

revv00 — Wed, 08 Apr 2026 05:48:31 +0000

Article URL: https://arxiv.org/abs/2602.16066

Comments URL: https://news.ycombinator.com/item?id=47685848

Points: 1

# Comments: 1

New comment by revv00 in "Embarrassingly simple self-distillation improves code generation"

revv00 — Tue, 07 Apr 2026 02:21:56 +0000

Indeed very insightful findings. Just wonder how it will affect visual reason tasks and other reasoning tasks.

New comment by revv00 in "A macOS bug that causes TCP networking to stop working after 49.7 days"

revv00 — Mon, 06 Apr 2026 23:13:52 +0000

Orz! A kindly reminder for rebooting.

New comment by revv00 in "NYC families need over $125k in income to live in any borough"

revv00 — Mon, 06 Apr 2026 23:11:05 +0000

Even 87k is a huge number, is it due to some selection bias?

New comment by revv00 in "My Google Workspace account suspension"

revv00 — Mon, 06 Apr 2026 13:12:37 +0000

The least evil giant might be Yahoo!, although nowadays no one uses their service due to their inability product and tech.