Top all-time · #15149
Weco AI
Company blog · weco.ai/blogai
Weco AI is the company blog at weco.ai, writing about ai. Since 2024, 5 different Hacker News readers have linked it 6 times, most often “AIDE²: The First Evidence of Recursive Self-Improvement”. It is still writing: the last post was on 2026-07-14, and you can follow it by feed.
Too few sampled comments (1) to say how often HN recommends it.?
Why HN links it
HN readers link Weco AI's blog for work on reliable LLM use, reward hacking and GPU kernel evaluation. One commenter points readers to a post that covers a validator passing a result that is useless.
Summarised by Claude from a sample of 1 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.
Weco is a research and product lab building recursively self-improving AI.
Most-linked posts?
- AIDE²: The First Evidence of Recursive Self-Improvement3 people · 3 mentions · first linked 2026-07-14
What Hacker News says?
> but is slower than the baseline. Validator says pass, result is useless You should read this blog, they cover this exact scenario - https://www.weco.ai/blog/first-evidence-of-recursive-self-im... > One domain that suffers from this particularly is GPU kernel engineering. We adopt our previous idea for detecting reward hacking from SpecBench and apply that to a set of KernelBench tasks, measuring whether the speedu…
Comment linking /blog/first-evidence-of-recursive-self-improvement
on “DSLs Enable Reliable Use of LLMs”Submitted AIDE²: First Evidence of Recursive Self-Improvement as a story · 3 points
HN discussionSubmitted AIDE²: The First Evidence of Recursive Self-Improvement as a story · 3 points
HN discussionSubmitted Show HN: Is autoresearch better than classic hyperparameter tuning? as a story · 2 points
HN discussionSubmitted AI agent achieves human-level performance in Kaggle competitions as a story · 2 points
HN discussionSubmitted Show HN: Weco Aide, an AutoML Agent as a story · 2 points
HN discussion