Hacker News: xiaoyu2006

New comment by xiaoyu2006 in "Detecting LLM-Generated Texts with “Classical” Machine Learning"

xiaoyu2006 — Thu, 16 Jul 2026 17:10:19 +0000

I assume different models will have different distribution, so it has to be kept updated?

New comment by xiaoyu2006 in "OnePlus halts operations in USA and Europe"

xiaoyu2006 — Thu, 16 Jul 2026 12:58:05 +0000

Not quite worth its high price

New comment by xiaoyu2006 in "OnePlus halts operations in USA and Europe"

xiaoyu2006 — Thu, 16 Jul 2026 11:30:43 +0000

They don't sell well in Asia. It's mainly xiaomi, oppo, vivo and huawei.

New comment by xiaoyu2006 in "Rewriting Bun in Rust"

xiaoyu2006 — Thu, 09 Jul 2026 10:44:09 +0000

Ownership is a very limiting coding constraint, and bypassing methods, when ultimately needed (e.g. cell), cannot be engineered to be ergonomical.

New comment by xiaoyu2006 in "Grok 4.5"

xiaoyu2006 — Thu, 09 Jul 2026 10:26:59 +0000

Not playing the politics game is also a political stances. There is no way to avoid it.

New comment by xiaoyu2006 in "Grok 4.5"

xiaoyu2006 — Thu, 09 Jul 2026 10:26:01 +0000

Corporation machine are not individuals (geeks). They chase solely profit while we have (at least some) agency on supporting the moral sides.

New comment by xiaoyu2006 in "Grok 4.5"

xiaoyu2006 — Thu, 09 Jul 2026 10:19:51 +0000

(Just fyi the correct answer is that organizer must report the organizing beforehand or it would be illegal.) Model labs have to censor their models in order to publish them, which is not equlv to model lab management members actively showing their political stances.

New comment by xiaoyu2006 in "Show HN: A statically typed, cross-platform, easily bootstrappable build system"

xiaoyu2006 — Sat, 04 Jul 2026 03:06:55 +0000

I always wonder why use a DSL while you can just borrow some existing PL?

Why are there more top grades at university? ChatGPT is to blame

xiaoyu2006 — Mon, 29 Jun 2026 01:17:25 +0000

Article URL: https://english.elpais.com/technology/2026-05-21/why-are-there-more-top-grades-at-university-chatgpt-is-to-blame.html

Comments URL: https://news.ycombinator.com/item?id=48713636

Points: 4

# Comments: 1

New comment by xiaoyu2006 in "Bashblog – a single bash script to create blogs"

xiaoyu2006 — Sun, 28 Jun 2026 09:12:11 +0000

...for fun!

New comment by xiaoyu2006 in "Fusion Programming Language"

xiaoyu2006 — Sat, 27 Jun 2026 09:29:40 +0000

Bad idea. It either takes a useless subset of all languages or do weird abstraction that wouldn't fit in a language otherwise.

Position: Stop Anthropomorphizing Intermediate Tokens as Reasoning/Thinking Trac

xiaoyu2006 — Fri, 26 Jun 2026 06:46:56 +0000

Article URL: https://arxiv.org/abs/2504.09762

Comments URL: https://news.ycombinator.com/item?id=48683190

Points: 4

# Comments: 0

New comment by xiaoyu2006 in "A website that lists websites to submit your website to"

xiaoyu2006 — Thu, 18 Jun 2026 16:47:30 +0000

...recursion?!

New comment by xiaoyu2006 in "I hate compilers"

xiaoyu2006 — Thu, 18 Jun 2026 05:52:11 +0000

It should not. Abstraction in software engineering brings intelligence. (compression correlates to intelligence)

New comment by xiaoyu2006 in "Epic Games announces Lore version control system"

xiaoyu2006 — Wed, 17 Jun 2026 18:08:58 +0000

this is such a good laugh. peak comedy

Math Education, and LLM

xiaoyu2006 — Wed, 17 Jun 2026 11:40:24 +0000

Article URL: https://ycao.net/posts/math-education-llm

Comments URL: https://news.ycombinator.com/item?id=48568928

Points: 1

# Comments: 1

New comment by xiaoyu2006 in "GLM-5.2 is the new leading open weights model on Artificial Analysis"

xiaoyu2006 — Wed, 17 Jun 2026 11:06:13 +0000

Some models are heavily subsidized. Total params & active params are better measurement of inference cost.

New comment by xiaoyu2006 in "GLM-5.2 is the new leading open weights model on Artificial Analysis"

xiaoyu2006 — Wed, 17 Jun 2026 11:04:39 +0000

This open source model is quite near SOTA with only 700B/40B MoE. Truly efficient.

Ask HN: How do you deal with the feeling of "loss of control" with AI coding?

xiaoyu2006 — Mon, 15 Jun 2026 14:58:00 +0000

I have to admit that newest models and harness tools have gotten really good and produce working code really fast with even mediocre prompts "wish programming".

The problem is more on my side than the LLM side. I feel I am 1) losing control of my project, even my hobby ones, because LLM outputs code so fast that I am 2) too lazy to manually review all of them, even if, indeed, I had the overall architecture design choices and sometimes straight up pseudo-code in natural language. But NL is, by definition, not formal, and programming is the process of expanding informal demands into formal languages. It is still fundamentally different from abstraction (like importing a lib) as in software engineering as a means to reduce complexity. The result is that not-going-into-nitty-gritty-impl-details builds the feeling of loss-of-control and -confidence.

This may not read as a coherent submission since it's just my random unsettling thoughts, but I am sure at least some of you fellow HNers feel (felt?) the same. What is your take on this?

Comments URL: https://news.ycombinator.com/item?id=48542239

Points: 3

# Comments: 3

New comment by xiaoyu2006 in "Ask HN: What does your local LLM setup looks like?"

xiaoyu2006 — Mon, 15 Jun 2026 06:57:10 +0000

V100 32G SXM2 adapted to PCIE. Running llamacpp with Q4KM Qwen 3.6 27B or Gemma 4 31B. I use them when I feel of privacy is important or I just want to mess around.