Hacker News: winwang

New comment by winwang in "Rust Threads on the GPU"

winwang — Wed, 15 Apr 2026 22:45:02 +0000

Each SM should have 4 independent SMSPs (32 lanes each), no? Effectively a "4-core" task-parallel system per SM.

New comment by winwang in "McGridsort: Warping Grids for GPU k-way mergesort"

winwang — Tue, 07 Apr 2026 17:55:15 +0000

Had a fun little idea for a weird GPU/SIMD k-way mergesort a couple years back, finally decided to write it up! (Anti-)jumpscare: no hard perf numbers in the post (though I have profiled it somewhat already).

McGridsort: Warping Grids for GPU k-way mergesort

winwang — Tue, 07 Apr 2026 17:55:15 +0000

Article URL: https://winwang.blog/posts/mcgridsort

Comments URL: https://news.ycombinator.com/item?id=47678925

Points: 3

# Comments: 1

New comment by winwang in "ARC-AGI-3"

winwang — Thu, 26 Mar 2026 23:34:19 +0000

Interestingly, I find that the models generalize decently well as long as the "training" (more analogous to that for humans) fits in (small enough) context. That's to say, "in-context learning" seems good enough for real use.

But of course, that's not quite "long term"

New comment by winwang in "ARC-AGI-3"

winwang — Thu, 26 Mar 2026 19:17:45 +0000

How much of this is expectations setting by the heights models reach? i.e. of we could assess a consistent floor of model performance in a vacuum, would we say it's better at "AGI" than the bottom 0.1% of humans?

New comment by winwang in "iPhone 17 Pro Demonstrated Running a 400B LLM"

winwang — Mon, 23 Mar 2026 17:09:11 +0000

It would be much worse if it had said "You are absolutely wrong to be confused", haha.

New comment by winwang in "Non-Messing-Up++: Diagonal Sorting and Young Tableaux"

winwang — Mon, 23 Mar 2026 12:45:57 +0000

Hey HN, I figured to just share this for feedback despite its dry-ness and small-idea-ness.

Non-Messing-Up++: Diagonal Sorting and Young Tableaux

winwang — Mon, 23 Mar 2026 12:45:57 +0000

Article URL: https://winwang.blog/posts/non-messing-up++

Comments URL: https://news.ycombinator.com/item?id=47488726

Points: 14

# Comments: 1

New comment by winwang in "Dataframe 1.0.0.0"

winwang — Mon, 23 Mar 2026 12:33:42 +0000

(no idea but) I feel like changing the first number has a psychological issue, but the 2nd number feels more important than just "minor" sometimes. So may as well let the schema set the mind free?

New comment by winwang in "Our commitment to Windows quality"

winwang — Fri, 20 Mar 2026 20:04:13 +0000

...I almost thought it was a parody site!

New comment by winwang in "Ask HN: What is it like being in a CS major program these days?"

winwang — Mon, 16 Mar 2026 12:43:16 +0000

Interesting. I've felt like it's never been easier to learn things, but I suppose that's not quite the same as "acquiring new skills". I don't know if it applies, but it's always been easy to take the easy way out?

I feel like AI has made it a bit easier to do harder things too.

New comment by winwang in "LLM Writing Tropes.md"

winwang — Sun, 08 Mar 2026 23:40:15 +0000

Yeah that's somewhat close to what I meant, though there's an irony here in that your comment (and this one) are pretty reddit-esque.

New comment by winwang in "LLM Writing Tropes.md"

winwang — Sun, 08 Mar 2026 21:47:26 +0000

I don't think lived experience matters too much to me. In some sense, AI has very unique "lived" experience, which is what creates the voice it uses ("doesn't have a voice" seems like an impossibility to me by definition).

I find AI very "human-esque", and its "self-reported" phenomenology is very entertaining to me, at least.

I also think AI writing might feel trashy also because most human writing is trashy.

New comment by winwang in "Google Workspace CLI"

winwang — Thu, 05 Mar 2026 02:54:06 +0000

Really interesting. I was thinking about something similar regarding the shape of code. I have no qualms recommending my agents take static analysis to the extreme, though it would cumbersome for most people.

New comment by winwang in "Nobody gets promoted for simplicity"

winwang — Wed, 04 Mar 2026 14:13:39 +0000

What about someone inexperienced but skeptical, using AI to learn + fix their own code before opening the PR?

New comment by winwang in "Agentic Engineering Patterns"

winwang — Wed, 04 Mar 2026 09:09:55 +0000

Linear walkthrough: I ask my agents to give me a numbered tree. Controlling tree size specifies granularity. Numbering means it's simple to refer to points for discussion.

Other things that I feel are useful:

- Very strict typing/static analysis

- Denying tool usage with a hook telling the agent why+what they should do (instead of simple denial, or dangerously accepting everything)

- Using different models for code review

New comment by winwang in "Intel's make-or-break 18A process node debuts for data center with 288-core Xeon"

winwang — Tue, 03 Mar 2026 22:08:16 +0000

If you have enough cores, you could pool the L1 together for makeshift RAM!

New comment by winwang in "Elevated Errors in Claude.ai"

winwang — Tue, 03 Mar 2026 09:21:48 +0000

Kind of agreed. I like vibe coding as "just" another tool. It's nice to review code in IDE (well, VSCode), make changes without fully refactoring, and have the AI "autocomplete". Interesting, sometimes way faster + easier to refactor by hand because of IDE tooling.

The ways that agents actually make me "faster" are typically: 1. more fun to slog through tedious/annoying parts 2. fast code review iterations 3. parallel agents

New comment by winwang in "[dead]"

winwang — Tue, 03 Mar 2026 05:00:14 +0000

Not sure, but https://status.claude.com/ uptime is pretty spotty. Funnily, the latest bar is still green despite there being an incident (and the messages even acknowledge this)

New comment by winwang in "If AI writes code, should the session be part of the commit?"

winwang — Mon, 02 Mar 2026 16:26:07 +0000

Interesting! I actually split up larger goals into two plan files: one detailed plan for design, and one "exec plan" which is effectively a build graph but the nodes are individual agents and what they should do. I throw the two-plan-file thing into a protocol md file along with a code/review loop.