vinhnx

6 ms·

61.
GPT-5.6 – August Updates [pdf](cdn.openai.com)
2 points by vinhnx 1mo ago|0 comments
62.
by vinhnx 1mo ago
Small model-facing surface, more responsibility in the harness. I'm exploring that tradeoff in Rust with VT Code: https://github.com/vinhnx/VTCode
63.
by vinhnx 1mo ago
I'm building roughly this in Rust: single binary, XDG-native, with Landlock/seccomp, scoped filesystem access, default-deny egress, and policy-based approvals. https://github.com/vinhnx/VTCode
64.
Designing a Query System(bal-e.org)
2 points by vinhnx 1mo ago|0 comments
65.
BorrowSanitizer(borrowsanitizer.com)
3 points by vinhnx 1mo ago|0 comments
66.
Tests for a PDF(blog.lvmbdv.dev)
2 points by vinhnx 1mo ago|0 comments
67.
Do-178B(en.wikipedia.org)
2 points by vinhnx 1mo ago|0 comments
68.
Why is it all in the kernel?(lawrencecpaulson.github.io)
1 points by vinhnx 1mo ago|1 comments
69.
Indirected Reality – Language, Truth, and LLMs(queue.acm.org)
1 points by vinhnx 1mo ago|1 comments
70.
by vinhnx 2mo ago
Because Luna feels right and just enough for me.
71.
I'm (mostly) picking models on speed now, not intelligence(martinalderson.com)
3 points by vinhnx 2mo ago|0 comments
72.
by vinhnx 2mo ago
This month marks 1 year since I first started self-research and building my own coding agent VT Code [0]. I've been learning and experiment with a lot of things to first adapt myself and to learn about AI native/harness engineerin
73.
Flint: A Visualization Language for the AI Era(microsoft.github.io)
278 points by vinhnx 2mo ago|69 comments
74.
by vinhnx 2mo ago
GPT-5.6 Luna xhigh/max from Codex is my daily driver now. Good economics, good performance. It sits in the most attractive quadrant on the Intelligence Index.
75.
by vinhnx 2mo ago
I reached a mid-run usage cap and burn all my Plus subscription usage. I'm hoping there will be another reset.
76.
Rio-vt and librio: Rio's terminal engine, now embeddable(rioterm.com)
54 points by vinhnx 2mo ago|17 comments
77.
Kimi K3 developed MiniTriton: a compact Triton-like compiler(github.com)
2 points by vinhnx 2mo ago|0 comments
78.
Kimi-K3 Technical Report [pdf](github.com)
391 points by vinhnx 2mo ago|183 comments
79.
by vinhnx 2mo ago
Kimi K3's full weight is now live on Hugging Face https://huggingface.co/moonshotai/Kimi-K3
80.
by vinhnx 2mo ago
Less than 2 hours left to the Kimi moment. It’s been more than 2 years since the DeepSeek moment that shook the world.
81.
Teach yourself programming in ten years (1998)(norvig.com)
185 points by vinhnx 2mo ago|132 comments
82.
A Visual Journey Through Async Rust(github.com)
3 points by vinhnx 2mo ago|0 comments
83.
Git rebase -i is not that scary(cachebag.sh)
119 points by vinhnx 2mo ago|148 comments
84.
Systems and Delays(martin.janiczek.cz)
84 points by vinhnx 2mo ago|17 comments
85.
Zig by Example(zigbyexample.neocities.org)
4 points by vinhnx 2mo ago|0 comments
86.
The End of Physical Games: What We Lost When Games Went Digital [video](youtube.com)
2 points by vinhnx 2mo ago|0 comments
87.
by vinhnx 2mo ago
Fair criticism, I understand your finding. It's not everyone's taste, when it from AI-generated contents.
88.
by vinhnx 2mo ago
Thank you, I have had updated the video link here Video Special | Anthropic's Claude Opus 5 + https://www.youtube.com/watch?v=8Vdofv2vQ_M + https://www.youtube.com/watch?v=q-jHHx3J8m8
89.
by vinhnx 2mo ago
The benchmark appears to have a mistake, as Opus 5 and Fable 5 score 53.4% and 53.5%, respectively, for the Agentic Coding row (FrontierCode v1.1). But Opus 5 is the highlight.
90.
by vinhnx 2mo ago
For anyone wanting a faster overview, I used NotebookLM to create a brief video summary after going through the system card and announcement blog using a cinematic video overview. Link: https://www.youtube.com/watch?v=SUFBhv

More ›