Hacker News: eyk19

New comment by eyk19 in "GPT-6 Sol and Luna"

eyk19 — Tue, 22 Sep 2026 18:17:21 +0000

Luna really is "intelligence to cheap to meter" by now

New comment by eyk19 in "It's getting harder to focus every day"

eyk19 — Fri, 24 Jul 2026 14:03:18 +0000

As much as I love and embrace LLMs, they are clearly eliminating all flow state that was previously inevitable when coding something

New comment by eyk19 in "Local AI needs to be the norm"

eyk19 — Sun, 10 May 2026 20:33:18 +0000

Apple stock is going to skyrocket

New comment by eyk19 in "Redox OS has adopted a Certificate of Origin policy and a strict no-LLM policy"

eyk19 — Tue, 10 Mar 2026 10:12:36 +0000

The point isn't that agent output is magically better; it's that reviewing your own agent's output is way cheaper (intellectually) than reviewing a stranger's, because you've written the plan by yourself. And 'slop' is mostly what you get when you don't have a clear plan to verify against. Maintainers writing detailed specs for their own agents is a very different thing from someone vibe coding a feature request

New comment by eyk19 in "Redox OS has adopted a Certificate of Origin policy and a strict no-LLM policy"

eyk19 — Tue, 10 Mar 2026 10:10:30 +0000

Definitely, but that's kind of my point: the maintainers are still going to be way better at all of that than some random contributor who just wants a feature, vibe codes it, and barely tests it. The maintainers already know the codebase, they understand the implications of changes, and they can write much better plans for the agent to follow, which they can verify against. Having a great plan written down that you can verify against drastically lowers the risk of LLM-generated code

New comment by eyk19 in "Redox OS has adopted a Certificate of Origin policy and a strict no-LLM policy"

eyk19 — Tue, 10 Mar 2026 09:40:54 +0000

I feel like the pattern here is donate compute, not code. If agents are writing most of the software anyway, why deal with the overhead of reviewing other people's PRs? You're basically reviewing someone else's agent output when you could just run your own.

Maintainers could just accept feature requests, point their own agents at them using donated compute, and skip the whole review dance. You get code that actually matches the project's style and conventions, and nobody has to spend time cleaning up after a stranger's slightly-off take on how things should work.

Apple: STARFlow-V, a Normalizing Flow Model for Causal Video Generation

eyk19 — Mon, 01 Dec 2025 17:14:24 +0000

Article URL: https://huggingface.co/apple/starflow

Comments URL: https://news.ycombinator.com/item?id=46109997

Points: 4

# Comments: 0

New Version of Siri to 'Lean' on Google Gemini

eyk19 — Mon, 03 Nov 2025 09:32:09 +0000

Article URL: https://www.macrumors.com/2025/11/02/new-version-of-siri-to-lean-on-google-gemini/

Comments URL: https://news.ycombinator.com/item?id=45797316

Points: 3

# Comments: 0

New comment by eyk19 in "Migrating from AWS to Hetzner"

eyk19 — Fri, 17 Oct 2025 10:42:54 +0000

We've experienced something similar: for compute-heavy rendering tasks, AWS just wasn't good enough. EC2 machines with the same spec perform much worse than Hetzner machines

Daily Aspirin intake slashes colon cancer relapse risk by 55%

eyk19 — Fri, 19 Sep 2025 09:15:09 +0000

Article URL: https://news.ki.se/common-inexpensive-drug-halves-recurrence-in-colorectal-cancer

Comments URL: https://news.ycombinator.com/item?id=45299590

Points: 25

# Comments: 14