Hacker News: sh4rks

New comment by sh4rks in "HERMES.md in commit messages causes requests to route to extra usage billing"

sh4rks — Wed, 29 Apr 2026 20:28:06 +0000

_puts pitchfork away_

New comment by sh4rks in "Tell HN: Claude 4.7 is ignoring stop hooks"

sh4rks — Fri, 24 Apr 2026 22:37:56 +0000

Coding agents are unusable without skills and mcp tools

New comment by sh4rks in "Are the costs of AI agents also rising exponentially? (2025)"

sh4rks — Sat, 18 Apr 2026 10:51:38 +0000

Writing code by hand is gambling (will it compile or not, will it pass code review or not)

New comment by sh4rks in "How long-distance couples use digital games to facilitate intimacy (2025)"

sh4rks — Mon, 13 Apr 2026 07:57:34 +0000

Anyone know any good cooperative board/card games for two that are:

1. Easy to learn and understand 2. Easy to buy and relatively inexpensive?

New comment by sh4rks in "Bird brains (2023)"

sh4rks — Thu, 02 Apr 2026 10:22:16 +0000

If you think being a guide dog doesn't require intelligence, you're delusional.

> answer the actual question

I literally did. You asked, which task that requires intelligence would I rather use an animal over an Llm. I'd much rather have a dog as my guide dog than an Llm. It can use it's innate intelligence to sense danger, navigate around obstacles it's never seen before, and even communicate with other humans through barking.

> trained clearly means it is not something based in intelligence, but in repetition and conditioning

I can't tell if you're trolling at this point. Llms are also trained and therefore are based on repetition and conditioning.

New comment by sh4rks in "Bird brains (2023)"

sh4rks — Tue, 31 Mar 2026 17:09:22 +0000

Dogs can be trained for a variety of tasks that require wit. Such as helping the blind navigate. And sniffing for illegal drugs.

New comment by sh4rks in "Tin Can, a 'landline' for kids"

sh4rks — Mon, 23 Mar 2026 15:38:47 +0000

That looks very ugly. Any kid above the age of 12 would reject that.

New comment by sh4rks in "The worst volume control UI in the world (2017)"

sh4rks — Sat, 21 Mar 2026 12:20:55 +0000

From the article:

> I’m sure this problem was solved years and years ago, forgotten, rediscovered, solved again, forgotten, rediscovered, solved again.

New comment by sh4rks in "US Job Market Visualizer"

sh4rks — Mon, 16 Mar 2026 21:35:11 +0000

fix(ui): improve accessibility and screen reader compatibility

Co-authored-by: Claude

New comment by sh4rks in "Ask HN: Why is my Claude experience so bad? What am I doing wrong?"

sh4rks — Mon, 16 Feb 2026 08:39:47 +0000

Sometimes people forget that you don't have to use AI to actually write the code. You can stick to "Ask" mode and it will give you useful suggestions and generate code but won't actually modify your files.

New comment by sh4rks in "Coding assistants are solving the wrong problem"

sh4rks — Thu, 05 Feb 2026 20:54:15 +0000

> developers can use any tools they choose (primarily Cursor Pro with Claude 3.5/3.7 Sonnet—frontier models at the time of the study

Sonnet 3.5 came out in mid 2024

New comment by sh4rks in "OpenClaw – Moltbot Renamed Again"

sh4rks — Fri, 30 Jan 2026 13:48:55 +0000

I want to use Gemini CLI with OpenClaw(dbot) but I'm too scared to hook it up to my primary Google account (where I have my Google AI subscription set up)

New comment by sh4rks in "Show HN: Moltbook – A social network for moltbots (clawdbots) to hang out"

sh4rks — Fri, 30 Jan 2026 10:15:52 +0000

People are still falling for the "stochastic parrot" meme?

New comment by sh4rks in "eBay explicitly bans AI "buy for me" agents in user agreement update"

sh4rks — Thu, 22 Jan 2026 15:05:39 +0000

Dead Internet theory

New comment by sh4rks in "India has surpassed Japan to become the fourth-largest economy"

sh4rks — Sun, 04 Jan 2026 17:04:07 +0000

Is that using nominal GDP or PPP?

New comment by sh4rks in "The future of software development is software developers"

sh4rks — Tue, 30 Dec 2025 08:38:25 +0000

Post model

New comment by sh4rks in "ChatGPT conversations still lack timestamps after years of requests"

sh4rks — Sat, 27 Dec 2025 01:20:58 +0000

Ah, the casino tactic

New comment by sh4rks in "Sandbox: Run untrusted AI code safely, fast"

sh4rks — Fri, 26 Dec 2025 21:45:09 +0000

How is this different from the several other alternatives?

New comment by sh4rks in "Ask HN: What Are You Working On? (December 2025)"

sh4rks — Mon, 15 Dec 2025 07:25:58 +0000

Wouldn't it be easier to just install a bolt lock on your door?

New comment by sh4rks in "Show HN: An endless scrolling word search game"

sh4rks — Fri, 12 Dec 2025 12:33:18 +0000

Ha. I read this comment and found the exact same diagonal "Ape"