Hacker News: anabis

New comment by anabis in "Kimi-K3 on HuggingFace"

anabis — Tue, 28 Jul 2026 07:45:51 +0000

I heard Muse Spark 1.1 isn't too shabby. Since Meta compute >> Chinese labs, we'll see how it goes from here.

New comment by anabis in "Residential Proxies Are a National Security Threat"

anabis — Tue, 28 Jul 2026 03:38:31 +0000

It's an anti-consumer evil practice, but if it is a national threat that's on the nation's security infra.

New comment by anabis in "Anthropic says Alibaba illicitly extracted Claude AI model capabilities"

anabis — Thu, 25 Jun 2026 02:30:14 +0000

Incentive is for users in general to release sessions (sans PII, credentials) so all AI get better and there is alternatives. Even if China didn't do this, I don't see frontier labs being able to charge premium over others for long. RSI maybe?

New comment by anabis in "Anthropic apologizes for invisible Claude Fable guardrails"

anabis — Fri, 12 Jun 2026 07:27:47 +0000

OpenAI did this first.

> In addition to safety training, automated classifier-based monitors detect signals of suspicious cyber activity and route high-risk traffic to a less cyber-capable model (GPT-5.2).

https://developers.openai.com/codex/concepts/cyber-safety

New comment by anabis in "Rich Sutton on AI creativity and discovery"

anabis — Wed, 10 Jun 2026 05:46:49 +0000

He seems to be saying that Claude Code can make discoveries. Does anyone think novel discoveries can be made from systems created by supervised learning only, and attempting to do so?

> Claude-Code, which have brought true advances in ... programming. ... these systems have found things that are both novel and good.

New comment by anabis in "The Eternal Sloptember"

anabis — Tue, 26 May 2026 00:33:59 +0000

Yet the Eternal September is what made the modern AI possible. I asked Cowork for share of the corpus before / after the event, and the corpus before it is <1%, which fits my hunch.

New comment by anabis in "Copy Fail"

anabis — Thu, 30 Apr 2026 04:20:18 +0000

Many things, such as ksmbd seems ill-advised when looked at from security. New AI driven exploits era will likely make projects more wary to adding functions.

New comment by anabis in "GPT-5.5"

anabis — Fri, 24 Apr 2026 06:45:45 +0000

Laziness is a virtue, but when I asked GPT-5.4 to test scenarios A and B with screenshots, it re-used screenshots from A for B, defeating the purpose.

New comment by anabis in "Over-editing refers to a model modifying code beyond what is necessary"

anabis — Thu, 23 Apr 2026 00:56:50 +0000

Are you mostly using the Composer model?

New comment by anabis in "Changes to GitHub Copilot individual plans"

anabis — Wed, 22 Apr 2026 06:37:24 +0000

Copilot was there in AI based development first with tab completions.

Now, it may be the right call to immediately give up and shutdown after Opus 4.5, but models and subscriptions are in flux right now, so the right call is not at all obvious to me.

The agentic AI models could be commoditized, some model may excel in one area of SWE, while others are good for another area, local models may be at least good enough for 80%, and cloud usage could fall to 20%, etc. etc.

Staying in the market and providing multi-model and harness options (Claude and Codex usable in Copilot) is good for the market, even if you don't use it.

New comment by anabis in "Changes to GitHub Copilot individual plans"

anabis — Wed, 22 Apr 2026 01:10:38 +0000

AI should decide the level of model needed, and fallback if it fails. It mostly is a UX problem. Why do I need to specify the level of model beforehand? Many problems don't allow decision pre-implementation.

New comment by anabis in "Android CLI: Build Android apps 3x faster using any agent"

anabis — Sat, 18 Apr 2026 01:30:26 +0000

Thx. In other comment, but setting Java options for proxy settings worked. I was guessing annotations would be based on UI framework borders(which I can toggle on screen via debug menu), but looks it's based on image recognition locally. Maybe you can have UI component borders as option(haven't heavily used it yet, so throwing around ideas)

New comment by anabis in "Android CLI: Build Android apps 3x faster using any agent"

anabis — Sat, 18 Apr 2026 00:58:27 +0000

MS showing "view summary" button for all meetings, then doing bait-and-switch to tell you to buy Copilot license (on a corporate seat no less, where regular users don't have purchasing decision power) is top annoyance now

New comment by anabis in "Android CLI: Build Android apps 3x faster using any agent"

anabis — Sat, 18 Apr 2026 00:54:02 +0000

Tryed both. Copilot said it was wrongly utf16, but it was utf8 according to vscode.

New comment by anabis in "Android CLI: Build Android apps 3x faster using any agent"

anabis — Fri, 17 Apr 2026 06:16:14 +0000

I got a workaround a la GH Copilot:

> android skills list

Picked up JAVA_TOOL_OPTIONS: -Djava.net.useSystemProxies=true

New comment by anabis in "Android CLI: Build Android apps 3x faster using any agent"

anabis — Fri, 17 Apr 2026 03:39:26 +0000

Taking screenshots, optionally with component borders highlighted, and operating the UI with element names like "button1" instead of tap 200,30 looks useful. If I could get it to work.

New comment by anabis in "Android CLI: Build Android apps 3x faster using any agent"

anabis — Fri, 17 Apr 2026 03:32:39 +0000

The install command shown for Windows is 404.

`curl -fsSL https://dl.google.com/android/cli/latest/windows_x86_64/inst... | bash`

The URL shown for individual OSs work, but the script errors for me.

`curl.exe -fsSL https://dl.google.com/android/cli/latest/windows_x86_64/inst... -o "%TEMP%\i.cmd" && "%TEMP%\i.cmd"`

I manually downloaded the exe, but it say socket error. vibe coding is going strong!

New comment by anabis in "Project Glasswing: Securing critical software for the AI era"

anabis — Wed, 08 Apr 2026 01:58:42 +0000

Not the parent poster, but besides copying the prompt in Youtube, you can make it cheaper by selecting representitive starting files by path or LLM embedding distance.

Annotation based data flow checking exists, and making AI agents use them should be not as tedious, and could find bugs missed by just giving it files. The result from data flow checks can be fed to AI agents to verify.

New comment by anabis in "Show HN: I made a YouTube search form with advanced filters"

anabis — Mon, 06 Apr 2026 09:18:23 +0000

gemini @YouTube is decent too.

New comment by anabis in "Claude Code's source code has been leaked via a map file in their NPM registry"

anabis — Wed, 01 Apr 2026 06:31:06 +0000

Yeah, I thought Anthropic was for AI safety. Telling Ai to not be honest is a bad sign.