Hacker News: rockinghigh

New comment by rockinghigh in "EU to legislate about Chat Control behind closed doors"

rockinghigh — Sun, 28 Jun 2026 15:48:01 +0000

The US is also falling behind Chinese manufacturing. They had to ban Chinese cars because legacy American automakers couldn't compete.

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

rockinghigh — Fri, 12 Jun 2026 02:18:36 +0000

Same here, I fine tune LLMs for specific use cases. How can I trust Anthropic models not to introduce bugs to preserve their moat?

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

rockinghigh — Fri, 12 Jun 2026 02:15:59 +0000

Imagine if your IDE started injecting bugs into your project just because your code looked like it implemented a competing IDE.

New comment by rockinghigh in "That Methyl Methacrylate Tank"

rockinghigh — Tue, 26 May 2026 22:20:24 +0000

Consumers don't control zoning laws or risk mitigation details.

New comment by rockinghigh in "The IBM-ification of Google?"

rockinghigh — Fri, 22 May 2026 03:37:27 +0000

Gemini tends to be faster and the Flash and Flash Lite models outperform ChatGPT's equivalent models.

New comment by rockinghigh in "We're testing new ad formats in Search and expanding our Direct Offers pilot"

rockinghigh — Thu, 21 May 2026 22:37:06 +0000

It may be a push from advertisers who want access to this format. Google Search competes for their money against the competition.

New comment by rockinghigh in "We're testing new ad formats in Search and expanding our Direct Offers pilot"

rockinghigh — Thu, 21 May 2026 22:34:13 +0000

They own tons of engagement data associated with that index from being the default search engine on most devices.

New comment by rockinghigh in "Anthropic acquires Stainless"

rockinghigh — Mon, 18 May 2026 18:53:17 +0000

Why wouldn't getting more customers the plan? Anthropic doesn't acquire companies to have a lower market share. There is clearly a consolidation and a rush to get as much of the developer market as possible.

New comment by rockinghigh in "Kimi K2.6: Advancing open-source coding"

rockinghigh — Mon, 20 Apr 2026 19:23:57 +0000

The MoE experts are quantized to int4, all other weights like the shared expert weights are excluded from quantization and use bf16.

New comment by rockinghigh in "Kimi K2.6: Advancing open-source coding"

rockinghigh — Mon, 20 Apr 2026 19:17:40 +0000

They could release data to back up that claim.

New comment by rockinghigh in "Kimi K2.6: Advancing open-source coding"

rockinghigh — Mon, 20 Apr 2026 19:16:04 +0000

Are there any protections from industrial espionage when using Anthropic, Cursor, Gemini, or OpenAI?

New comment by rockinghigh in "Oracle slashes 30k jobs"

rockinghigh — Tue, 31 Mar 2026 16:06:12 +0000

Their revenue was $57.4 billion last year. Just in Q4; cloud revenue $6.7 billion, cloud infrastructure $3.0 billion, cloud application $3.7 billion, Fusion Cloud ERP $1.0 billion, NetSuite cloud ERP $1.0 billion.

New comment by rockinghigh in "Leanstral: Open-source agent for trustworthy coding and formal proof engineering"

rockinghigh — Mon, 16 Mar 2026 22:18:54 +0000

It's the number of attempts at answering the question.

New comment by rockinghigh in "Yann LeCun raises $1B to build AI that understands the physical world"

rockinghigh — Wed, 11 Mar 2026 00:24:13 +0000

He founded the team that worked on fasttext, llama and other similarly impactful projects.

New comment by rockinghigh in "Yann LeCun raises $1B to build AI that understands the physical world"

rockinghigh — Wed, 11 Mar 2026 00:22:41 +0000

He founded FAIR and the team in Paris that ultimately worked on the early Llama versions.

New comment by rockinghigh in "Tony Hoare has died"

rockinghigh — Tue, 10 Mar 2026 21:10:52 +0000

It can also be used to simplify existing code bases.

New comment by rockinghigh in "Claude Sonnet 4.6"

rockinghigh — Wed, 18 Feb 2026 06:32:47 +0000

It's a lot simpler. These models are not optimized for ambiguous riddles.

New comment by rockinghigh in "GLM-5: Targeting complex systems engineering and long-horizon agentic tasks"

rockinghigh — Wed, 11 Feb 2026 18:14:45 +0000

How is this riddle relevant to a coding model?

New comment by rockinghigh in "GLM-5: Targeting complex systems engineering and long-horizon agentic tasks"

rockinghigh — Wed, 11 Feb 2026 18:12:40 +0000

They do. Pretty much all agentic models call linting, compiling and testing tools as part of their flow.

New comment by rockinghigh in "AI makes the easy part easier and the hard part harder"

rockinghigh — Mon, 09 Feb 2026 00:55:51 +0000

It's called problem decomposition and agentic coding systems do some of this by themselves now: generate a plan, break the tasks into subgoals, implement first subgoal, test if it works, continue.