Hacker News: mingqiz

New comment by mingqiz in "Our position on open-weights models"

mingqiz — Tue, 28 Jul 2026 02:28:13 +0000

Makes sense. Let the rest of the world use open weight models and let us government review both closed models and open models, and only allow us citizen and American companies use the censored models by paying selected trusted providers who unfortunately needs to charge hefty fee for the additional security work. And that provider just happens to be Anthropic who just happens to unavoidably make some money.

New comment by mingqiz in "Kimi K3, Qwen 3.8, and Anthropic's (Potential) Unravelling"

mingqiz — Mon, 20 Jul 2026 16:26:47 +0000

Not true. Enterprise pay for api. If i am allowed to use fable withouta budget, i can spend 3k per day. Using combination of cheaper models (and better at least in my experience) can lower this to 300. Considering the layoffs, thw companiesdo care a lot of this cost. Price matters more than quality for many use cases. For example, i can writea game where npc are all k3/5.6 level intelligence if the price is cheap enough.

New comment by mingqiz in "GPT-5.6"

mingqiz — Fri, 10 Jul 2026 13:59:05 +0000

This. Claude is very good at instruction and treat my mistake in prompt as law. Gpt is just smarter at understanding my intent.

New comment by mingqiz in "GPT-5.6"

mingqiz — Fri, 10 Jul 2026 13:47:34 +0000

It's the other way actually. Cc is a better harness but gpt models are just so much better in my personal experience (at least for backend) ever since 5.4.

New comment by mingqiz in "Bun has an open PR adding shared-memory threads to JavaScriptCore"

mingqiz — Sun, 21 Jun 2026 07:39:14 +0000

I honestly trust code reviewed by llms multiple rounds more than human reviewed code. Here is the thing. I don't know the human who wrote the code but i am familiar with the model they use, which might already be better than many human.

New comment by mingqiz in "The Token Compression Illusion: Why I'm Skeptical of RTK"

mingqiz — Thu, 18 Jun 2026 21:44:42 +0000

But Claude especially copy opensource ideas after they are widely used for months

New comment by mingqiz in "Apple Foundation Models"

mingqiz — Mon, 15 Jun 2026 16:43:29 +0000

Isn't that what they are doing already? The model is already guarded and hidden and i only get to send it what i want. Talk with it to clarify my requirements. And i can switch to a different provider for cheaper/better results.

New comment by mingqiz in "The current AI pricing was always going to go away"

mingqiz — Fri, 22 May 2026 18:53:14 +0000

Deepseek dropped their price permanently.Now v4 pro costs 3.48% of the output token price of opus 4.7. (now that opus 4.7 has a more costly tokenizer). Is it fair to say that Anthropic/OpenAI is marching to bankruptcy as opensource models like deepseek improves? Note that deepseek also supports 500 concurrent requests per account for any individuals for v4 pro.