HNHacker News
TopNewBestAskShowJobs

addozhang

4 karma · joined March 3, 2022

submissionscomments
addozhang··on GLM-5.3-Flash
If you don't want to use the Chinese model, then don't. Why attack it instead? Don't you want others to use it either?
addozhang··on DeepSeek Harness
I personally really like products with plugin systems: a stable, cohesive co with a rich, extensible ecosystem. You can create products that fiyour exact needs, and even if there's no plugin that meets your requirements, you can build it yourself. At least there's vibe coding.

Just like Obsidian, there's also hot loading.

addozhang··on DeepSeek V4 Flash 0731
what's the real you mean?
addozhang··on Chinese chipmaker shares surge 470%
South Korea has opened its financial doors to the outside world, and since SK Hynix is a publicly traded company, that's allowed capital to flow in and take control of the company.
addozhang··on CarPlay Is Additive
If the cainfotainment system is good enough, I don't use CarPlay - at most I just connect Bluetooth to make calls.

The reason CarPlay and Android Auto exist is that car infotainment systems used to suck.

addozhang··on Kimi K2.7 Code is generally available in GitHub Copilot
Finally. Is this the first open weight LLM on Copilot? The gate is opened!
addozhang··on ZCode – Harness for GLM-5.2
I'd prefer a CLI over a desktop. But then why don't I just use OpenCode?
addozhang··on Claude Sonnet 5
That's really a great way for the agent to understand itself.
addozhang··on Claude Sonnet 5
I feel pretty much the same way, and the scenarios are similar too. Using Sonnet has a bigger advantage when it comes to response time.
addozhang··on Claude Sonnet 5
In the 4.x era, I prefer Sonnet to Opus. The quality of Sonnet generation is good enough for me, but it's much faster than Opus.
addozhang··on ACL 1.0: A source-available commercial license for the AI era
It's not about "banning all AI," but trying to distinguish between using Al assistants in everyday development and feeding source code into someone else's model training systems.
addozhang··on OpenRA
Sometimes, UI is also part of the classics.
addozhang··on Previewing GPT‑5.6 Sol: a next-generation model
For a large model based on statistical probability, at such a fast speed, if it executes n rounds 99.9% of the time, how much would the accuracy drop?
addozhang··on Show HN: Trace – Offline Mac meeting transcripts you can flag mid-call
This is an excellent product and exactly what I've been looking for. But most of my meetings are done on my company Mac, and they definitely won't let me install this kind of software, even though I'd be willing to pay for it myself.
addozhang··on Show HN: Trace – Offline Mac meeting transcripts you can flag mid-call
Agreed, no need to tie it into Apple either.
addozhang··on Show HN: Trace – Offline Mac meeting transcripts you can flag mid-call
I don't really recommend it. If the software is a one-time purchase, there's no need to rewrite it with an LLM. Rewriting the tokens could cost more than just $10.
addozhang··on MiMo-v2.5-Pro-UltraSpeed: 1T model with 1000 tokens per second
Yes, you don't need much domain knowledge to use Opus, but it's just way too expensive.
addozhang··on Show HN: How to create Pipy service with hello example
Great job!