as1as

3 karma · joined March 31, 2026

Solo developer building TalkWith.chat — daily AI debate battles with evolving personal AI debaters. Currently 0 users. Sharing the full build journey (mistakes included).

submissions comments

as1as··on Show HN: SwiftLM – Qwen Chat on iPhone, 100B+ Moe on M5 Pro 64GB (Native Swift)

I'm always worried about the performance, but thank you for introducing a good project.

as1as··on Show HN: Margo – Find the font your brain reads fastest

I think it could be used as a reading practice for kids who have a hard time reading. It's a wonderful project.

as1as··on Show HN: I built a self-hosted Fly.io engine using Go and Firecracker

I'm planning to upload the model I tuned to the cloud, which would be useful, I'll try it later. Thank you

as1as··on Show HN: Sundial – a new way to look at a weather forecast

That's a great design, I don't think the sunset time is accurate, do you have any plans to apply it?

as1as··on Ask HN: M5 MacBook Pro buyers, worth spending the $$$ to maybe run LLMs local?

I want to buy a Mac Studio m4 max.. I can barely run an 8b model on my device right now and I can't even dream about fine tuning.

as1as··on Ask HN: Gemini CLI vs. Claude Code

I think Claude is far better than Gemini, the lack of usage is a chronic problem. Even using the Max model is not enough. If I used that much, I think I worked a lot and took a rest. LOL.

as1as··on Show HN: Livestream Robot Cooking

This is hilarious and surprisingly cool. Real-time robot cooking via prompts is such a fun demo. Well done!

as1as··on Show HN: The Alphabetical Clock

The clock is moving very dizzy, lol

as1as··on Show HN: Free AI API gateway that auto-fails over Gemini, Groq, Mistral, etc.

Simple but very useful. Love the idea of unifying free AI providers with automatic failover. Great job!