Hacker News: aw123

New comment by aw123 in "How to Train a Gen AI Kick Drum Model on Your Old Linux Desktop with 6GB VRAM"

aw123 — Thu, 16 Jul 2026 18:26:39 +0000

How far are we from getting a general model that can resynthesize any instrumental audio sound without fiddling with any knobs, so that we can recreate instruments we hear from any song? Seems like it should exist by now?

New comment by aw123 in "Anthropic: Developing a Claude Code competitor using Claude Code is banned"

aw123 — Sun, 11 Jan 2026 22:39:17 +0000

I've seen dozens of "experts" all over the internet claim that they're "subsidizing costs" with the coding plans despite no evidence whatsoever. Despite the fact that various sources from OpenAI, Deepseek, model inference providers have suggested the contrary, that inference is very profitable with very high margins.

New comment by aw123 in "Anthropic: Developing a Claude Code competitor using Claude Code is banned"

aw123 — Sun, 11 Jan 2026 21:45:21 +0000

source: pulled out of your a**

New comment by aw123 in "Mercury Coder: frontier diffusion LLM generating 1000+ tok/sec on commodity GPUs"

aw123 — Fri, 28 Feb 2025 23:30:59 +0000

so if I understand correctly, you remask some tokens that were previously unmasked?

Ask HN: Should I bother learning low-level languages like Rust or CUDA?

aw123 — Sat, 21 Dec 2024 02:06:30 +0000

So I mainly do JS/TS/React and python/pytorch, basically webdev stuff and ML stuff. Rust and CUDA keep popping up over and over in my various social media feeds. I've looked into learning these, but the problem is there isn't a problem I want to solve right now that would require these. In webdev land I haven't yet run into a problem I couldn't solve by importing a package someone else wrote in nodejs. In ML land sometimes I want to make pytorch go faster, but all the complexity of CUDA seems daunting. I recently discovered Triton, and it seems that you can simply just write the code in python and add decorators.

Also numba/taichi seems to speed up python code by a ton.

With the world going so fast, should I spend any time learning these lower-level programming languages like Rust or CUDA? I'm not a hardware person btw, at least not right now. I guess I'm basically torn between thinking I should learn these things cause it would be cool and whether it's just a waste of time.

Comments URL: https://news.ycombinator.com/item?id=42476946

Points: 3

# Comments: 1

New comment by aw123 in "Liquid Foundation Models: Our First Series of Generative AI Models"

aw123 — Wed, 02 Oct 2024 10:36:07 +0000

Interesting. Though it seems like they are still incentivized to distill their bigger models downward for more tokens/s.

Ask HN: Where can I learn more about the tech stacks of various companies?

aw123 — Sat, 07 Sep 2024 02:41:08 +0000

I'm trying to study how various different companies' tech stacks. I found https://www.backblaze.com/cloud-storage/case-studies to be good. Anyone know of any other good resources?

Comments URL: https://news.ycombinator.com/item?id=41471154

Points: 3

# Comments: 0

Ask HN: How to make a globally fast search engine without using AWS?

aw123 — Wed, 28 Aug 2024 01:41:47 +0000

Forgive me as I'm sort of a novice but after looking at how sluggish some of the small search engines like Kagi are compared to Google or DDG I've been thinking about what the best most cost effective way for a single person/small team to build an app that is low latency across the world without breaking the bank and without relying heavily on the big cloud providers.

If we only have one database in one location then that wouldn't work, right? We'd need a distributed solution so what could we use? For serverless functions to start maybe we use Cloudflare workers? Maybe we'd then have a strategy for moving off cloud and towards buying/renting more datacenter space as we scale?

I'm interested in hearing thoughts about this. If you were to build a fast search engine that could scale to millions of users today, how would you do it?

Comments URL: https://news.ycombinator.com/item?id=41375187

Points: 1

# Comments: 5

New comment by aw123 in "Limitless: Personalized AI powered by what you've seen, said, and heard"

aw123 — Mon, 15 Apr 2024 21:07:21 +0000

So Rewind will be the local version and Limitless will be the cloud version? Will Rewind still be coming to windows?

Ask HN: Why don't more OSS companies adopt licenses that exclude big companies?

aw123 — Mon, 15 Apr 2024 05:47:25 +0000

I see a lot of complaining about big companies just taking OSS software and not contributing back. Why don't more companies just adopt a new license that excludes big companies from using their software in certain ways? For example: if you had 100+ million users before January 1, 2025, your company and its subsidiaries are not allowed to use or distribute this software for commercial purposes. This would level the playing field and give the small guys an advantage over big companies, lowering the barrier to entry and increasing competition. This is a net benefit for the world, no?

Comments URL: https://news.ycombinator.com/item?id=40037408

Points: 2

# Comments: 4