Hacker News: ericlewis

New comment by ericlewis in "Teenage Engineering's PO-32 acoustic modem and synth implementation"

ericlewis — Thu, 02 Apr 2026 14:29:00 +0000

much simpler, just pull apart the MT VST :P

I am working on PO-33, doing EMFI.

Teenage Engineering's PO-32 acoustic modem and synth implementation

ericlewis — Sat, 28 Mar 2026 01:04:56 +0000

Article URL: https://github.com/ericlewis/libpo32

Comments URL: https://news.ycombinator.com/item?id=47550433

Points: 148

# Comments: 36

New comment by ericlewis in "Grok 4 Launch [video]"

ericlewis — Thu, 10 Jul 2025 17:01:20 +0000

I still dont understand why people point to this chart as any sort of meaning. Cost per task is a fairly arbitrary X axis and in no way representing any sort of time scale.. I would love to be told how they didn't underprice their model and give it an arbitrary amount of time to work.

New comment by ericlewis in "Man suffers chemical burn that lasted months after squeezing limes"

ericlewis — Fri, 29 Nov 2024 22:10:04 +0000

US, GA here. My mom was big on tanning and warned of us this (lemons are also bad). I believe she said something about it being used on purpose for tanning, but that you had to be careful or you would badly burn. She probably did that around the late 80s or early 90s.

New comment by ericlewis in "Private Cloud Compute Security Guide"

ericlewis — Wed, 06 Nov 2024 18:24:24 +0000

As a person who works on this kinda stuff I know what they mean. It’s very hard to debug things totally blind.

New comment by ericlewis in "Addition Is All You Need for Energy-Efficient Language Models"

ericlewis — Wed, 09 Oct 2024 13:55:30 +0000

Higher the precision the better. Use what works within your memory constraints.

New comment by ericlewis in "OpenAI Threatening to Ban Users for Asking Strawberry About Its Reasoning"

ericlewis — Wed, 18 Sep 2024 22:07:11 +0000

If you are using the user attribution with OpenAI (as you should) then they will block that users id and the rest of your app will be fine.

New comment by ericlewis in "OpenAI threatens to revoke o1 access for asking it about its chain of thought"

ericlewis — Sat, 14 Sep 2024 00:54:39 +0000

The theory is that you increase the context with more relevant tokens to the problem at hand, as well as its solutions, which in theory makes it more likely to predict the correct solution.

New comment by ericlewis in "Hardware Acceleration of LLMs: A comprehensive survey and comparison"

ericlewis — Mon, 09 Sep 2024 01:45:50 +0000

This reminds of a TPU.

New comment by ericlewis in "Comma.ai: Refactoring for Growth"

ericlewis — Mon, 12 Aug 2024 05:53:45 +0000

I have one of these. It’s not there.

New comment by ericlewis in "ChatGPT unexpectedly began speaking in a user's cloned voice during testing"

ericlewis — Sun, 11 Aug 2024 19:36:32 +0000

Perhaps it’s because I know human beings that have the exact same operation and failure mode as the LLM here and I’m probably not the only one. Failing at something you’ve never seen and faking through it is a very human endeavor.

New comment by ericlewis in "Tesla's FSD – A Useless Technology Demo"

ericlewis — Mon, 17 Jun 2024 02:13:55 +0000

My wife is capable of “automatic driving” and I am not. She describes it as a flow state.

New comment by ericlewis in "Ask HN: Help me understand RAG vs. fine tuning for building a coding partner"

ericlewis — Mon, 17 Jun 2024 01:40:38 +0000

You would need both. RAG will have to be part of the process if you want to generate or work on an existing project. Since it stands for Retrieval Augmented Generation. A fancy way of saying an LLM can access documents.

New comment by ericlewis in "Ask HN: Help me understand RAG vs. fine tuning for building a coding partner"

ericlewis — Mon, 17 Jun 2024 01:37:50 +0000

GPT-4 is fine-tunable.

https://platform.openai.com/docs/guides/fine-tuning

“ Fine-tuning for GPT-4 is in an experimental access program - eligible users can request access in the fine-tuning UI when creating a new fine-tuning job.”

You need only have fine-tuned a 3.5 model ime.

New comment by ericlewis in "Ask HN: Predictions for when GPT-5 will be released and how safe it will be?"

ericlewis — Mon, 10 Jun 2024 13:53:08 +0000

ChatGPT is in-fact that. GPT-4 the API is not that. There is clearly no RAG, you bring that yourself, same with toxicity filtering, again - a different api, up to you to implement. etc. There is probably something going on before the input reaches GPT-4 and we are definitely not seeing the raw output from the LLM. But the layers over GPT-4 the API are thinner than ChatGPT.

The Assistants API is closer to ChatGPT & what you are describing.

Edit: it should be noted that the Assistants API is somewhat model agnostic as well, so the product part of this isn't part of the inference system.

New comment by ericlewis in "Anthropic Courses"

ericlewis — Wed, 05 Jun 2024 00:33:05 +0000

It’s kinda specific to Claude, but the overall concept and really a lot of the api shape is similar for most other LLMs, especially ones that adopt the OpenAI api model.

This is probably best described as a tutorial for how to use tools with Claude.

New comment by ericlewis in "React 19 Beta"

ericlewis — Sat, 27 Apr 2024 02:34:37 +0000

useEffect is best thought of as a synchronization mechanism between external state and react state and the function it returns may actually be better thought of as a side effect. UseEffect makes sense in this case because you aren’t on the side!

New comment by ericlewis in "Humane AI – Pico Laser Projection – AI Twist on an Old Scam (2023)"

ericlewis — Thu, 18 Apr 2024 00:21:56 +0000

I wanted to replicate many parts of the Ai pin and did research into laser projectors. It’s probably a MEMs projector. The cyan color they use is custom, at least. Minifying that tech and making it robust enough to be knocked around on a wearable device is pretty challenging and I couldn’t find anything of a similar size off the shelf. They also run pretty hot and apparently the pin has heat issues (and according to folks at CES (?) using the laser a lot makes it over heat.

New comment by ericlewis in "Show HN: CompressX, my FFmpeg wrapper for macOS"

ericlewis — Wed, 10 Apr 2024 17:49:39 +0000

Not that it really matters I guess, but electron apps are chunky. Huge bin size and large memory usage.

New comment by ericlewis in "Google Axion Processors – Arm-based CPUs designed for the data center"

ericlewis — Tue, 09 Apr 2024 14:35:07 +0000

Xcode cloud isn’t running on Apple silicon, arguably a place where it would make tons of sense.