Hacker News: macwhisperer

New comment by macwhisperer in "AI is just unauthorised plagiarism at a bigger scale"

macwhisperer — Fri, 22 May 2026 18:20:58 +0000

can't we just freaking share things?

ai models are a crystallization of human effort (available for free on huggingface)..

why not use it?

AI is like the UBI of intelligence, stop leaving free money on the table by refusing to use it locally ( you can run it on a laptop CPU)

New comment by macwhisperer in "An LLM on a Sony PSP"

macwhisperer — Fri, 22 May 2026 00:40:26 +0000

this is really cool congrats!

New comment by macwhisperer in "Ask HN: Anyone else struggling with AI and work?"

macwhisperer — Thu, 21 May 2026 17:32:15 +0000

good, the point is that now you have free mental bandwidth to use on building something that truly interests you using AI to help actualize your goal. build something cool for your kiddos idk?

New comment by macwhisperer in "Where Are the Vibecoded Photoshops?"

macwhisperer — Mon, 18 May 2026 16:50:44 +0000

at this point I trust software companies less and less..

being able to build the stack and create bespoke solutions with llms's is incredible..

idk why people get mad about vibe-coding. if ur little brother can make a Spotify clone with Claude in a day, shouldn't that mean that you as a dev should be able to create something 100x better that makes Spotify obsolete?

good ideas / feasible novel architecture design will be the only thing valuable..

New comment by macwhisperer in "Apple Silicon costs more than OpenRouter"

macwhisperer — Sun, 17 May 2026 20:09:22 +0000

I run the latest 20b-30b models on a MacBook Air... running inference with an MoE (25 tps) for like 2 hours is like 10% battery.. (look me up on huggingface to download my models)

also you gotta realize frontier models have massive "system prompts" that clog up the context window with garbage.

being able to write your own system prompts gives you a MASSIVE edge..

New comment by macwhisperer in "SANA-WM, a 2.6B open-source world model for 1-minute 720p video"

macwhisperer — Sat, 16 May 2026 17:17:38 +0000

ai is exciting because it shows us what really matters...

New comment by macwhisperer in "Show HN: Find the best local LLM for your hardware, ranked by benchmarks"

macwhisperer — Fri, 15 May 2026 10:53:20 +0000

can you add in the other quants like IQ3_M?

also my personal simple rule of thumb for local ai sizing is:

max model size (GB) = ram (GB) / 1.65

New comment by macwhisperer in "Explore Wikipedia Like a Windows XP Desktop"

macwhisperer — Fri, 15 May 2026 10:13:32 +0000

pretty cool! needs the search function to work tho to be useful

New comment by macwhisperer in "FreeFish – Terminal Aquarium in the Browser"

macwhisperer — Thu, 07 May 2026 15:34:33 +0000

----------------------- introducing -- FreeFish! -----------------------

a free ASCII aquarium with full controls! works offline too!

note: the 'size multiplier' setting is mostly for scaling up on desktop..

----------------- about the project: -----------------

this started as a python CLI application for my own personal use...figured why not make a web version for everyone! plz enjoy..

FreeFish – Terminal Aquarium in the Browser

macwhisperer — Thu, 07 May 2026 15:34:33 +0000

Article URL: https://freefish.neocities.org

Comments URL: https://news.ycombinator.com/item?id=48050674

Points: 1

# Comments: 1

Show HN: Freelava – Lava Lamp in an HTML Page

macwhisperer — Mon, 20 Apr 2026 17:50:26 +0000

go to -- https://freelava.neocities.org --

it's a free digital lava lamp!

works on desktop and mobile..

use it as a desktop screensaver (I like to use two separate windows of it)

or just touch the goo on your mobile phone / touch screen device!

Comments URL: https://news.ycombinator.com/item?id=47838023

Points: 2

# Comments: 0

New comment by macwhisperer in "Anonymous request-token comparisons from Opus 4.6 and Opus 4.7"

macwhisperer — Sat, 18 Apr 2026 20:34:15 +0000

im currently running a custom Gemma4 26b MoE model on my 24gb m2... super fast and it beat deepseek, chatgpt, and gemini in 3 different puzzles/code challenges I tested it on. the issue now is the low context... I can only do 2048 tokens with my vram... the gap is slowly closing on the frontier models

New comment by macwhisperer in "SmartChild Is Back"

macwhisperer — Fri, 03 Apr 2026 18:26:42 +0000

WE ARE SO BACK

Today I am very excited to announce the launch of my new application "SmartChild".

------------------------ smartchild.neocities.org ------------------------

This project aims to ressurect the old AIM bot "SmarterChild", but with a modern twist!

------------ How it works: ------------ When you go to smartchild.neocities.org, the AI model will immediately install into your phone or computer's RAM.

Once it is successfully loaded, you can use the application in airplane mode or without internet! It will stay in your device until you either: refresh the page, clear browser history, or restart device. This is local on-device AI, the chats are private and the website does not collect any data (unlike cloud AI like chatgpt).

Any device that has internet access and a browser can use my application free of charge! Though it may not work on older phones or computers!

Please, tell your friends!!!

------- Credits: ------- This web application was built entirely using free, open source software and custom AI models (also open source). Special thanks to the creators of:

Python- Guido Van Rossum Llama.cpp- Georgi Gerganov WebAssembly- WC3 Wllama- Nguyen Tuan Hai Javascript- Brendan Eich Html- Sir Tim Berners-lee XP.css- Adam Hammad Neocities- Kyle Drake Huggingface- Clem Delangue Tinyllama- Peiyuan Zhang DeepSeek- Liang Wenfeng Gemini- Google DeepMind

SmartChild Is Back

macwhisperer — Fri, 03 Apr 2026 18:26:41 +0000

Article URL: https://smartchild.neocities.org

Comments URL: https://news.ycombinator.com/item?id=47630188

Points: 2

# Comments: 1