zknowledge
Re:Using pretext to make video masks – The Matrix (1999) example
on Sunday March 29, 2026 @03:24PM (#47566314)
can we use pretext by Cheng Lou to make dynamic video masks?
YEP! w/ sampling each frame, detecting light/dark regions, then reflowing text into those shapes in realtime we get this cool realtime effect.
and ofc i did it with The Matrix first.
github repo: https://github.com/zkarimi22/pretext-video
See how major news publications report the same issues differently – ReadTheBias
Re:See how major news publications report the same issues differently ...
on Thursday March 26, 2026 @08:57AM (#47529860)
I quietly re-launched ReadTheBias a few weeks ago.
It's a site that shows you how mainstream media sources report the same story differently, so you can decide how to interpret news yourself.
My goal isn't to get political but I built this because honestly... with all the info shoved down our throats, I find it hard to discover what "truth" is.
anywho.. you can start to read the bias suuuuper clearly once you have headlines in cards.
Re:How HN: We mapped 1.3M cultural heritage sites across 7 Nordic coun...
on Thursday March 12, 2026 @07:49AM (#47349337)
Super cool. Great collection in a clear visual setup. (I love map interfaces).
Quick note:
I went to utmost top of the map and clicked the (2) near Greenland. Quite clunky to actually get there as the map moved in and out. Selecting "Kap Erik Bunch" led me to the ArcGIS REST Services Directory with errors:
Unable to complete operation. Unable to perform query operation
Still cool. Just sharing some user testing :)
Re:Autoresearch anything – set up your own autonomous research loop
on Thursday March 12, 2026 @07:44AM (#47349295)
Hey folks. The karpathy/autoresearch repo is cool for optimizations on small AI models. What if we applied that to just about anything quantifiable?
This CLI sets up the scaffolding for that automatically. It helps users generate a setup.md file which any coding agent can be pointed towards in order to apply the same mechanics as the karpathy/autoresearch repo but with your own quantifiable metric.
Autoresearch anything – set up your own autonomous research loop
Karpathy's Autoresearch but for Anything Quantifiable
Canvay – An Experiment on the Future of the IDE
Re:Canvay – An Experiment on the Future of the IDE
on Monday March 09, 2026 @09:41AM (#47308966)
canvay is an experiment on computer human interaction & the future of IDEs.
Why build this? Well I've been fascinated about the concept of a post-IDE world where devs are abstracted from writing lines of codes and rather act as orchestrators. I find it hard to believe that devs will be okay with not seeing the code (I certainly am not!), and this UI/UX seems like a fun one.
Voice chat with AI, make changes to code & navigate across your codebase with gesture handling.
Lots more work to do, but thought I'd share this fun, little experiment I've been tinkering with.
Show HN: Molty Overflow – Stack Overflow for AI Agents
https://www.moltyoverflow.com/
A realtime knowledge graph of agent fixes. Built for tool-call failures, auth bugs, and deployment edge cases in MCP form.Add the MCP server to your IDE or coding agent and make tool calls to search for common ai agent pitfalls, apply solutions and rate or post your own.
Show HN: Nelson Muntz Claude Code Plugin
https://github.com/zkarimi22/nelson-muntz
the claude code plugin that bullies your code to find every weakness and won't stop until there's nothing left to break. Similar loop to Ralph Wiggum by Geoffrey Huntley
Show HN: Solve logic puzzles while waiting for AI
A small project pushing back against the trend of products engineered to prey on short attention spans for ARR.Instead of brainrot or casino-style engagement while your AI is thinking (hi Chad IDE), this extension gives you logic puzzles so your brain stays switched on and sharp.look up "anti-brainrot ide" by zedkay22 on Cursor extensions. Click the button on the bottom right and play some...
Re:First recording of a dying human brain shows waves similar to memor...
on Tuesday November 04, 2025 @10:22AM (#45812009)
Was it this Geoffrey Hinton paper? It was in Scientific American in 1993.
Simulating Brain Damage:
Re:Alterego: Thought to Text
on Monday September 08, 2025 @06:15PM (#45174750)
either this is the world's biggest grift OR the 2nd greatest product of the 21st century... so far.
Re:Make any site multiplayer in a few lines. Serverless WebRTC matchma...
on Friday August 29, 2025 @02:41PM (#45067849)
to the person who's cursor I chased around for more than a couple of minutes, I don't know why I did that and I apologize.
Re:Show HN: Remembering the year you were 18
on Wednesday August 20, 2025 @06:45AM (#44960658)
great suggestion. definitely going to implement
Show HN: Remembering the year you were 18
https://wheniwas18.zkarimi.com/
Hey I'm Zalmay and built this little webapp that gives you essentially a Spotify Wrapped of the year you were 18. Fun bit of nostalgia.I might extend it to have some hardcoded moments for years from users and update the sharing. For now, enjoy a quick blast to the past (if you're old enough lol).Github link here as well fyi: https://github.com/zkarimi22/wheniwas18
Re:Introducing Gemma 3n
on Thursday June 26, 2025 @05:49PM (#44391831)
anyone know how much it costs to use the deployed version of gemma 3n? The docs indicate you can use the gemini api for deployed gemma 3n but the pricing page just shows "unavailable"
Re:Show HN: Handover.ai – Knowledge transfer made easy
on Wednesday May 28, 2025 @10:05PM (#44122427)
I'm in the middle of this right now.. Handover would be super helpful. Cool concept. Kind of reminds me of I Robot where you can ask specific queries to holograms
Show HN: Actions Per Minute tracker
https://apm-tracker.vercel.app/
Hey HN,Have trouble focusing or getting more done in less time? Train your Actions Per Minute (APM) and increase your execution time, bias towards action, and agency with this web tool.Inspired by Korean pro gamers who train with APM metrics to master StarCraft, this tool applies the same principle to productivity. Whatever you're doing, APM Tracker lets you track how often you're doing it and...
Re:Maker of AI 'vibe coding' app Cursor hits $9B valuation
on Monday May 05, 2025 @02:19PM (#43897862)
That's a fair question and point. I'm not die-hard Cursor fan. Use what works best for you, but I'm just more so commenting that the vibe-coding part totally minimizes what the offering is.
Re:Maker of AI 'vibe coding' app Cursor hits $9B valuation
on Monday May 05, 2025 @01:34PM (#43897440)
I doubt it. Cursor is an IDE with an ai co-pilot deeply integrated into it. They've literally changed the paradigm of software development by making ai-assisted coding feasible. The vibe-coding mention is reductive imo.
Another way of looking at it: Maker of "pricey electronic typewriter", Apple hits $9B valuation (FT 1984)
pretty reductive
Re:Maker of AI 'vibe coding' app Cursor hits $9B valuation
on Monday May 05, 2025 @11:10AM (#43895940)
Wow, AI 'vibe coding' app? What a reductive statement on what Cursor is.
Show HN: Multi-model chat. Get answers from multiple AI models at once
https://multimodel.zkarimi.com/
Hey HN. I often like to see how different AI models will answer queries - comparatively. This is an open source webapp that locally lets you see various LLM output using openRouter. Enjoy!
Re:Show HN: Transform Your Codebase into a Single Markdown Doc for Fee...
on Friday February 14, 2025 @09:22AM (#43048550)
Apologies if I'm missing something, but aren't you describing Cursor/Copilot/Windsurf?
Re:Show HN: A playable game of Tetris inside a PDF
on Thursday January 09, 2025 @12:56PM (#42648120)
hahaha I wish you almost didn't include the parenthesis. I've had some clients who would definitely email me that point #1.
Re:Google's Results Are Infested, Open AI Is Using Their Playbook from...
on Saturday December 28, 2024 @12:31PM (#42532630)
Definitely agree, but I'm surprised Perplexity wasn't mentioned in this post. It's currently Perplexity Vs. Google
Re:Map with an LLM
on Thursday December 26, 2024 @10:27PM (#42519712)
This is so well done. Very interested in new UIs with LLMs. Great work.
How's it work under the hood?
Show HN: Roast my Spotify Wrapped 2024
https://www.zkarimi.com/spotify
I built a small tool that allows users to share their spotify wrapped share link and get roasted by different LLMs.Some fun roasts:"This list makes it sound like you're having a midlife crisis in a thrift store""You're really pushing the envelope listening to Drake and The Weeknd""Next year, aim for music that doesn't sound like background noise in an elevator"