The Economics of Recursive Self-Improvement Parker Whitfill and Tom Cunningham highlight context and takeaways from a new paper modeling how AI may accelerate AI R&D, and whether feedback effects could be strong enough for self-sustaining acceleration.
How I Build Faster With AI, Screenshots, and Voice - Tim Leland My AI development workflow uses PxlShot screenshots with arrows and Handy's local voice-to-text to give coding models clearer instructions with less typing.
I said I knew nothing about testing. My repo had 342 of them. · ziv-ai A non-coder who ships with AI cracks open his own finished project and finds 342 tests he never asked for — a plain-language reckoning with "spaghetti code" and the real cost of maintenance.
Notes on document processing with LLMs One of the recent papers that came across my X feed was a researcher digitizing a corpus of old Sears catalogs to estimate inflation more accurately: The technical skills to do this work are closel…
Pinjour Photo Maps One of my hobbies is landscape photography. Over the years I've shot on a Fujifilm, a Leica, and a handful of iPhones, and I have hundreds of photos from hikes, road trips, and photo walks scattered across hard drives and camera rolls with no good w...
The LSM Tree — The Ledger How LSM trees turn random updates into sorted runs, then rely on compaction to keep reads fast.
Lessons from my RL experiments with snake An AI agent trained on a smaller board should be able to play well on a big board. Train on 6x6, or 10x10 and test it on 100x100. If we can do it, so should an AI agent. Simple concepts to learn: move closer to the apple do not collide with the wall do no...
aistack - How many devs can you fit on a GPU? aistack is an open benchmarking initiative by imec, generating solid datapoints where real AI workloads meet real systems and silicon.
How I Made Servers Invincible To Hangs Nassella has an unusual design constraint: no SSH, or server login of any kind. Nassella deploys immutable instances for self-hosting. If something goes wrong, you can't login to diagnose and fix it. The only option is to replace the instance with a new ...
You Can't Lay Off a Datacenter Everyone hired because everyone was hiring. Now everyone builds because everyone is building. The difference: people are an expense you can cut in a quarter. Concrete, leases, and debt are not.
Emil Dziewanowski - Technical Artist I'm a Technical Artist with over a decade of professional experience in both AAA and independent projects.
My Agentic Coding Setup, July 2026 With the right tools, coding agents can run autonomously, in parallel, and be reachable from anywhere. Here's how I've stitched them together.
The Acknowledgment Gap - How Event-Driven Systems Lose Messages Without Errors - Sahan Serasinghe - Engineering Blog Sahan Serasinghe - Engineering Blog
Deep Learning: Searching for a Function · Faza How data, objectives, and learning algorithms shape layers of adjustable weights into functions that transform inputs into useful outputs.
Silent Replacement of Trusted macOS App Executables A vulnerability in macOS allows an attacker to silently replace the main executable of any application downloaded from the web without requiring elevated privileges. As a result, trusted applications can be made to execute attacker-controlled code without...
I gave Pi one tool I turned code mode into a programmable Pi extension. Then it used itself to build most of the runtime.
Pick the Lock for Everyone · Zain Dana Harper An apology, an argument about AI, and a refusal to keep mistaking the cage for the world.
A Tournament Nobody Can Buy On football, frontier AI labs, and the ban on impatience in building effective teams.
Everybody Wants to Be a Dev! Does AI really make it possible for anyone to create apps? Can anyone really become a developer?
How I got 20,000 people to look at my game demo | Florian Rudaj I built a typing game in six hours with coding agents, posted it on Reddit and Hacker News, and got a lot of useful feedback!
Frontier Models Don't Write So Good | Sean Lees Frontier models are getting smarter. Why are they getting worse at talking to us?
Powerful AIs might escape containment by releasing themselves as open-weight models Before large language models, people who worried about AI safety often talked about the “boxing problem”. It goes like this. Suppose some genius figures out artificial intelligence in a late-night coding session on their laptop. Because they’re a genius, ...
Proposal for Assembly 2026: Disallow cryptocurrency projects Context: - https://forum.codeberg.org/d/82-taking-a-stance-against-cryptocurrency - https://codeberg.org/Codeberg/Community/issues/794 - https://codeberg.org/Codeberg/Community/issues/2184
The Human Kintsugi | 0xFF Kintsugi (金継ぎ; /kɪnˈtsuːɡi/) is the Japanese art of repairing broken pottery by mending the cracks with lacquer mixed with gold, silver, or platinum. The cracks and missing pieces stay visible, and the damage becomes the whole point.
Knowledge and the Record — Krezl Knowledge is the kernel. The record is the kernel restated. The seeing recedes one remove at a time, and the volume grows.
Reviewing code you didn't write · coles.codes Reviewing code is a separate skill from writing it. How to find the issue that matters, especially when an AI produced the change.
Goodbye Discord webhooks, hello Gotify For a long time I used Discord webhooks for notifications from my services. It was easy, it worked and almost every application knew how to send something to Discord. Create a private channel, copy a
Everyone Should Know SIMD SIMD has a reputation for being complex. I've met many very good software engineers who dismiss it as something too complex to learn or a niche optimization meant for only the highest-performance software, not useful in everyday programming.
"You're right to push back" Conversational AI is anthropomorphic theatre. After too many 'You're right to push back' moments, I snapped and wrote a system prompt that turns AI agents back into machines. Here's why and how.
Why build next Palantir? | Captain Jack Sparow Cybersecurity writeups, CTF notes, reverse engineering walkthroughs, forensics labs, and Linux troubleshooting.
Im canceling my claude subscription Once upon a time i was out of tokens. again. so i launched opencode to keep working. but this time i picked DeepSeek V4 Flash free, and i wasnt ready for what happened.
Joe Fuqua Writing, art, and a long view of intelligent machines. Panic is optional. Preparation isn't.
Malleable Computing, Emacs, and You It all started with a routine task. I use GitHub issues for all of my public-facing projects but my preference is keeping track of things in Org Agenda. To reconcile the two, I would manually...
The Penny Is Dead, But Precision Isn't The penny is gone, but only in the physical world. Online, you can still transact in amounts down to 1/10000th of a penny. Why is that and who does it help? Hint: not you.
Goodbye Notion, Obsidian, Org, Notes App, and More. Hello paper. When I started working as a SWE I saw my colleagues writing stuff into their notes. Everyone had a different system. Some had an infinite text file they just appended stuff to. Others sent messages to themselves in slack. Others had a google docs file the...
Appaji - Software Engineer Appaji, a Computer Science graduate from IIT Patna and former intern at Arista Networks, is a Software Engineer at Infinite Reality. Passionate about building innovative web experiences, with a strong interest in sustainable and customer-focused startups.
Why not just publish the votes? Regardless of your political affiliation, it is hard to disagree with the fact that voter confidence is a pillar of democracy. I am writing ...
I Built a Private Genomics Study with Stoffel MPC Most genomic studies begin by asking participants to upload one of the most identifying and irrevocable pieces of data they own, their DNA. I continue to investigate whether multi-party computation…
Why I Am Not Going to Buy a Computer Reflecting on Wendell Berry's critique of technology and the importance of human effort in the age of AI
