60articles · curated from global AI & Data Engineering sources
How nb2lite-skill-agy packages Google's gemini-3.1-flash-lite-image as an Antigravity skill + MCP server — with multi-turn stateful edits, a simple install guide, and a dogfooded cover image.
In March, the CEO of a research lab with zero products closed a $1.03 billion seed round — the...
Hetzner is experimenting with LLM inference. That is not a sentence I expected to write, but I think...
For quite a while now, I've had the feeling that AI agents are surrounded by this mystical aura....
Most production AI tasks are not LLM tasks. I replaced a 7B model with a tiny 2.4 MB Go classifier: rules first, small model, LLM last.
Those who have been following me for a while will know that teaching and mentoring are a Big Deal™️...
3.6...
I've spent my career helping people build software. This is a series of letters about what happens...
The first time I bolted AI onto a data warehouse, the plan looked reasonable on a
What it took to deploy, why the QAT checkpoints refuse to load, and what one flex-start v6e chip is actually worth — measured live.
In the last entry I got Gemma-4's 128-expert MoE running on an inf2.24xlarge and signed off with...
Imagine your AI assistant just produced 200 lines of code. Legally, you may not own a single line of...
Running an agent on your own hardware fixes data sovereignty and nothing else. Prompt injection, silent provenance failures, and privilege escalation all survive the move to local. Here's where local agents are genuinely safe to deploy in 2026, and where 'local' is just a comforting word.
I trained a model that scored 4 percentage points better than the one I ship. It's gone. I deleted it...
Fit calculators check memory. Leaderboards check general skill. Neither checks your task. Measured results for local quants on two machines, reproducible from raw outputs.
Scope note (read first): this describes a system I built and operate to develop an unannounced game....
Scope note (read first): this describes a system I built and operate to develop an unannounced game....
In the previous article I wrote: "Mechanizability-scanner. Building this next." And: "Cross-model...
Porting Gemma-4 31B (dense) to AWS Inferentia2: the tensor-parallel recipe that worked at 12B collapses at 31B, NxD ModelBuilder saves it — and then a passing validation lies to your face.
This article is my submission for the Agents of SigNoz Hackathon: Blog Track, where participants...
Part 1 of Retrieval-Augmented Self-Recall — the research track behind Claude Code, Beyond the Prompt....
I've been hearing the word "harness" thrown around a lot lately. I assumed it just meant "the IDE" or...
A bug showed up in my personal project last month. Nothing dramatic - a value wasn't updating the way...
My laptop was sitting idle with the fan at full tilt. Nothing was running that I knew of. The culprit...
A payment landed for exactly half an invoice's value. The payer's email matched the customer on file....
A stranger on DEV.to said "run this experiment." I ran it at n=600. Here's what happened — including...
My memory gate passed 16 out of 16 frozen cases. Then I blocked the article. Not because the run...
Cloud AI is a subscription trap. I built an exit that fits in my jacket pocket and runs at 3...
I just got back from my vacation which, like 99% of Polish people, I spent in Croatia. Dobar dan!...
Two years ago I open-sourced KeyEcho, a small desktop app that plays a mechanical-keyboard sound the...
When I first got involved in improving AI models for our IoT devices spread across different parts of...
I Tested 300+ Models. Then I Killed the Benchmark. Let It Break — part 1 Tags: #ai #llm...
My gut screamed scam. My tool said 'I don't know' — and that refusal is what actually cracked the case. An experiment in why honest CLIs beat confident ones.
Yes-Brainer is a council of AI models for the decisions that aren't no-brainers. One question fans...
There is a growing sentiment in engineering circles right now that documentation is a relic of the...
antirez.com
A field report on running Google's Gemma-4 on AWS Inferentia2: mixed attention heads, the vLLM / optimum-neuron / NxD dead-ends, and the neuronx-cc compiler limits.
Thanks to Dipankar Sarkar, Mike Czerwinski, Max Quimby, and Ponsubash Raj R for the detailed comments...
One thing kept bothering me about retrieval systems: they often destroy the document before...
Playing the game of writing technical post was funny two years ago. Now, I am a mod and it is...
Previous Related Post: Part 1 This particular rabbit hole was created due to a previous...
File-system checks ask "did the script run?" Neural gates ask "did the constraint actually change the...
PR volume went up, ticket quality didn't, and the gap got filled with LLMs on both sides of the review: bots reviewing, bots replying, bots occasionally arguing with bots about priorities that only existed in a teammate's head. Our CEO named the actual problem, and it's bigger than code review.
Every "agent quality gate" I tested shares one fatal assumption: that an LLM can judge whether an LLM...
Quick note before we dive in — I know I've been off track from the iOS/Swift series lately. I just...
A post on X last week: "The fact that people can't even reply to posts without AI anymore says a...
RAG retrieval strategies without a vector database: BM25, keyword indices, knowledge-in-bundle — and when embeddings earn their cost.
Intro AI Avatar is a completely free app that lets your VRoid (VRM) 3D avatar animate in...
Reading Lilian Weng's harness engineering survey as a reliability engineer — what self-improving harness papers actually show, and the three invariants every working loop converges on.
We stopped measuring retrieval quality by how many tokens we could fit into the prompt. When...
If you are building applications on top of foundational models like OpenAI or Anthropic, you already...
Frontier models confidently cite prominent papers that were retracted after their training cutoff, with no idea. I measured it: they flag the famous old retractions they learned, and miss the recent ones entirely. A registry lookup is the only fix, and no smarter model changes that.
I didn't go to a university for computer science. I have a B.Tech in Geophysics. What I know about...
How we bolted opt-in Apple Metal and NVIDIA CUDA backends onto a pure-C Qwen3-TTS engine — resident fused pipelines, server request-batching, and measuring it all on a Mac mini M2 rented by the hour. Plus the two 'obvious' optimizations we killed with data.
Take control of your AI models with our newest skill, designed to make local fine-tuning efficient.
The demo ran 50 times without a failure. Then we shipped it. Three days into production, the agent...
Four provenance-tracking arms, identical gates, one uncompacted oracle: measuring exactly what memory compaction does to agent gate decisions — including the preregistered hypothesis that failed.
If someone compromised one of your project's dependencies today, would they be able to steal your...
Most LLM API error handling starts out looking like normal HTTP error handling. If it is a 429,...
I built a 36-pattern checklist to catch AI writing tells in my own drafts, calibrated against...