Whatever you do, don't lose the plot
The other day, I was working on translating data from a file into json to visualize the results. Easy enough, given that this was an operation I’d done countless times in the past. Still, I blanked.…
Re: People Are Not Friction
Dave Rupert puts words to the feeling in the air: the unspoken promise of AI is that you can automate away all the tasks and people who stand in your way. Sometimes I feel like there’s a palpable…
Decafclaw: Building an AI Agent in a Week
TL;DR: I set out to spend an afternoon understanding OpenClaw. A week later I'd built my own agent with 80+ tools, a web UI, and self-reflection. Here's how that happened.
Tabstack + OpenClaw sitting in a tree
TL;DR: I built a skill to hook OpenClaw up to Tabstack - and it seems like a pretty great upgrade to the web fetch tool it comes with out of the box. Oh yeah, and OpenClaw is not nearly as scary as…
AI-Augmented CI/CD: Shift Left Security Without Risk
AI code review in CI/CD without prompt injection. Defensive patterns: three security tiers, isolated execution, no secrets in prompts.
What is an AI Skill?
A skill can be just plain text written conversationally, it can be a list of rules, mathematical expressions, or even the foundational code that you want your AI to readily modify and apply.……
Something I Can Never Have
"AI isn't lightening workloads," The Wall Street Journal reported last week. "It's making them more intense."Well, yes. This is, in fact, how "workload" works…
Occasional Weeknote #25
The 25th weeknote is a bit short and uneventful. No real reason other than I’m just in a bit of a creative funk for the moment and wading through some general Life Overwhelm (which I feel like…
This example starts with a chi-square but ends with a lesson on how even well-written prompts can result in hallucinations.
A research study counted how often ChatGPT made up citations for three different categories of mental disorders (binge eating, body dysmorphic, and major depressive). They used a chi-square to…
Does Computer Science Still Exist?
I have always seen the computer as a tool that gives us new ways to think. Does it still? As I write this, Hack, the text-based dungeon crawler from my childhood, has just been fully ported to…
10 tips for using Claude Cowork
The most powerful way to use AI is as a collaborative mind you design intentional systems around starting with a clear sense of what you're optimizing for. The post 10 tips for using Claude Cowork…
Claude Code - Programmer's Partner and Torah Scholar?
With Passover fast approaching, I've been putting AI to use in my Seder prep. Specifically, I've been using it to discover and unpack different topics of conversation I could explore with my guests.…
Using Autoresearch Project to Build The Fastestest Java Decompiler
Most people first see this Andrej's Autoresearch project as an ML autotuning setup: an agent edits one file ( train.py ), runs short experiments, and keeps only measurable improvements. Under the…
Unlock Your NotebookLM Data with Gemini 3.1 Pro for Next-Level AI Analysis
NotebookLM in Gemini is the money. Here’s what I mean. Google’s Gemini 3.1 is a crazy smart model, especially the pro version. But even with a 1 million token context window (short term…
No thank you i will not change how i write so you don't think i am an llm
In the instructions for the final essay in a module i took last year, i was warned that the essays would be checked for ai plagiarism and that we should take the initiative to run our own essays…
The Future of SaaS Is Agentic
TL;DR: The future of SaaS is agentic, but agentic SaaS is not just a chatbot layered on top of APIs and a dashboard. Traditional SaaS was built for users to operate software manually; agentic SaaS…
Claude Cowork Guide: 50+ Tested Tips on Plugins, Skills, Sub-Agents, and Memory
Karo Zieminski spent nine days breaking Claude Cowork before writing this guide: I've seen enough of shallow tutorials that simply rephrase the official docs to know I wanted to do something…
AI News – March 20, 2026
Quick Brief Featured story today: “Ooh, new drama just dropped 👀” And more models & releases news from the AI universe – Photo by Markus Kammermann on Unsplash. Today’s Hot Posts…
Bots Will Outnumber Humans Online by 2027
Cloudflare CEO Matthew Prince just predicted that AI bot traffic will exceed human traffic online by 2027. His company powers approximately 20% of all websites and is used by approximately 80% of…
How to fail at healthcare, education, and technology in one week
Sometimes it feels like governments are actively trying to make everything worse. “Affordable” healthcare that no one can afford. AI regulations designed to ensure 450 million Europeans…
Get AI-Ready With Erik: Retrieval Augmented Generation
Get AI-Ready With Erik: Retrieval Augmented Generation Summary In this video, I delve into Retrieval Augmented Generation (RAG) and its application in SQL Server databases, particularly focusing on…
Will the last human please turn out the lights?
I have a reputation as an AI apologist. I doubt this post will change anyone’s mind. But I have to address the recent news that Hachette canceled the US release of a book called Shy Girl and…
Amazon's Novel Approach to a Newfangled AI Device: a Phone?
It's perhaps not as crazy as it sounds. Just don't say "Fire Phone"...
Designing the Modern Data Center Network for AI Workloads
The fastest GPU is only as fast as the slowest packet. For years, data center design has been driven by compute density, virtualization efficiency, and east-west traffic patterns dominated by many…
Why Is Everyone Supposed to Die If Machines Can Think?
If you only listen to spokespersons for AI companies, you'll have a skewed view of how AI is actually being integrated into the workplace. You probably don't need to convince a developer to include…
Stop Hiring Cooks. Start Building Kitchens.
Everyone says "we can use AI for that." But reliable, repeatable AI requires infrastructure. Here's why developers need to build the kitchen, not just hire the cook.
Stop Prompting, Start Briefing
For the past year I have been teaching technical teams a five-step system for getting better results from AI. I call it the Prompt Cycle. Teams that adopt it see immediate improvement because they…
When AI Goes Rogue and Other Security Wake-up Calls
Amazona and Meta face AI security crisis, OpenAI acquires Astral for Python development, quantum computers get $5M healthcare challenge, and Pentagon plans classified AI training environments.
Are you late to AI?
You're not too late for AI. This encouraging read reframes the AI boom, helps you overcome hesitation, and invites you to start exploring today.
EnshittifAIcation
Yesterday morning, first thing after waking up, I checked my emails. One of them was from a client - a sharp person, but not a tech expert - forwarding a message from one of their "digital…
Agents (or Humans) in Goal-Directed and Goalless Environments: On Pipelines, Priors, and the Rhythm Between Exploration and Exploitation
“It is not knowledge, but the act of learning, not possession but the act of getting there, which grants the greatest enjoyment.” – Gauss (letter to Bolyai, 1808) Give an AI Agent a…
I’ve had Mark’s email sitting in my inbox all week and I just keep on reading this. I’m posting it h...
I’ve had Mark’s email sitting in my inbox all week and I just keep on reading this. I’m posting it here so I can keep on coming back to it in the future: Hot take: AI is going to be better at helping…
Teaching a Transformer to Read DNA: How EabhaSeq Finally Works
Months of failed tests, a breakthrough in reference-conditioned generation, and what happened when synthetic training met real T21 samples.
What We Could Be Doing Instead of AI
There’s been a lot of money spent on AI. I think AI is a mechanical turk. I think there are a few better uses of the investment. Some boring things are investing in urban density, universal health…
Review your own AI-generated code
Years ago, when we started to have 5+ contributors to our codebase at Lyft, we had to establish a code review process with high enough value to justify the friction it caused. However, "value" is…
AI Vision Is a Game Mechanic Now
There’s a category of game that couldn’t have shipped two years ago. Not because the hardware didn’t exist, or the game design theory wasn’t there, or the players weren’t ready. The core mechanic was…
[Veille] Fortunes et impasses de la critique de l’IA
https://danslesalgorithmes.net/2026/03/17/quitgpt-1-4-de-la-politisation-du-boycott/ Hubert Guillaud livre un dossier en quatre parties sur les critiques de l’IA (où « IA » ne désigne pas le domaine…
Does Bad Code in Your Context Window Misalign LLM Recommendations?
Put away your Betteridge’s Law of Headlines, I genuinely don’t know the answer :) In “Training large language models on narrow tasks can lead to broad misalignment”, Betley et al found that…
Tokens Are Now a Recruiting Tool
Jensen Huang, CEO of Nvidia, said that tokens are now a recruiting tool (and he wants to give engineers half their salary in tokens)....
The right way to be right
Plus: How and why AI might kill us all. (#567)
F2F #79: When will taste die?
Taste won’t die in the AI age - it just changes jobs. As creation gets cheaper and AI slop floods every feed, the scarce skill shifts from writing code or content to knowing what to make, what to…
The Abstraction Layer
Are we entering a world where code is another abstraction layer? And, how closely will the next wave of engineers pay attention to it?
The Feed Framed the Last Election and AI Answers Will Frame the Next
After 2024, Democrats drew the right immediate lesson: get better at being online. More native creators. More distributed media. More willingness to compete in the places where politics, culture, and…
Fun with Claude
Like everyone, I’m busy developing expertise in harnessing AI, vibe coding in my spare time. I have a couple of product ideas I’d like to wrangle, but they’re going to take some…
Why Building CLIs Is One of the Best Practices in Vibe Coding
If you've spent any time building tools for your own workflow, you've probably discovered it: there's something deeply satisfying about crafting a good command-line interface tool that fits exactly…
Fanout Query Analysis
When AI models like Gemini, GPT or Nova answer a question using web search, they don’t just run your query as-is. They generate their own internal search queries, or fanout queries. A single…
Experiences with AI Tools
After resisting it for a long time, I decided to plunge into using agentic AI tools to help me with various tasks. Here is a synopsis of where we are at in late March, 2026: Task: Update python code…
The Post-SOLID Era: When Code Becomes FLUID
Last year I explored the idea of FLUID software as a design philosophy. This piece explores the next layer of that shift: how AI changes the software delivery model itself.For decades, software…
Cognitive Helmets for the AI Bicycle Part 2: The Sometimes-Wrong Bot
The Sometimes-Wrong Bot When I was a postdoc, reflective of the inability of the university to contain my multitudes, I was cross-appointed in three disciplines (Cognitive Science, Computer Science…