The Token-Burn Trap: Silicon Valley's AI Metric Is a Competitive Liability
Token spend is not a productivity metric. It's an effort metric. And effort is not outcome.
Token spend is not a productivity metric. It's an effort metric. And effort is not outcome.
The “old” Internet isn’t gone, our expectations have changed.
Surplus AI coding tokens are wasted capacity. Use them for targeted code reviews, UI audits, product analysis, and marketing - with copy-paste prompts for each.
Dmitri's blog about programming, Clojure, and software development
AI coding agents have made code production nearly free. Drawing on insights from Kent Beck, Paul Ford, and Simon Willison, this post argues that the value has shifted from writing code to defining systems — contracts, invariants, SLAs, and verification.
On April 8, 2026, I had a conversation with Claude (Anthropic) that went from a loose intuition about dark matter to a testable astrophysics prediction in peer-reviewable form, in one sitting. From my phone.
The deadliest thing of all isn't to do bad things, it's to do nothing at all. If you do wrong you can always learn from it and redirect yourself, but if you do nothing you won't ever get to that point. In this reality, you can either do things or not do things…
and why you should still read the code
Sometimes, I find myself talking while my audience has a puzzled look on their face. It doesn't matter how much I prepared my speech, the message is just not getting through. But then they ask: What a
Exactly a year ago today, I changed something about myself that had been part of my life for ten years. What that "thing" is won’t be mentioned here, so a...
Ways to stabilize benchmark results, to reduce noise.
🧵 Most People Overcomplicate This System Design Question
What if, instead of speeding everything up, AI let us build intentional businesses? We're all so busy arguing about whether AI will replace programmers, kill creative jobs, or consolidate power into fewer and fewer companies. I think there might be a quiet…
Keeping code clean helps both humans and agents. Coding agents are limited by context, and poorly organized code means they consume more tokens and perform worse. Here are some practical steps to keep your code clean for both humans and agents.
We built a synthetic clinical dataset with known ground-truth patterns and fed it to 18 automated data analysis tools. Most of them confidently reported incorrect results.
Scroll inside the stage with the mouse wheel or with your finger. It is only a machine, but it already understands the gesture. Have fun! …
I was clearing my desktop recently when I came across a graphic I made couple of years ago to describe people in my life to another friend. How I interpret different people in my life. Looking back at it couple of years later I found it interesting, looking ba…
The computer system aboard the current Artemis II lunar space mission is from a different world that the one from the Apollo era. Apollo astronauts navigated to the lunar surface using a computer with a 1-MHz processor and roughly 4 kilobytes of erasable memor…
After almost twenty years on the platform, EFF is logging off of X. This isn’t a decision we made lightly, but it might be overdue.
Thunderbird is a free email application that’s easy to set up and customize - and it’s loaded with great features!
Claude sometimes sends messages to itself and then thinks those messages come from the user. This is categorically distinct from hallucinations or missing permissions.
Insights and guidance from our engineering team on how Astral secures its tools.
The ag manufacturing giant will also make digital diagnostic, maintenance, and repair tools available to third parties for 10 years.
A basic introduction to USB for people that don't need to know what happens on the wire
Reverse engineering the traffic system from Pizza Tycoon (1994): how the original devs drove dozens of cars through a city with almost no CPU budget.
Discover powerful applications such as Little Snitch Mini, Little Snitch, LaunchBar and Micro Snitch.
Easy and intuitive Kalman Filter tutorial
Live status of the Strait of Hormuz
In early March, I noticed approximately $180 in unexpected charges to my Anthropic account. I’m a Claude Max subscriber, and between March 3-5, I received 16 separate “Extra Usage” invoices ranging from $10-$13 each, all in quick succession of one another. How…
Introducing Muse Spark: Scaling Towards Personal Superintelligence
Today, we’re excited to introduce Muse Spark, the first in the Muse family of models developed by Meta Superintelligence Labs. Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration…