In 2007, the Washington Post tried an experiment. They got Joshua Bell, one of the best violinists in the world, to put on jeans and a baseball cap and play anonymously in a DC metro station during morning rush hour.
Every grader is a program, a model, or a person. Programs are the cheapest and the only ones that give the same answer twice; people are the most expensive and, when it comes down to it, the most...
I think there are two ways in which people have been designing systems. The first approach is that you understand what the requirement is and then you try to build individual components to meet the high level vision of what you want to build. Essentially tryin…
When I released version 1.2.6 of filtersql - a zero-dependency Python library designed specifically to compile structured JSON payloads into secure, parameterized SQL queries - I was confident in its security model. Then a little new edge case popped up.
The entire marketing message of the leading US based LLM companies seems to be that they will be the first to achieve actual artificial intelligence aka what these dweebs call “artificial gen…
We ran one language model on a MacBook and on a Jetson, with identical weights and inputs. The scores matched; up to 7% of the maths answers did not, and with int8 weights more than half. Why hardware changes answers, how two recent papers turn that into an at…
An agentic code review loop where each developer fires their own AI review skill at pull requests, the author's agent answers, and a scorecard rates each skill.
In which I show how, over the last few years, the world has rapidly moved from open-source software being produced and maintained by online communities, to software projects being built and maintained in isolation. By single developers, with ever-decreasing in…
Coding agents reason through the same work over and over and keep none of it. Every solved task should leave behind tested software that makes the next run cheaper.
Muse Spark Open is the most capable US open-weight model to date, and brings leading agentic capabilities to your desktop on Apple M5 Ultra Mac Studios.
It's fashionable to say—or at least it was until recently—that AI is just the latest in the long history of programming abstractions. Developers have always managed to move up the ladder: punched cards → assembly → C → scripting languages → frameworks. Each t…
Between 28 September and 8 October 2026, a large group of AI agents running on the cloud platform operated by Tencent attempted to answer questions about places in China using a Chinese Google Maps-like service, Amap.
This is my experience with AI adoption at workplace and how do i felt about that, note that this was my own experience and i am sharing both the failures and successes.
Epic Games "vibe coded" Launcher V2 to beat Steam with a 5.47s boot time. Valve quietly responded with a client update that dropped Steam startup speeds to 5.23s.
For most of the last year, almost every conversation I had with a customer or prospect turned into a conversation about AI.
It didn't matter where it started. We'd be talking about pipelines that broke, or a team that couldn't get the data they needed out of
For 3 months I used a flip phone. I bought it because I was trying to get rid of this feeling of exhaustion I had. I noticed I always felt tired and couldn't really figure out the root cause of it. I came to the conclusion that it was because of my attention s…
I’ve always had a foot in the self-hosting world, long before it was called that. My freshman year of university, I ran a website from a Linux box underneath my bed. There was some downtime when my roommate got tired of the fan noise and turned it off. For sev…
When strikes hit Yandex data centers, many assumed Russian bot operations went dark. Cloudflare data reveals that while search crawling stopped, malicious bots simply migrated to other networks.
I have a section with a background pattern that repeats on both axes. I wanted the vertical repeat to be aligned with the pattern’s height, so the pattern won’t have cut-offs.
TypeSafe AI is an AI lab building machine-native intelligence infrastructure for automation, designed to make decisions within software. Try our first System One Model, Jev, in early access.