We are at an inflection point. I did something amazing today. Having played around with Muse, Meta’s AI agent over the past couple of days, cautiously probing its capabilities, I asked it to …
Measuring Uncertainty How do you measure “uncertainty”?
That may seem like an odd question. But let’s just dive right into it, because starting down this path of inquiry will lead us step by step to the definition of the fascinating concept of Shannon entropy.…
I recently had an experience while doing a simple Google search that was so profoundly weird that it stopped me in my tracks.
Understanding this Google se...
Claude spent 119,000 tokens to produce a 107-token program that didn't work, in a programming language that nobody uses? Or that the code you make with Claude is about 2% of your token spend? If so, read this.
As of this writing, the appeals court has upheld the Pentagon’s labeling of Anthropic as a supply chain risk. And while there is always an element of politics involved, it’s worth considering what a supply chain risk is from NIST:
SFisms defines “meat proxy” as:
A person who forwards AI-generated text, code, or other output without reading, understanding, or validating it. The person acts only as a relay between the AI system and the intended recipient.
Being a meat-proxy is to knowled…
It’s common to compare the current AI takeover of software engineering to the rise of AI in chess. Chess AIs went from much weaker than serious players to much…
Large-scale agentic training and evaluation with large language models (LLMs) rely on isolated, stateful execution environments in which models inspect repositories, invoke tools, execute commands, and interact with task-specific services. These workloads crea…
Four changes to the n-gram caches of llama.cpp make drafting up to 41.6x faster, load the static cache up to 23.5x faster, and lower peak memory up to 2.65x.
Since high school, my thoughts have often turned to death, a subject that has remained ever-present in my mind. What has changed is that I have moved from wanting to kill myself, to viewing death as a minor inconvenience.
For the folks who found this blog from Anthropic’s site, welcome! For everyone else, I’d better give some context. Last month, I had a blog post titled “It Only Counts When AI Get…
On October 23rd every iPhone app becomes a foldable app. What earlier transitions and Android's foldables teach, and what the iPhone Duo simulator did to my Big Sky Field Guide.
Drum Machines (LLMs) have no soul -- Like so many others I’ve been thinking about the future of my career in software. These LLMs have changed everything. What’s actually left for me? What’s my defensible moat that won’t be crossed before I’ve even finished pl…
In the near future, an engineer looks back on the rise and quiet disappointment of machine intelligence. After years spent waiting for the future to arrive, it's time for him to turn off the lights.