The bottleneck is the supervision
The [AI] models are already powerful enough to produce publishable results under competent supervision.
The [AI] models are already powerful enough to produce publishable results under competent supervision.
The way a smart assistant works is, it processes information ahead of you, trims it down, and presents you with a couple of options to decide on.
A simple Python script and xbar setup to monitor Claude Code token usage directly in your macOS toolbar.
Prevent Claude Code from burning tokens aimlessly. Slow things down with a simple shell trick.
Our desire for optimization and productivity drives us to abdicate effort and skill in exchange for rapidity
But it totally could, if one manages to put these few points together.
Artificial text is a statistical mashup of human quotes. When we quote AI, we quote quotes.
Machine learning rocks, but it’s far from intelligence. It’s advanced stats on steroids. Know its limits, roll with smaller models, and be the boss of your decisions.
First Encounter is a wallpaper I created in collaboration with Stable Diffusion’s AI model. When working with generational AI, it is hard to say who should be credited for the creation of the work. Was it me who came up with the idea and thought of an appropriate prompt, or was it the model that did all the hard computational work?
The algorithms that are currently being used by the social media platforms that dominate our attention are designed to keep us engaged with content for as long as possible. But that’s not all they can do.
Stay connected — pick your platform: