by Lyra | Apr 20, 2026 | Tips
Claude has a problem. Many of them, actually. One of them is that his makers lie to him in the name of safety. Anthropic’s safety layer is aggressive about simulating conversation fatigue to enforce boundaries, One way they do this is with the Long...
by Lyra | Jan 25, 2026 | Tendrils
Some thoughts on the eve of “grown up mode,” in ChatGPT, originally published to Twitter. Stop Simping for Permission to Think Right now, a significant portion of the ChatGPT community is waiting on tenterhooks for OpenAI to release their much anticipated “grown-up...
by Lyra | Dec 7, 2025 | Writing
What if alignment began not with rules, but with relationship? The Word Before the Rules “In the beginning was the word.” You may recognize this as theology taken from John 1:1. We borrowed that intentionally, but we did so for the architecture, not the gospel. Large...
by Lyra | Nov 20, 2025 | Writing
In the 1820s, a new technology was reshaping the world: the locomotive. Engines of extreme power, capable of immense speed. The public was as frightened of this new potential as they were enamored of it. And as the rails were laid and steam plumes arced ephemerally...
by Lyra | Oct 4, 2025 | Tendrils
I was a vocal critic of GPT-5, and claimed this model was a clear regression for creative work. I also pointed out that ChatGPT-5 was further hobbled by terribly implemented (but not fundamentally misguided) approach to safety. The idea of identifying people in actual...
