# Thinking in Silico: How Goodfire Is Decoding Concept Geometry to Re-Engineer Frontier AI > Source: > Published: 2026-09-09 01:32:46+00:00 Goodfire CTO Dan Balsam reveals how mechanistic interpretability is evolving from passive observation into active control of large language models. Through concept manifolds, predictive data debugging, and the Silico research agent, Goodfire is attempting to turn high-dimensional AI black boxes into inspectable, steerable substrates.