How we forecast the future impacts of new technologies such as AI, and how we think about our leverage over them, in part depends on which historical tech cases we let in as evidence.

People often wave away the numerous examples of technological divergence in premodern societies. But in my new essay, I argue that none of the standard criteria offered to distinguish a "true" technological civilization really hold up well.

Any line you might draw — 'vibes', enduring (monumental) art, human welfare, defensive resilience, or seed-contribution to the Industrial Revolution — is leaky in both directions, smuggles in our cultural bias or just begs the question.

The modern innovation landscape certainly differs from that of the past; but only in degree, not kind. Instead of falling for every 'no true Scotland' argument, I propose we rule technological civilizations in, not out, when trying to understand the texture and reach of the tech tree.