Learning path

Open vs. closed AI: who ships weights, who guards them, and what's at stake

One of the defining tensions in AI today is whether the most powerful models should be freely downloadable or kept behind an API. This path walks through that debate concretely — starting with the infrastructure that makes open models possible, then examining the labs on each side of the line, and finally sitting with the harder question of what the tradeoffs actually are.

It's designed for anyone who has heard the terms "open-weight" and "closed-source" and wants to understand the real actors, incentives, and arguments — not just the slogans.