Who Built AI — And What Must We Give Back

We love to talk about AI as this next frontier. But make no mistake: today’s AI doesn’t spring from magic. It rides on the foundations laid by countless people — open source libraries, research papers, blog posts, experiments, code samples — freely shared over years. The “raw material” of intelligence is human labor and knowledge made public.

In “Abundant Intelligence,” Sam Altman argues that access to AI may one day be a fundamental human right. That’s not mere rhetoric. It’s a recognition that the commons of human knowledge is being expropriated into closed systems — systems controlled by the few who own the models, the data, the compute. If that’s true, then yes — there’s a moral debt.

Altman doesn’t just speak — he sketches out ambitious infrastructure plans: a “factory” that can produce a gigawatt of AI infrastructure every week. He’s shifting the conversation: AI is no longer just algorithms and code; it’s energy, chips, power grids, and physical scale. Who controls compute and power wields real influence.

If we believe the future should be shared, not hoarded, we need to match his ambitions with reciprocity. “Giving back” should mean:

• Publishing weights, architectures, training methods
• Revenue-sharing or attribution systems for original contributors
• Building public compute capacity — especially in underserved places
• Policy that treats compute, intelligence, and infrastructure as public goods

Because make no mistake: the question is not just who owns the AI — but who owns the infrastructure behind the AI. If intelligence will depend on who controls energy and compute, then democratizing access isn’t optional. It’s essential.