The Paperclip Problem in AI

The way that I think about it is that we're familiar with the paperclip problem in AI. If you had an artificial general intelligence that became sufficiently powerful to start scaling up its own intelligence, and it had been designed to optimize the function of paperclip factories, if its central prerogative was more paperclips, it would literally consume the world to create more paperclips. And so everything that contributes to the production of capital, the system wants to be as fungible as possible. The problem with that is that human beings are the biggest product of capital. But something like the breakdown of the family makes sense from capital's perspective because if you have the best comparative advantage available to you living

Transcript

Play full episode

Transcript

Episode notes

The AI-powered Podcast Player

Save insights by tapping your headphones, chat with episodes, discover the best highlights - and more!

Get the app