This idea of algorithms having “final destinations” is compelling, but the article (sadly) does little to explain what that technically means—touching only on what simulations are for—and doesn’t expand enough on what that destination currently looks like.
“The thing about predictive simulations is that the model only simulates whatever constraints it is provided. The simulated world only cares about what it’s told to care about. Given the capitalistic motives of big tech, that usually means one thing. Shareholder value.”
Is the simulated end game right now one that finally sucks out /all/ of the money?