3 ms·
Use our super computers to allocate resources appropriately? I'm not sure what you are getting at. Marx's critique of capitalism is since a few rich people ow
by yters 6y ago
Use our super computers to allocate resources appropriately? I'm not sure what you are getting at.
Marx's critique of capitalism is since a few rich people owned all the means of production, they can oppress the poor, and neglect to give them what is needed for normal life. The promise of communism is that by taking the means of production out of the hands of a few greedy individuals and somehow running the mechanisms with an objective metric to ensure all people's needs are met we will end up with a society where the poor are not oppressed by the rich, and all can live lives free from things like starvation and lack of shelter or medical care.
Of course, when the communist revolutions took the means of production out of the hands of capitalists, the new hands were no more noble, and continued to oppress with even greater brutality.
One might hope then for a communism run by objective, malice free computers by encoding human needs as a goal function for algorithmic optimizers. Within a materialist paradigm this is at least theoretically possible.
- chordalkeyboard 6y ago> Use our super computers to allocate resources appropriately? I'm not sure what you are getting at. How do you translate the results of the model into a plan for allocating resources appropriately? You're either going to neglect consumer preference because you "know better" in which case its not clear what you're optimizing in your model; or you're going to have a high-fidelity simulation of what the consumers do when they bid on products in the market anyway. > One might hope then for a communism run by objective, malice free computers by encoding human needs as a goal function for algorithmic optimizers. Within a materialist paradigm this is at least theoretically possible. My perspective is that, to the extent that you are attempting to encode human needs without inserting your own biases and values, you would be simulating what humans would do in a free market. We can establish the existence of edge cases like humans falling victim to cognitive biases and alter our model so that we only simulate "very rational, well informed" humans; but as soon as we do that we open the door to give our own biases and values higher precedence than the desires and values of the humans we are simulating. Also its not clear how the output of a machine would be used to allocate resources in an economy. Whether someone prefers one option to another depends on the relative cost to themselves, whether someone wants to sell one option or another depends on the relative profit. The market essentially consists of a vast system of PDE's with hidden parameters that are solved by humans iteratively in distributed fashion. You can tell Joe Sixpack that after the 7th meal, he exclusively prefers hot dogs to hamburgers unless the price of beef drops below 1.05x the price of hot dogs, but not before, and when the price of beef reaches 2x the price of hot dogs he prefers beef again but for conspicuous consumption and not for the taste. But its not clear how you would discover that before he discovered it himself, and even if true its not clear that he would prefer hot dogs to hamburgers before he went through the 7 meal process of discovering that himself.
- yters 6y agoNo need to simulate people. Just collect their preferences through an online UI, like grocery shopping.
- chordalkeyboard 6y agostated preference is frequently different from demonstrated preference. if you fail to simulate people then you will fail to produce an accurate representation of their desired outcome.