Hacker News Reader

Comment by mathattack | Hacker News Reader

Is such planning possible? Decentralized models work because centralized planning doesn’t.

I concur with the aims. Solving pandemics, climate change and nuclear proliferation require a different toolset than what succeeded for reducing global poverty.

View on HN

Centralized planning didn't work. And my assumption has always been that it was the result of a slow-to-respond system. Today we have a way to react to economic signals almost instantaneously. I think it's worth trying again in a smaller way i.e. centralized planning of the meat industry (someone smarter than me can figure out what industry is actually best to do a trial run of modern centralized planning).

You might be interested by what Cockshott has to say.

You don't think meat packing is centralized? How you you think factories do their work? Where do you think subsidies come from? What does centralized mean?

Why would I want a man in Washington to guide Montana in the form of government though?

It is not like decentralized planning always work better. It depends on the task. What works in my opinion is iteration, feedback and experimentation. You don’t merely dogmatically state centralized or decentralized work better. Instead you experiment, iterate and refine. You learn from others.

Can centrally planned iteration and testing work better? I’m not trying to be dogmatic - I’m trying to look for a 3rd option beyond centralized and decentralized.

Sure, research and development in bell labs was pretty good. Doubt we'd get all that stuff from random people without exotic equipment and cash. Of course they made risks of their own money, not other people's money

Reply on news.ycombinator.com