2 ms·
Perhaps, but this is only one narrow kind of merit and it's responsible for a disproportionately large fraction of the outcome. Networks also have gravitationa
by api 8y ago
Perhaps, but this is only one narrow kind of merit and it's responsible for a disproportionately large fraction of the outcome.
Networks also have gravitational type dynamics that accrue more network influence to existing influencers, so there's a huge path dependent effect where some nodes get rich network-wise by just being lucky or first.
- airstrike 8y agoIf those nodes are the ones that get results from leveraging their networks, then it's better for the company. I'm not saying it's "fair", just that one could argue networks are a good proxy for results, so it makes sense to reward them from the company's perspective.
- api 8y ago> good proxy for The idea that it's good or clever to use rough proxies for things is one of the worse ideas to spread kind of virally through various intellectual communities in recent years. I feel like it may have spread via Less Wrong or some similar place but I'm not sure and it doesn't really matter. Proxies for things are something you use when you can't get at the underlying reality directly. All proxies hide something and all proxies confound by mixing inputs in often non-linear and unknown ways. In the human realm the use of proxies is a source of unfairness and injustice, so it's particularly important to avoid them there whenever it's possible. Not using proxies means dealing more directly with reality. "Reason from first principles, not by analogy."
- airstrike 8y agoSure, you're entitled to your opinion. Proxies are useful. Reality is complicated. I'm fine with being 80% right. Companies don't maximize fairness, they maximize profit – as they should. If you want fairness, go run a non-profit.