What we got wrong about Uber and Lyft | Hacker News Reader

What we got wrong about Uber and Lyft

nytimes.com

4 pointsblakesterz4 comments

> Uber and Lyft turned out to be more like taxis.

Who could have possibly predicted that a new, cheaper, and more convenient taxi service (that saves money by dodging taxi regulation/taxes and by classifying its employees as independent contractors) would be used as a taxi service?

And who could have possibly predicted that a cheaper and more convenient taxi service would compete with public transit?

Another example of social scientists not being able to predict outcomes in the same way that physical scientists are able to predict within their fields. Perhaps there never were "experts", that the article still points to.

Social scientists study complex things. Physical scientists study simple things. It's not really a surprise that the latter are more accurate than the former.

But that doesn't mean it's irrelevant. We got along just fine without knowing the mass of the Higgs boson. But we must make decisions about organizing society. Like all sciences that's going to consist of making hypotheses, testing them, and revising them when they're wrong.

As with medicine, we don't stop making decisions just because we don't have complete information or perfect models. We have to make decisions today. Sometimes they're wrong. But failed experiments are the essence of science. They're what drives the next hypothesis.

Sociologists and economists would love to have something as simple as particles or molecules to work with. They don't. They have human beings, which are insanely complex, and even harder to experiment on -- for logistical as well as ethical reasons. The remarkable thing is that there are patterns we can follow at all.

I'm not criticizing attempts at understanding things. It's just that it's not science in any sense as the physical sciences are. You really can't run experiments since you can't control all the variables and the behaviors of people. You can't even accurately measure outcomes or contributing factors. When you don't have the tools that can produce predictions of reasonable accuracy, then perhaps we shouldn't be calling it science and pretending that what these social scientists are telling us is in any way accurate. We've had social scientists try to address all sorts of maladies, such as suicide, lack of interest in education, high divorce rates, etc., and most of these parameters have not improved. If they simply would study, describe, and debate these matters, it would be one thing. But they, the "experts" pretend they know the answers when they're as often right as wrong.

Reply on news.ycombinator.com