You are using an older browser version. Please use a supported version for the best MSN experience.

How will Machine Learning affect economics?

The Huffington Post The Huffington Post 28/03/2016 Quora
BRAIN COMPUTER © Christian Lagereek via Getty Images BRAIN COMPUTER

These questions originally appeared on Quora - the knowledge sharing network where compelling questions are answered by people with unique insights.Answers by Jon Levin, Professor of Economics at Stanford University, on Quora.Q: How will Machine Learning affect economics?

A: Machine learning methods are really powerful for fitting predictive models and for doing classification on large-scale, high-dimensional data. These are the data we increasingly use in economics. So I think there's no doubt many machine learning methods will get used more and more often.

One area that's going to get a lot of attention is combining machine learning with causal inference. A big fraction of empirical microeconomics is about finding ways to exploit natural experiments, whether by using instrumental variables, regression discontinuity, matching, difference-in-difference estimators, or other methods.

Large-scale data has great advantages in terms of finding natural experiments (to take a trivial example, if you want to measure how a July 15 price change affected sales, it's much more powerful to have daily sales data than monthly sales data). But for the most part economists trying to estimate causal models on large-scale data are using traditional methods like fixed effects linear regression. Having some easy to use alternatives would probably make a significant difference in empirical research.

I actually think one way machine learning (or really, more data) will affect the field is that after a while it will re-energize economic theory. The reason is that we are going to generate all kinds of new interesting facts - about individual behavior, labor markets, firm productivity, the macro-economy - and having a bunch of new and possibly disconnected or contradictory facts makes a great starting point for new models and theories.

Q: Are MOOCs here to stay or an intermediate step in the evolution of education?

A: The aspect of MOOCs that I hope is here to stay is the way they have allowed pretty much anyone in the world to have access to the same teachers as students at Stanford, Harvard, Yale, etc. That's an amazingly powerful idea with a lot of social value.

The initial structure of MOOCS already has evolved a fair amount. Initially I think there was a view that students would take classes concurrently, that online interaction would substitute for the classroom experience, and that semester-length classes were a natural unit for instruction. These principles seem to have all evolved.

The bigger issue with MOOCs is the tension between creating incentives for high-quality content and the goal of making the content available for free (the "open" in MOOC). To some extent people may put classes online as an experiment or because it's rewarding to reach a large audience. But to get people to pour a huge amount of time and effort into producing a great online educational experience, you need a revenue model so they can be rewarded.

Platforms such as YouTube have figured out how to make great content available for free, either using advertising or by making the online content a hook for something that does generate revenue (e.g.concerts or fame that leads to revenue opportunities). MOOC platforms may figure this out too - by charging for certification, or making the online material a component of a mixed online/off-line program, or by creating and charging for certain classes that aren't open.

I'm optimistic that at least one of these models will work out because I think there is a huge incentive for universities to find ways to make online education work, and for employers to find ways to use online classes for training. Hopefully they will work out with enough success to support a lot of terrific online teaching that's open to anyone at zero cost.

Q: Is there any hope for disruption in the US wireless carrier business?

A: AT&T and Verizon have very strong market positions. They each have around a third of wireless subscriptions in the United States. They are spending $5 billion a year or more to improve their wireless networks.

T-Mobile is the one rival carrier that's gained significant market share. It's gone from around 10 percent to 15 percent of wireless subscriptions. In a narrow sense it also has been disruptive. For instance it changed the way people sign up for plans so it's not necessary to commit for two years. Historically T-Mobile hasn't had a network with equally broad coverage to AT&T or Verizon, but if it may be able to acquire more low-frequency spectrum, which could help it compete with AT&T and Verizon.

These companies all use the same technology, and the technology has large scale economies. So the more interesting question is whether the industry could be disrupted by an alternative approach, for instance a business that relied almost entirely on wi-fi or some other type of very local wireless communication. So far, no one has figured out the mix of technology and economics that would make this work, but I don't know that it's impossible.

These questionsoriginally appeared on Quora. - the knowledge sharing network where compelling questions are answered by people with unique insights. You can follow Quora on Twitter, Facebook, and Google+.

More questions:​

More from Huffington Post

The Huffington Post
The Huffington Post
image beaconimage beaconimage beacon