Average is over, Economics

Average is over: Future of Economics

For the final part of my Average is over posts I want to share some of Tyler Cowan’s thoughts on the future of Economics.

In chapter 11 of the book he writes about the end of average science. He argues that science is going to become harder to understand for three reasons:

  • The problems are ever more complex and simple, intuitive, big breakthroughs are unlikely.
  • Individual contributions are becoming more specialised, and
  • Soon intelligent machines will become researchers in their own right.

He goes on to explain how machine intelligence and specialisation are reshaping theoretical mathematics and physics, and how the average age of Nobel Prize winners has increased over time. Once machine science goes beyond human understanding – think of proofs in the multi-dimensions of string theory – science will also become harder to regulate. But what does all that mean for Economics?

Cowan argues that in recent times data gathering and crunching has been pushing out theoretical intuition. (if this sound interesting, also read Noah Smith’s post on the death of theory) Economists still like their models but, “the real action and value-add comes from the data and its handling, including data from field experiments, laboratory experiments and from randomised control trials”. Sooner rather than later, big data will reinforce the use of machine intelligence in Economics. Data from, for example, your social media profiles, online shopping and loyalty cards will be looked at by a computer programme and it will search for patterns in more complex ways than researchers can do. Cowan hopes that this will reinforce our understanding of some basic regularities behind economic phenomena. He goes on to write:

We are not far a way from having a single de facto, more or less unified, empirical social science. In that social science, researchers invest a lot in learning empirical techniques and then invest some marginal energies in the simpler theories that surround their chosen field of study. Finally, they spend their research time looking for new data sets, or looking to create that data, whether by detective work or by lab and field experiments.

I’m sure this sounds familiar.

He goes on to write that economists who favour more intuitive approaches will have to take a different approach to survive. They will focus less on producing their own original research and will become clearinghouses for and evaluators of the work of others. That means translating the work of others for fellow economists and for the public. It is about seeking out, absorbing and evaluating information.

If you are blogging already, this will sound familiar too.

Average is over, higher education, MOOC

Average is over: Future of Academia

In Average is over Tyler Cowan argues that the future of work will be about working with intelligent machines and how effectively we will do that, will depend a lot on education.

Part of the story about education is online education as one of the places where new information technologies is playing a role. Think MOOCs, but also of blogs, Twitter, Wikipedia, YouTube and TED. Cowan argues that there are new ways of learning and they are based on the following principles:

  • Time-shifting,
  • User control,
  • Direct feedback,
  • Online communities, and
  • The packaging of information into much smaller bits than the traditional lecture or texbook chapter.

So does he believe in the “disruption” and “revolution” that people have been blogging about? Not quite. He highlights some of the important features of the economics of online education that will influence the role it can play: that it will be cheap, that it is flexible, that the profits from teaching innovations will be high and that it allows for much more precise measurement of learning. But then goes on to argue that online education and learning games will supplement current systems.

He makes an interesting point about who will provide online education to supplement current systems. The MOOC model is to be replicable and universal. The business model of universities is to market the quality of exclusivity. Though Harvard and Princeton provide MOOCs, they are unlikely to dilute their brands by providing credits and having thousands of partial alums walking around.

Cowan says that online education will be egalitarian in a specific way: smart, motivated learners from non-elite communities will use it to get ahead, but it will not push the uninterested student to the head of the pack. “This kind of learning is driven by the hunger for knowledge, not by a desire to show off your talent or to “signal” as we economists say”.

Given all this, what will the new world of face-to-face teaching look like?

Probably something like the so-called flipped classroom that everyone is talking about: watch a video on YouTube, come to class for some discussion, practice to homework problems on the e-learning system. Cowan writes: “It will become increasingly apparent how much of current education is driven by human weakness, namely the inability of most students to simply sit down and learning something on their own”. He argues that the better performing students will be treated like the chess prodigy’s are today – they use intelligent software to teach themselves, they cooperate with each other and they get guidance and feedback from coaches. “Their conscientiousness, and the understanding that high wages await them in the world, will enforce hard work and discipline”. The lesser performing students will specialise in receiving motivation. In this, there are some choice bits that made me smile:

Education will become more like the Marines, full of discipline and team spirit.

The teacher is first and foremost a role model and a motivator and to some extent an entertainer.

In the longer run, professors will need to become more like motivational coaches and missionaries.

We could think of the forthcoming educational model as professor as impressario.

Let’s treat professors more like athletics coaches, personal therapists, and preachers, because that is what they will evolve to be.

 And then finally:

Of course, educational institutions are not ready to admit how much they share with churches. These temples of secularism don’t want to admit that they are about simple tasks such as motivating the slugs or acculturating people into the work habits and sociological expectations of the so-called educated class.

I think that it is an excellent chapter and well worth the read.

And what do I think it means for me and colleagues in the School? For example, should I jump into edtech and develop an e-guide, make video’s or apps? It depends.

If your resource is uniquely interesting and well explained, you could have a world market for it. If you are just animating your PowerPoints because the Dean says so, you are probably better off curating resources than creating it. I think that there is already a lot of good resources out there. My contribution can be in how I put them together for my students.

Whether you decide to create or curate, the next question you should ask yourself is whether it will help the conscientious students to teach themselves, or whether it is aimed at inspiring and motivating the others. I frequently hear the hope expressed that interactive pdf’s or slick mobile apps will engage this new generation. But I fear that nice graphics and pop-up messages are the sort of things that everyone gets used to very quickly. If we accept the idea of being coaches and preachers I think that it is obvious that you should spend your extra time on your live show in class, on getting to know your student, on being that real-life exemplar – not creating more e-resources.

I think that we should all think back to the professors who inspired us in the days before e-learning. Were they the ones who read from the book, or had perfect over-head projector transparencies, or a nice list of prescribed reading? The modern day equivalents are the e-guides and apps. Or were they the ones who could motivate and inspire? Now put that in your twitter.

Average is over, future of work

Average is over: Future of work

I suppose that it started with Beyond 2000 and today I always click on the Twitter links to stories about technology and how it is shaping the future. So, speaking at a leadership event hosted for first-year students, I thought the natural choice is to talk about the future of work. There is a lot of stuff out there on the internet and I decided to anchor the story with Tyler Cowan’s new book Average is over. But this is not a book review, it is some of my thoughts, inspired by the book and a bunch of other sources.

I started drawing on a piece called Better than human, in Wired magazine: suppose that before the end of the century 7 out of 10 people will lose their jobs. In the early 19th century 70% of American workers lived on farms and today automation has eliminated all but 1% of their jobs. This will happen again on farms, in factories and to white collar jobs. It is driven by globalisation and what Cowan calls the increasing productivity of intelligent machines.

Before we get to what this means in terms of job polarisation, inequality and average being over, lots of people need some convincing about the technology. The internet is awash with interesting sites an stories:

  • The blog RobotEnomics has categories on AI, driverless cars, drones, the machine economy.
  • For the possible future of manufacturing have a look at the factory for Tesla’s model S.
  • For logistics have a look at these Kiva robots at work (the telling part is how personal it gets but the people are referred to as “warehouse associates”). PS. We already know that working for Amazon is no picnic, but even those jobs are clearly in danger.
  • And it is not only large-scale manufacturing or distribution processes that are being changed by technology. Have a look at Baxter. He is designed to work alongside people, is easy to train and relatively cheap. Get Baxter and a 3D printer and traditional blue collar jobs are changing.
  • And then there is iRobot’s Ava 500 robot: “telepresence-on-a-stick”. I could send it to a few of my meetings.

At a very basic level the point is that in the future we need to work with intelligent machines and that makes for the importance of education and training.

The Wired article breaks down the changes in the workplace of the future into four parts:

  • Current jobs that humans cannot do, but machines can: think of making a computer chip, or internet search, or high-frequency trading.
  • Jobs that humans do today, but machines will eventually do better: think of auto pilots, computerised mortgage appraisal, x-ray analysis, pretrial evidence gathering, telemarketing.
  • Jobs that only humans will be able to do, at first: the advances lie in speech recognition, Watson winning at Jeopardy, algorithms writing like journalists and deciding what movies should be made.
  • Robot jobs that we cannot imagine yet.

Average is over emphasises that we are not in a race against machines but with them. The key question is whether you will be good at working with intelligent machines. Cowan’s best examples are from the world of free-style chess – humans have to decide on the openings and strategy but software is used for tactical play, humans have to look for the gaps or turning points in the game, but they use the software to play out the probabilities many moves deep. The future belongs to effective man-machine teams. But he thinks that does not mean that everyone has to learn to write code. Only that you will have to understand how the systems work and what their failings are likely to be.

This has any number of implications for everything (as a scary side note, read Noah Smith’s post on drones and the end of the age of the gun), but for the future of work it means futher polarisation. It means pressure on middle-skill, middle-wage jobs. A lot has been written about this already, so I am only going to give a few links:

And the end result is likely to be more inquality. Owners of capital and supermanagers will have a greater share of wealth (Piketty, again!).

In the book Cowan goes on to explain who will prosper in the new wold. You can read the short version on the NYT Opinionator blog.

  • If education and training is everything for working with intelligent machines, then open education will benefit the conscientious.
  • People who listen to computers will benefit: from recommending web pages, to products, to partners on dating sites etc., computers can deal with more data and parse out better decisions that most of us who go with our gut. And your smart phone will push this info to you.
  • If you have the human touch to manage and motivate. We still need people for that.

In the book he also proposes other ways in which people will deal with this changing world of under-employment and lower incomes. Amongst other things he hopes that they will move to live in more affordable places and waste less money on useless stuff.

But before we get to the future I still want to write posts about what this all means for higher education and for Economics. It really is a thought-provoking book.

future of work, machine intelligence

Thinking about the future of work

I agreed to do a talk at a student leadership get together next weekend and have been wondering what an economist can talk about. A first thought was to tell a roots of development story, throw in some institutions and then leadership. But that might be a lot to think about if you’re a first-year student spending your Saturday morning listening to lesser-known experts.

So I though the future of work, the data society and machine intelligence story could be interesting. I don’t have anything ready yet, but want to share a few links:

  • Johan probably made me think about this when he wrote about technology, robots and MOOCs why his and my jobs are probably safe.
  • I wrote about the knowledge sharing, co-working story last year, with reference to MOOCs.
  • If you are still worried that open-education and MOOCs might disrupt your academic work, Acemoglu and some co-authors has a math model showing that web-based educational technologies will complement your work.
  • Last week I also caught a Project Syndicate post on the displacement of workers by intelligent machines.

The bulk of my story will probably be inspired by Tyler Cowan’s book Average is over. I am about a third of the way in and he makes some interesting points about teams working with intelligent machines and the lessons from freestyle chess. Earlier this year he shared a few thoughts on the NYT Opinionator blog on who will do well in the future: the conscientious, people who listen to computers, those with a marketing touch and motivators.

And of course I plan to mention movies: Gattaca, Minority report and Elysium. Any other good ones that anyone would like to recommend for machine intelligence driven future dystopia? For things turning out really bad there is Terminator and The Matrix. Battlestar Galactica? But maybe this is going too far from “the future of work” topic!