Interview with Andrew McAfee

MIT’s Andrew McAfee is one of the world’s most influential thinkers working in issues at the intersection of technology, the economy, and society.  He is a principal research scientist at the MIT Sloan School of Management, and the cofounder and co-director of the MIT’s Initiative on the Digital Economy.

McAfee’s 2014 book The Second Machine Age: Work, Progress and Prosperity in a Time of Brilliant Technologies, co-authored with his MIT colleague Erik Brynjolfsson, argues that we are entering a a time where technology – in particular the automation of cognitive tasks – will impact society in a scale comparable to the impact of the “First Machine Age,” or Industrial Revolution.

Waltham, Massachusetts, the home of the Waltham Review, has a great deal of experience with Industrial Revolutions.  The city was the birthplace of the First Industrial Revolution in North America, which began with Francis Cabot Lowell’s epoch-making theft of British intellectual property in order to found the first integrated cotton mill.  The mill was established at a point along the Charles River where a ten-foot drop creates abundant water power, and its creation was a turning point in American industrial, financial, and labor history.  Today, Waltham is home to many of the leading companies and venture capital firms of the Second Machine Age.  A high-profile example is the defense technology firm Boston Dynamics, whose robots display amazing and alarming agility on both two legs and on four.

I had the chance to pose eleven questions to McAfee, giving him an opportunity to describe his views on the present and future of work and society with readers of the Waltham Review.

 

WR: Thinking about your career before your most recent book, about ten years ago, you invented the term “Enterprise 2.0.”  To what does this term refer? 

AM: I used this term to refer to the use of what we now call “social media” explicitly for business purposes within the company and up and down the supply chain.  Using blogs, wikis, and all of these new tools to get corporate work done.  At that time, almost all corporate collaboration was accomplished through email.  But now we have tools like “Chatter” from Salesforce.com and “Slack”, which are incredibly popular.  So now it feels like, ten years later on, Enterprise 2.0 might finally be here.

 

WR: In your book The Second Machine Age, you highlight both the promise and the disruption of the new information technology revolution.  It’s safe to say that you are in the optimistic camp when it comes to technology and economic growth, and the noted economist Robert Gordon is in the more pessimistic camp.  What differences are behind this difference between your outlooks?

AM: I think there are a couple of things going on.  One is that Bob is just a lot less impressed by new technology developments than I and my coauthor are.  Bob very rightly points out that technologies like electrification and the development of the internal combustion engine were big deals – it is almost impossible to overestimate how important these developments were for economic growth in America, and also for improvements in well being.  I grant all of that.

I look at what is going on now: we have the interconnection of the world’s population with the Internet and mobile phones, we have the ability of all of those people to access for the first time  a decent chunk of humanity’s accumulated knowledge and contribute to it, we have the fact that we really do now have some kind of artificial intelligence for the first time, which I think of it as a digital colleague that can look at huge amounts of information that can find patterns in those data and tee things up for us and present them to us, look at them with us.

Then, if I go beyond the purely digital, we can edit the code of life with arbitrary precision using the new CRISPR technology.  And, from what I understand, we are finally getting a decent picture of how proteins are created inside of cells.  This means we might now be able to create our own proteins – we will no longer be limited to the proteins that nature gave us.

When I debate Bob, the line I always use is this “if you are not impressed by these things, and you do not think they are going to be transformative over the course of the 21^st cent, then you really are hard to impress.”

I think that the other difference is that Bob is laser-focused on economic growth, according specifically to the ways that we have been measuring economic growth, i.e. GDP growth.  And GDP growth is really important, and we should watch it, but it is not the same thing as improvement in human welfare and human well being.  And I look at all of these new developments, and think about what they will do for human welfare – regardless of what they will do for measured GDP – and I cannot help but be optimistic.

 

WR: You have some specific policy prescriptions for dealing with the disruptions of the Second machine age.  Can you talk about them? 

AM: One set of prescriptions is this: let’s just get the basics right.  Let’s do the completely uncontroversial things right.  I think that Alan Blinder has a line: “the areas where there is the most agreement among economists are the areas where they are listened to the least.”  And those are things like infrastructure investment, liberal immigration policies, especially for higher skilled immigrants, education reform, fostering entrepreneurship, and investing in basic research.

So I always hum to myself “EIEIO” – “Education, Infrastructure, Entrepreneurship, Immigration, and Original research.”  Those are uncontroversial things, and yet we are doing a mediocre to lousy job at them.

And then secondly, there are some areas where there is a bit more controversy, but we still have a point of view. Erik and I wrote in Foreign Affairs this year, and outlined some policy goals.  One goal is just to have fluidity and flexibility wherever you can.  Because times are changing quickly, and rigidity is a bad response to that.  The other is putting in place policies that directly promote or encourage good, old-fashioned work.  A Negative Income Tax or an expansion of the Earned Income Tax Credit encourages work, and a universal basic income doesn’t encourage work.  So policy should tilt in the direction of approaches that promote work and labor force participation.

 

WR: The causes of this change are far from fully clear, but we do seem to be seeing – both in the United States and within (as opposed to between) other countries worldwide – an increasing concentration of income and wealth at the very top.  Is there any chance that capitalism and our current form of technological progress are inconsistent in some ways with democracy?

AM: There is always that chance, but I don’t believe that it is the case.  If we are honestly concerned about the top 1 percent or the top 0.1 percent, then, OK, great, let’s raise marginal tax rates.  Let’s increase the estate tax.  There are tools to reduce that kind of economic inequality, and we have used these tools before.  The big Reagan tax cuts reduced the top marginal tax rate in the US a huge amount.

But more fundamentally, I don’t think our problem is that we have too many super-wealthy people in the US.  I think our problem is that we don’t have enough solidly middle-class people.  And that’s not a 1-to-1 story.  The rich are not robbing the middle class – that is not what is happening.

If you look at demagogues who are running for office in rich countries, a lot of their support is coming from the middle class.   They are not backed up by crazy plutocrats.  They are backed by people who feel left behind, marginalized, and excluded from the conversation, and part of that is economic.

 

WR: You are, in general, an optimist.  Let’s try to be pessimistic for a second.  What are the national and global risks that keep you up at night? 

AM: First of all, I don’t think I’m an optimist by temperament. I am an optimist by evidence.  Most of the things that we care about have been getting better.  Looking at data makes me really optimistic.

Now the stuff I’m nervous about.  First of all a couple of trends in the data, that mainly have to do with the stagnation and the hollowing out of a big chunk of people.  It’s hard for me to see how that trend is going to spontaneously reverse itself.  That is a policy challenge for us – what do we do about that?

Looking at global warming makes me pessimistic.   I don’t know exactly how pessimistic to be; I am not enough of an expert in the specific science there.  But cooking our planet is not good.

Now, when I look at the improvement trajectories in technologies like solar power, I get extremely optimistic.  And the way I think about our future is this: I just don’t know if these new technologies are going to ride to the rescue of the planet quickly enough.  I just don’t know.

And then the last huge question mark for me that can lead to some pessimism.  I see that we are democratizing access to some incredibly powerful technologies.  In general, we should welcome that.  We want to put better tools in the hands of more people.  As we do that, we are putting better tools in the hands of madmen, and psychopaths, and our enemies, and different flavors of bad actors.

 

WR: Do you have a policy prescription to deal with this democratization in access to destructive power?

AM: I have a policy non-prescription: don’t try to preemptively restrict access to these tools.  For example, bad guys can, and I think will, use drones to fly explosives into things, unfortunately.  Should we take drones off the market?  Absolutely not.  Should we forbid any more research on CRISPR, since it is such a powerful technology?  No, we should not do that.

And there is a really interesting battle of perspectives on this topic.  Because these are really powerful technologies, spreading quickly, and in the face of that, some people think we need more “upstream government” – the phrase I have heard.  I think I understand the motivation for that, but I think it is a terrible idea.

 

WR: Why is it a terrible idea?

AM: I cannot think of the body of bureaucrats that I would trust to accurately look into the future of technological progress and guide us in the best possible directions.  Even if those bureaucrats were the smartest people around, and completely disinterested and unbiased, I would not trust them to succeed. And to be clear, people like that do not exist.

What I favor is this.  There is another phrase you hear, called “permissionless innovation” – in other words, don’t make me submit what I want to do to any oversight or signoff body.  I think we will get much better outcomes with permissionless innovation.

 

WR: Of all of the new technologies that you are seeing emerge right now, which one do you think has the chance to change our society and economy the most rapidly? 

AM: That is a really hard question, particularly with the “most rapidly” at the end of it.  Time scales are hard to figure out.  But I would say that “deep learning” – which is getting so much attention these days – is a super powerful and transformative technology.

 

WR: What might deep learning have the potential to transform?

AM: A couple of examples.  Deep learning is the technology behind the Go-playing computer, which is now the world’s best Go player.  And that is kind of trivial example: Go is just a very difficult, subtle strategy game.

The team that built that technology then applied it very quickly to managing the energy at one of Google’s data centers.  And becoming more energy efficient throughout the economy is kind of a big deal.  As soon as they flipped the switch, that data center became 15 percent more energy effective.  We are going to see more examples like that all throughout the economy.

 

WR: Thinking about a specific futuristic technological dream and nightmare, I just have to ask this: how far away are we from having Freeman Dyson’s self-replicating robots? 

AM: We could build robots that could build robots today.  But self-replicating with some kind of autonomous will, and a consciousness that may or may not line up with ours?  I am going to quote Andrew Ng, a rock-star AI scientist.  He says “worrying about that kind of thing is like worrying about overpopulation on Mars.”

 

WR: Last question.  Is there any question that you’re feeling disappointed right now that I didn’t ask you? 

AM: No. I think you have done a remarkably thorough job!

 

WR: Thank you very much, Professor McAfee. 

 

---

(Artwork: Pieter Brueghel the Elder (Dutch, 1525-1569)  The Corn Harvest  1565)