Monthly Archives: May 2014

WIRED article on the game of Go, and computers

I recommend this very good WIRED article about computers, go, and intelligence at a high level:

And while programmers are virtually unanimous in saying computers will eventually top the humans, many in the Go community are skeptical. “The question of whether they’ll get there is an open one,” says Will Lockhart, director of the Go documentary The Surrounding Game. “Those who are familiar with just how strong professionals really are, they’re not so sure.”

According to University of Sydney cognitive scientist and complex systems theorist Michael Harré, professional Go players behave in ways that are incredibly hard to predict. In a recent study, Harré analyzed Go players of various strengths, focusing on the predictability of their moves given a specific local configuration of stones. “The result was totally unexpected,” he says. “Moves became steadily more predictable until players reached near-professional level. But at that point, moves started getting less predictable, and we don’t know why. Our best guess is that information from the rest of the board started influencing decision-making in a unique way.”

All of nothing on immigration = failure

Kudos to Senator Marco Rubio for this insight:

“A comprehensive, single piece of legislation on any topic, but especially on immigration, is going to be very difficult to achieve,” Rubio, a potential presidential candidate, said when asked whether he’d push a large overhaul in the 2015-16 session. “We keep talking about the same issue now for 15 years, and everybody is doing this all-or-nothing approach. And all-or-nothing is going to leave you with nothing.”

Read more:

As I have been saying for some time, “comprehensive” in Washington is code for grandstanding.  I want a bill, not an issue. So do the vast majority of Americans, but unfortunately this is a minority position in DC — in both parties.  Rubio is not only on the right side of this issue, but he’s smart enough to have decoded the Senate’s machinations.

Low-hanging fruit abounds

This is a chart for the Cowenists – believers in my friend, Tyler Cowen’s, vision of technological progress slowing because humans have already reached all the low-hanging fruits of innovation. I’m of the belief, to borrow his metaphor, that growth means we are getting taller with each invention, making the higher fruit more accessible. What hath the computer wrought? Greater reach for inventors! A democracy of information!

Consider the amazing diffusion of smart phones, from Pew:

Figure 1

In TWO years, the percentage of people with No Cell Phone was cut in half, down to a level where choice, not income, was probably the deciding factor. The non-adopters are overwhelmingly elderly.

I think Tyler would agree that this is impressive, and maybe dents the pessimists’ case (and he is far more nuanced than the anti-progress pessimist). Nonetheless, I think this chart does much more violence to the Piketty argument that inequality is the defining issue of our time. Hogwash. The young are significantly lower-income than the elderly, yet they have much higher ownership rates of the most amazing piece of property ever, a tangible thing of monstrous value that did not exist half a generation ago.

If mind-blowing, inexpensive, mass-market technological property is the fruit of inequality, then let’s have more. No?

Inequality is (Relatively) Unimportant

I am inspired and bemused by the intense interest in Thomas Piketty’s book Capital in the Twenty-First Century. The formula for becoming the top-selling book (including fiction) in the modern era appears to include these core ingredients:

1. Load with rich, rock-solid historical data about something of keen interest along a political fault line. (In Piketty’s case this data is about income inequality, wealth inequality, and capital).  

2. Write hundreds of pages about the data using tones of caution and care about interpretation.

3. Conclude with wild claims, both about things related to the data as well as things unrelated to the data.

What is bemusing is how the assertion that inequality is “dangerous” has been allowed to run wildly through educated discussion. Do a Google search and you find pamphleteering, not economic science (witness this DeLong link) Maybe it is true, and if it is true, sign me up to stop it.  But is inequality actually important, dangerous, or significantly correlated with something, you know, real?  How does inequality rank against very tangible and measurable ills in our world: child mortality, deforestation, AIDS, slavery, or poverty? Here’s a hint: according to Gallup’s classic “Nation’s Most Important Problem” question, inequality has a mere 3 percent score, and it seems to have been that way for many decades. Inequality just does not rate as a real problem compared to unemployment, growth, health care, education, and honest to God poverty.

Piketty and his apologists assume that inequality matters. Is there any evidence to support the assumption?  Is there, say, a correlation in panel data of nations in Europe over a hundred years between wealth inequality and child mortality? Is it lagged, conditional on income levels and other things? Better yet: Is there a threshold to inequality, where it becomes dangerous above (or below) some Goldilocks zone?

Recall just last year when critics went ballistic over a Reinhart and Rogoff claim about the relationship between national debt and slower GDP growth. There was a 90% threshold  claim, such that debt above that level was dangerous, and some scholars found an error in the data about that threshold (but not the linear relationship) which was then the basis for a full-on paranoid liberal freak-out. As RR noted, even their critics found the same negative relationship between debt and growth, but merely challenged the threshold.

Now contrast the RR debate with this one. Piketty doesn’t even pretend to establish a relationship between inequality and something real.  Imagine Reinhart and Rogoff writing a massive tome and concluding with:  “____ is desirable up to a point. But beyond a certain level it is useless.” or “History suggests that this kind of _______ level is not only useless for growth, it can also lead to a capture of the political process …” which is exactly the ambiguous kind of assertion Piketty gets away with in interviews

I care deeply about poverty; it’s why I chose to study economics. Poverty is real. Inequality is a false idol. Those like Piketty who wish to make it the defining issue of our time have a burden to prove why democracy should shift resources away from other battles against disease, poverty, and pollution. It absolutely must be noted that one of the trade-offs a “war on inequality” will extract is a vast reduction in the number of legal immigrants — one million foreigners legally move to the U.S. under current law, but the Piketty war would drop that by 900,000. “Huddled masses yearning to breathe free?  Sorry, no more space for inequality.” Piketty’s agitation fuels the nationalist sentiment.

If the inequality alarmists cannot substantiate the “danger,” then they stand guilty of empty  ideological posturing.

Google, Women, Immigrants, and Lazlo Bock

I had never heard of Lazlo Bock until yesterday, when he and his work were recommended to me twice. Once at breakfast with some friends who are executives in Bay area startups. Then a few hours later, over coffee with a high-profile HR executive and scholar. Those of you who know that my own research into military leadership in particular, and organizational structure in general, can understand why Bock should have been on my radar before, and is very much on it now.  Bock is doing some very interesting things at Google, things that I think the Pentagon should be watching:

Google data-mines its approach to promoting women (WaPo, 4/2/2014) Don’t neglect the Q&A.

In the U.S., two-thirds of computer science degrees are awarded to non-U.S. citizens. They graduate and immediately go back to their home countries because we can’t keep them. Why not keep that knowledge, drive and passion here?

Bock’s own words (a Google ThinkInsights essays from 2011)

And if you think about it, if you’re an organization that says ‘our people are our greatest asset,’ you must default to open. It’s the only way to demonstrate to your employees that you believe they are trustworthy adults and have good judgment.

gDNA described by Bock (HBR essay)

People Science needs to be adaptive. By analyzing behaviors, attitudes, personality traits and perception over time, we aim to identify the biggest influencers of a satisfying and productive work experience. The data from gDNA allows us to flex our people practices in anticipation of our peoples’ needs.