Tuesday, January 28, 2020


I noticed that I haven't made a Diary post for a while and decided to now. Although these posts are sort of fillers, I think that they personalize the blog a little and prevent it from becoming a pure book-review website. As always, I have a hard time coming up with books to read. It seems to me that people who read a book a week must be completely undiscriminating. At this stage, the most satisfying books for me are long biographies of intellectuals written by talented biographers. This combination limits the number of suitable books, because there aren't many intellectuals whose lives interest me, and there aren't many good biographers either. Autobiographies have, in theory at least, a lot of potential, but in my experience they don't compare favorably to thorough biographies. For example, Confessions, by Jean-Jacques Rousseau, though probably one of the best autobiographies, is not as complete as the biography I read. Rousseau presented a one-sided view of his life without including the reflections of others, and, as a result, the full context of his development into an outcast remained unexplored. One might conclude, after reading beyond Rousseau's account of events, that it was his pigheadedness that got him into trouble and ruined his friendships. I could arrive at this conclusion only after reading lengthy biographies of Rousseau and Diderot. Another example, Stendhal's autobiography, was good for what it was, but it was written very hastily and didn't cover a very long period. I enjoyed reading Simone de Beauvoir's autobiographical writings up to a point but eventually concluded that her take on reality was incomplete. I sensed that her perspective was skewed, and that became a major distraction. If there were a good biography of her, I would prefer that, but I don't think there is one currently. The autobiographer is inevitably tempted to present a sanitized version of his or her life. A skilled biographer, on the other hand, has an opportunity to understand his or her subject better than the subject understands himself or herself. Even in the absence of such understanding, the reader is provided with additional facts that are useful for making such judgments. I find that process interesting, and my next project is to read a lengthy biography of Charles Darwin; in this instance I'm going to skip Darwin's autobiography altogether. As my reading tastes have evolved, I now think that most fiction is too inaccurate and contrived, most historical writing is too tendentious and general, and most scientific writing, though potentially informative, is emotionally unsatisfying. I am open to different approaches, but this is where I am now.

Of course, I have been following the Trump impeachment proceedings. What is surprising at the moment is that he has any Republican support at all. Nothing could be plainer than that he is a narcissist with little talent beyond self-promotion. None of his domestic or foreign policies are well-informed, and any honest economist will tell you that he deserves no credit for any perceived economic strength in the country. In the most recent insider-tell-all book it is revealed that Trump didn't know that India borders China or what Pearl Harbor was all about. The one thing that sticks in my mind is that if he hadn't swallowed the Russian propaganda about Ukraine he would not have been impeached in the first place. As George Conway, a conservative Republican, announced recently, Trump is actually pretty stupid. The cowardice of his supporters in the Senate is therefore striking. Since there are still no definite signs that the impeachment process will result in his removal from office, we can only hope that he won't be reelected if he's cleared. There is still a chance that witnesses such as John Bolton will turn the tables on Trump.

I might add that my latest readings indicate that automation is the primary force behind current economic and political trends in developed countries. It's always hard for me to get excited about politics, because complex issues such as this never make their way into political talking points. The Republicans in general and Donald Trump in particular have nothing to offer workers who have been displaced by automation, and Trump's attribution of job losses to unfair trade practices skirts the issue. The automation problem is so severe that the Democrats also seem afraid to address it directly. I think that we are only in the early phase of a development that will leave many more people unemployed in future years, and that the process will gradually move from manufacturing jobs to white collar jobs to professional jobs. Tech-savvy candidates like Andrew Yang are ahead of the curve, because they are promoting UBI now, and I'm sure that it will eventually become a political issue. However, the way things have been going, it seems unlikely that the democratic process will be effective for dealing with this particular problem. Democracy itself may be on the way out, because other systems are simply more efficient and effective.

We are having the dreariest winter since we moved to Vermont in 2011. Usually by mid-December the temperature remains below freezing until April and the snow doesn't melt. This year the temperatures have often been above freezing, there is little snow on the ground and it has rained recently. I have been looking at places to move to in northern Maine, where it still gets cold in the winter, but there are no comparable college towns up there that would balance the population of local people with well-educated ones, as we have here. I think that living in Middlebury would be a different experience if the college wasn't here, so I'm not about to move yet. William has completely adapted to his new cat door and ramp setup. He sleeps all day near the fire and gets locked in the basement at night. Late at night and in the morning he comes from the outside and paws on the porch door if he sees that someone is up. He doesn't seem to be catching much now, though he did catch a junco the other day.

Friday, January 24, 2020

Human Compatible: Artificial Intelligence and the Problem of Control

I've just finished reading this book (actually, a cheaper advance uncorrected proof) by Stuart Russell, which represents a continuation of my interest in AI as reflected in the books I discussed in 2017 and 2018 by Max Tegmark and Luke Dormehl. Russell is a prominent AI researcher at UC Berkeley, so his perspective is a little different from that of Tegmark, who is a physicist, and Dormehl, who is a journalist. I found Russell more focused on the important issues at hand, but some of the chapters weren't of much interest to me and I consequently spent little time on them. I got the sense that academic specialization has impeded Russell's perspective a little (it usually does), though he did manage to convey some of the urgency more than Tegmark did in his book and much more than Dormehl did in his.

As the title suggests, the ascent of AI poses new problems for humans that are best addressed now rather than later. Russell uses the analogy that if we knew that an asteroid was going to strike the earth in a few years, we wouldn't wait for its arrival to start planning. The main question is how to prevent AI from becoming dangerous to mankind as technology approaches Artificial General Intelligence (AGI), or, more loosely, superintelligence. I am not a programmer and don't care about programming, so I didn't pay much attention to those chapters. However, Russell also traces the history of AI and compares it to developments that have taken place in other scientific fields. He mentions a story about the physicists Ernest Rutherford and Leo Slizard. In 1933, Slizard read an article by Rutherford stating that although it was known that matter contains a great deal of energy, there was no practical way to extract it. The same day, Slizard thought of the idea of a nuclear chain reaction, which led directly to the first man-made, self-sustaining nuclear chain reaction in 1942. Russell thinks that although the path to AGI is not immediately clear, a new idea that makes it possible could suddenly emerge at any time. So much AI research is occurring at the moment that it would not be surprising at all to have AGI appear within a few decades. Russell believes that the main pieces for its development are already in place. The necessary hardware exists and the use of probabilistic, self-learning algorithms is currently producing useful results.

Most of the problems related to AGI that are addressed in this book concern control. Sections are devoted to finding ways of incorporating human values into algorithms. Russell casts a wide net, which I think is far wider than necessary. For example, he considered ideas from economics and philosophy that I don't think are relevant. I liked the fact that he disagreed with Stephen Pinker, who thinks that AI doesn't pose a risk. The main risk is that AGI will be far more intelligent than any human ever could be and that if it has goals that do not appropriately value human well-being, the consequences could be disastrous. The problem consists mainly of incorporating the right human-friendly behavior into the algorithms and enforcing such procedures in national and international agreements. From a formal standpoint, the agreements would be similar to current international agreements on potentially dangerous gene-editing techniques such as CRISPR.

Although Russell covers his bases well and discusses what life might be like if a safe AGI arrives, that is not the main theme of the book. I think that by focusing on the risk to human existence, he doesn't devote enough space to changes that might occur in favorable circumstances. Even while stating that this would be the most significant event in human history, he doesn't speculate much on how we would react to it. He mentions Ray Kurzweil's idea of merging human brains with robotic bodies and expresses some skepticism, as I do. He also touches on some of the social changes that would be likely to occur. Perhaps everyone will spend most of their time in advanced virtual realities. More immediately, it would be a dramatic shock if AGI, in short order, stopped climate change, created a source of unlimited clean energy and replaced most human jobs along with all governments. While on the surface these would represent positive changes, they would also be traumatic. New ways would have to be found for people to occupy their time. Up to now, people have been able to distinguish themselves by working harder, having better ideas or being more productive than their peers, and those human-level talents would pale in comparison to AGI. The very idea of "genius" would instantly become obsolete, because all of the most important ideas that were originated by humans could perhaps be discovered and improved upon in seconds by AGI. How would people increase their social status under such circumstances? How would human interactions be affected by the introduction of realistic androids? Russell repeatedly refers to the gorilla analogy, in which gorillas were once a dominant primate species but became endangered when humans came along: the same may apply to us when AGI arrives.

I am also a little surprised that Russell has little to say about the potential abuse of AGI by humans. It is to be expected that individuals will attempt to control it in order to achieve their personal objectives rather than to support the welfare of mankind. What would happen, for example, if Vladimir Putin took control of an AGI developed in Russia? This would probably be just as disastrous an outcome as the rogue AGI to which Russell devotes so much of the book.

Tuesday, January 14, 2020

Good Economics for Hard Times III

The chapter on climate change, "In Hot Water," focuses primarily on policies to reduce greenhouse gas emissions and does not address the long-term economic effects. Carbon taxes and exchanges are recommended. Wealthy countries, which are the primary creators of global warming, ought to assist poor nations, which, though not major polluters, continue to use processes that produce carbon dioxide emissions. One of the recurrent themes of the book is that humans tend to engage in "sticky" behavior, which in economics means that they don't like to move or change their habits. In India and other countries, people still cook with wood, which creates both carbon dioxide emissions and poor health. They also burn fields after harvesting, which causes pollution and is not a sound agricultural method. In cases like this, the authors recommend assistance from wealthy countries. Since poor countries are going to take the brunt of climate change, the countries that created it have some responsibility along with the financial resources to help.

The next chapter, "Player Piano," discusses automation, also limiting itself to the near-term effects. Robotics has already taken a toll on employment in manufacturing. The authors believe that this trend will continue and intensify, affecting many other industries and reducing relative wages for all but the tech-savvy. The chapter then abruptly shifts to tax policy, which is continued in the next chapter, "LEGIT.GOV." Banerjee and Duflo spend quite a few pages explaining why the policies of Ronald Reagan and Margaret Thatcher were ill-conceived and have caused negative consequences right up to the present. One idea, that lower taxation on the rich results in economic growth, has been disproven. Another idea, that welfare makes people lazy and unproductive, is also inaccurate. Reagan and Thatcher apparently were sincere, and at the time there was no research to contradict them. The authors, rather courageously I think, go on to attack the American Dream ideology, in which the country is seen as an open field in which anyone can get ahead with hard work, and in which those who fail simply haven't put in enough effort. What they emphasize is that the Reagan-Thatcher policies are responsible for increased inequality in the U.S. and U.K., in which the rich are under-taxed. The top income tax rates in both countries have dropped considerably, allowing a few people to become extremely wealthy – much wealthier than the wealthy of the early twentieth century. Furthermore, the centers of finance in New York and London have permitted some to become super-rich without any visible talents. Excessive wealth accumulation by a small minority, which is enabled by low top income tax rates and the absence of a wealth tax, results in insufficient funds for necessary social programs. Those whose wages have stagnated or evaporated due to automation and international competition eventually require guidance and sometimes need financial support from the social service arms of government.

Ever since the Reagan-Thatcher era, government has been portrayed as an obstacle to progress. Although governments are often inefficient and corrupt, the same is true of private businesses. However, the authors argue that government involvement is essential when inequality rises to a high level and the government becomes the only source able to help those in need. Many of the social issues that exist in the U.S. and U.K. today are relatively nonexistent in countries such as Denmark, which has higher taxation on the wealthy and more funds available for social programs. In this respect, the authors criticize Emmanuel Macron for eliminating the wealth tax in France and placing a surcharge on fuel. These actions had an instant polarizing effect that resulted in riots. Catering to the rich comes at the cost of social stability.

Banerjee and Duflo also devote quite a few pages to the political polarization in the U.S. Much of it, they say, can be attributed to the low self-esteem and anger of those who are suffering economically now. In the U.S., the Protestant work ethic, as invoked by Reagan, tells them that they are failures if they can't succeed financially. A portion of these people gives up, takes drugs and dies. Another portion gets angry and becomes a pawn for corrupt politicians like Donald Trump. As policy wonks, the authors are well-versed in which policies would defuse this situation if they had political support and the proper funding. It goes without saying that Trump doesn't understand or care about the issues and that his incompetent administration is incapable of addressing the underlying causes. But the blame doesn't lie completely on the Republicans, since the Democrats have also supported welfare "reform" and privatization. Rhetorically, Hillary Clinton contributed to her own political demise by referring to Trump supporters as "deplorables." This angered enough people that it may have cost her the election. Hillary Clinton and Barack Obama show all of the symptoms of "elites," and while I share some of their sentiments, displaying them in public can become a prelude to political death. As I've said, this is an anti-intellectual country, and some of the most popular memes in circulation, including the American Dream and American exceptionalism, border on pure fantasy.

The next-to-last chapter, "Cash and Care," discusses practical ways to assist the needy in both wealthy and poor countries. In each case, a form of Universal Basic Income is agreeable to the authors, though they think that it works better in poor countries. There is a lot of detail to wade through here, and I think this chapter would appeal more to someone in the field of Poverty Alleviation and Development Economics, like Duflo.

On the whole I found the book enlightening, because this kind of information rarely comes up in the news. It is encouraging to know that some real expertise exists in the subjects covered. Reading it is an easy way to rise above the nonsense that one faces on a daily basis. These important issues become drowned in misinformation, lies and bad jokes when encountered at the level of American politics. My only complaint about the book is that it doesn't extend its analysis out very far into the future, at which point both climate change and technological change will render the policy atmosphere quite different from the one we're in now.

Wednesday, January 8, 2020

Good Economics for Hard Times II

The next chapter, "Likes, Wants, and Needs," covers a wide range of research regarding human behavior. Part of it reflects the behavioral economics of Daniel Kahneman, Amos Tversky and Richard Thaler, and there is also a section on how social environments affect individual behavior. From my point of view, the latter is the most interesting part of the book so far, perhaps because I was not aware that this kind of research existed in economics. To sum it up in my preferred language, it is about the negative aspects of group selection in a eusocial species, i.e. ours. Rather than discussing this as a biologically inborn human characteristic, the authors simply look at conflicts that arise from people's tendency to behave in a tribal manner. In the contemporary context, this has resulted in political polarization in the U.S. and other parts of the world. There are also examples from the caste system in India. Populism in the U.S. is now based mainly on group identity, thus, politicians are able to gain support by proposing anti-immigration policies. The underlying ideas espoused by people like Trump don't hold up well to scrutiny and resemble propaganda more than fact. For example, "Nearly half of the residents in states with almost no immigrants – like Wyoming, Alabama, West Virginia, Kentucky, and Arkansas – believe immigrants represent a threat to American culture and values." The authors discuss Students for Fair Admission v. Harvard. There used to be discrimination against Jews at Harvard because they were too smart, and now there is discrimination against Asians. It is difficult for Harvard to defend its admissions practices because they seem to contain racist elements and, in the case of Asians who have good academic qualifications but are lacking in personality, Harvard believes they can be rejected. This may look bad from a politically correct standpoint, but I think Harvard has a point, since in their case they want graduates who will be the major leaders in the future, not simply good students, and this makes the definition of merit more complex than plain academic skills. Banerjee and Duflo also cite an interesting study in which Indian cricket players were divided into different teams. Some teams were composed of uniform castes and some were composed of mixed castes. The results indicated that those who played on mixed teams were more likely to befriend players from other castes. The lesson is that diversity need not be seen primarily as tool for rectifying social injustice, and that it may be more useful as a way of defusing class or group prejudices, with the goal of creating a more coherent and stable society. The more people from different groups mix, the less antagonistic they become toward each other. For similar reasons, I prefer to live in socioeconomically mixed neighborhoods, if only because homogeneous neighborhoods are more likely to produce a pernicious form of social identity.

The following chapter, "The End of Growth," is also quite interesting. It covers economic thought at an abstract level regarding the conditions that allow nations to achieve economic growth and reviews the ideas of optimists and pessimists among leading economists regarding the prospects for future growth. The main idea is that productivity increases cause economic growth. This usually occurs best under conditions in which a large, poor labor force is given access to new capital investments. According to Robert Solow, a pessimist, productivity growth gradually slows in wealthy countries and never recovers, whereas Paul Romer, an optimist, thinks that innovation can maintain high growth indefinitely. The economic growth of various countries is examined. The pattern in recent decades has been for some undeveloped regions to develop and become rich quickly. Usually this occurs when there are the right ingredients in place. Europe had tremendous economic growth after World War II, as did Japan a little later. The same phenomenon is occurring in China and other developing countries now. India is a problematic country for economists, since most of them would predict much more economic growth there than has occurred so far. Economic growth may not always be desirable, but the cumulative effect of it globally does seem to have been to reduce poverty worldwide. What I really admire about Banerjee and Duflo is their willingness to deflate myths. They are the first economists I've heard say that there is no proven way to sustain high economic growth indefinitely, and that the countries that have experienced it have done so in an erratic manner such that it is unpredictable. They specifically deny the validity of the Republican mantra regarding tax cuts for the rich: there is no evidence that they produce measurable economic growth. They also note that Mark Zuckerberg of Facebook is making a completely unsubstantiated claim when he says that the Internet will provide future economic growth: there is no evidence supporting that claim either. In my view, Facebook and other social media are simply substitutes for the traditional social contact that has been lost in wealthy countries. Even though I am a stockholder in Facebook, and it has been my best investment by far, I don't think that it is of any real value, and people would be better off if they found real friends. Needless to say, I don't have a Facebook account. I think that the main problem of economics is that the phenomena are too complex for people to understand. To put that in context, I would say that it is far more difficult than theoretical physics, and this explains why economics has never been a reliable tool except in highly circumscribed circumstances.

I am nearing the end of the book and should have one more post. The next chapter does, after all, seem to cover climate change.

Wednesday, January 1, 2020

Good Economics for Hard Times I

This is a recent book by Abhijit Banerjee and Esther Duflo, two (married) economists who won Nobels in 2019. Although economics is not an inspiring subject, I am finding that it is one of the very few academic fields which actually studies human behavior in a manner that could be of some value in matters of policy. This is particularly relevant at the moment, when so many politicians in the U.S. and elsewhere pay little or no attention to findings that are relevant to decisions that will definitely impact millions of people over the long term. I am a little surprised that sociologists don't have comparable publications, but, then, we live in capitalist societies, and economists have clout, whereas sociology professors are perceived as run-of-the-mill academics who produce work with no practical applications. I find books like this informative, because the authors actually know something and rarely appear in the news; they are replaced by ignorant TV personalities who cater to the wishes of politicians and corporations. Although biologists such as E.O. Wilson and Robert Sapolsky have long been writing about our animal natures, economics seems to be the only academic subject that sits at the intersection between science and public policy. Unfortunately, no economics books that I've read have made for exciting reading, and I read them mainly out of a sense of responsibility. Not only is E.O. Wilson a better writer than any economist I've ever read, but he also has a more profound understanding of life on Earth.

As far as I've read, there are chapters on immigration and trade. Donald Trump's name has come up a few times, but he is not really a target of the authors. As far as migration is concerned, generally migrants from poorer countries to wealthier countries do not cause significant disruptions, because they take jobs that people don't want in the wealthy countries. They enrich the local economies by making inexpensive services available to the existing inhabitants. Anecdotally, you can see this in Vermont, where the locals don't want to milk cows or pick apples, and migrants are enabling the survival of those industries. An exception occurs in the case of highly-educated immigrants arriving in wealthy countries. In this instance, they may drive wages down in high-paying industries. I know of one example in which a highly-educated American tech worker feels like a token American in a field that is being taken over by foreigners. Thus, in the case of migration, Trump has got it completely wrong, i.e. poor people should be allowed in and well-educated people should be restricted unless there are no comparable domestic workers. The chapter also examines problems related to migrations within countries. Many examples come from India, while similar circumstances occur in countries like the U.S. For a variety of reasons, some rational and some not, many people don't like to move. Often reduced economic circumstances can be improved significantly simply by retraining or moving to a different location. I noticed something like this when I lived in Dixon, Illinois, which is one hundred miles west of Chicago. Quite a few of the people there were poor because the local economy was a little depressed, and I was surprised to learn that some of them had never even been to Chicago once. They liked Dixon because their families lived there.

The chapter on trade is quite complex but readable. It looks closely at economic dogma that has been present since the early nineteenth century. Specifically, most economists still hold to modern versions of the ideas of comparative advantage and free trade. Comparative advantage and free trade do work in most circumstances, though in the literature at this stage there are many provisos and adjustments that need to be made. Comparative advantage, of course, means that each nation will be able to produce some goods and services at a lower cost than other nations and will therefore be able to compete and trade internationally even if it is weak in some industries. For example, China was poor but had cheap labor which gave the country an economic advantage in some manufacturing industries. The economic rise of China did hurt American workers in rural areas whose jobs were lost due to competition. However, the main point of the chapter is that internal adjustments in the U.S. could compensate for that damage. Overall, there is a slight net gain to the U.S. economy attributable to free trade, and it would make sense to leave free trade intact and allocate internal resources to help those who are hurt by it.  Banerjee and Duflo think that the elimination of tariffs would be fine if the losers to free trade were retrained and moved if necessary. Apparently the U.S. has an underfunded and underused Trade Adjustment Assistance program which has proven to be effective in assisting displaced workers. Increasing the funding and publicizing that program would be a better idea than imposing tariffs, which, under Trump, are helping the steel and aluminum industries but hurting industries associated with agriculture.

Anecdotally, moving has been advantageous to my family. My English ancestors converged on London during the Industrial Revolution and got better jobs. My father's move to the U.S. was not exactly a success story for him, but the rest of the family did eventually attain a higher standard of living. More dramatically, my Armenian ancestors moved from Armenia to what is now western Turkey long ago and became wealthy. When the Armenian genocide was taking shape, they moved to Greece, and later the U.S. and South Africa. In my case, I worked in the printing industry for thirty years, most of which were during its decline. Of the eight unrelated plants where I worked, located in Indiana, Kentucky and Illinois, five have closed. Although the printing industry is hardly affected by foreign competition, the same general rule about being willing to move applies to laid-off employees. For this reason I have very little sympathy for rural Americans who are reluctant to make necessary changes in their lives and vote for incompetent, corrupt politicians just because they would rather hear lies than face the truth.

I still have a long way to go in the book and should have one or two more posts to make on it. So far it hasn't said anything about adapting to the Anthropocene epoch or offered a critique of capitalism, which are probably beyond its scope. Another concern I have, which hasn't been discussed, is that centrally managed countries like China will at some point have tremendous global economic advantages as long as the people in power are competent and not corrupt. By applying research in economics and public policy, China may eventually become a much more functional country than the U.S., which is increasingly burdened by inept voters and efforts to decentralize the government. If Japan, South Korea and Taiwan were able to attain Western-like economies in such short periods, one can expect the same from China, but on a much larger scale.