Can an Artificial Intelligence Write A Good Blog Post?

A blog post by an artificial intelligence that has just been taught to write about an artificial intelligence is still far from being a good blog post by any human being.

There are many reasons why a blog post by an artificial intelligence is unlikely to be a good blog post by any human being.

First, the post is still a huge piece of written material, so it will be a big task for the AI to read it all. This is similar to reading a huge, long book, which is a huge task for you as an author.

Second, it is likely that the AI will use its knowledge about writing to create a very bad, misleading, or otherwise nonsensical blog post. In this case, the AI will be writing a blog post about its own stupidity.

Third, even if the blog post is not written by a computer but rather by a human author, the human author will not understand it and will not be able to correct it. This is because the blog post will include very basic, incorrect, or outdated knowledge about writing,

Consistency about elasticities, revisited

From Ross Rheingans-Yoo, the content is all his, I will not do a double indent:

  • If marginal wealth is taxed an additional 0.5%/yr at the high end, then fewer people will amass and invest that much wealth—some will instead disperse it among a wider number of family members, donate it to charitable or political causes, or spend it on expensive consumption. (Saez and Zucman, in their potential-revenue analyses, assume that this effect is quite small, and that the wealthy will mostly accept lower returns on wealth.)
  • Similarly, if the marginal opportunities to invest became worse by 0.5%/yr, fewer people would chose to invest, by the same token. Additionally,the effects should be the same size, as it’s the same decision-makers facing the same incentives!
  • But if pushing on the price (read: rate of return) has little effect on the quantity of investment, then pushing on the quantity of investment should have a large​ effect on the price! (Unless we’re at some magic kink in the supply curve for unspecified reasons…)
  • So a small amount of additional capital competing for investment opportunities should quickly reduce the competitive rate of return.

What’s the practical upshot? Well, if the authors’ assumptions about revenues are right, then Piketty’s r>gr>g “wealth spiral” can’t proceed unchecked, since capital simply can’t accumulate without competition quickly reducing the average rate of return back below gg.

Effort: The Unrecognized Contributor to US Income Inequality

That is a new paper by J. Rodrigo Fuentes and Edward E. Leamer:

This paper provides theory and evidence that worker effort has played an important role in the increase in income inequality in the United States between 1980 and 2016. The theory suggests that a worker needs to exert effort enough to pay the rental value of the physical and human capital, thus high effort and high pay for jobs operating expensive capital. With that as a foundation, we use data from the ACS surveys in 1980 and 2016 to estimate Mincer equations for six different education levels that explain wage incomes as a function of weekly hours worked and other worker features. One finding is a decline in annual income for high school graduates for all hours worked per week. We argue that the sharp decline in manufacturing jobs forces down wages of those with high school degrees who have precious few high-effort opportunities outside of manufacturing. Another finding is that incomes rose only for those with advanced degrees and with weekly hours in excess of 40. We attribute this to the natural talent needed to make a computer deliver exceptional value and to the relative ease with which long hours can be chosen when working over the Internet.

I like that last sentence in particular.

My Conversation with Ted Gioia

Here is the audio and transcript, the chat centered around music, including Ted’s new and fascinating book Music: A Subversive History.  We talk about music and tech, the Beatles, which songs and performers we are embarrassed to like, whether jazz still can be cool, Ted’s family background, why restaurants are noisier, why the blues are disappearing, Elton John, which countries are underrated for their musics, whether anyone loves the opera, whether musical innovation is still possible, and much much more.  Here are some excerpts:

GIOIA: …Spotify still isn’t profitable. I believe Spotify will become profitable, but they’re going to do it by putting the squeeze on people. Musicians will suffer even more, probably, in the future than they have in the past. What’s good for Spotify is not good for the whole music ecosystem.

Let me make one more point here. I think it’s very important. If you go back a few years ago, there was a value chain in music — started with the musician, worked for the record label. The records went to the record distributor. They went to the retailer, who sold the record to the consumer. At that point, everybody in that chain had a vested interest in a healthy music ecosystem in which people enjoyed songs. The more people enjoyed songs, the better business was for everybody.

That chain has been broken now. Apple would give away songs for free to sell devices. They don’t care about the viability of the music subeconomy. For them, it could be a loss leader. Google doesn’t care about music. They would give music away for free to sell ads. In fact, they do that on YouTube.

The fundamental change here is, you now have a distribution system for music in which some of the players do not have a vested interest in the broader musical experience and ecosystem. This is tremendously dangerous, and that’s the real reason why I fear the growth of streaming, is because the people involved in streaming don’t like music.

And:

COWEN: Do you think music today is helping the sexual revolution or hurting it? Speaking of Prince…

GIOIA: It’s very interesting. There’s market research and focus groups about how people use music in their day-to-day life. Take, for example, this: you’re going to bring a date back to your apartment for a romantic dinner. So what do you worry about?

Well, the first thing I have to worry about is, my place is a mess. I’ve got to clean it up. That’s number one. The second thing you worry about is, what food am I going to fix? But number three on people’s list — when you interview them — is the music because they understand the music is going to seal the deal. If there’s going to be something really romantic, that music is essential.

People will agonize for hours over which music to play. I think that we miss this. People view music as distance from people’s everyday life. But in fact, people put music to work every day, and one of the premier ways they do it is in romance.

COWEN: Let’s say you were not married, and you’re 27 years old, and you’re having a date over. What music do you put on in 2019 under those conditions?

GIOIA: It’s got to always be Sinatra.

COWEN: Because that is sexier? It’s generally appealing? It’s not going to offend anyone? Why?

GIOIA: I must say up front, I am no expert on seduction, so you’re now getting me out of my main level of expertise. But I would think that if you were a seducer, you would want something that was romantic on the surface but very sexualized right below that, and no one was better at these multilayered interpretations of lyrics than Frank Sinatra.

I always call them the Derrida of pop singing because there was always the surface level and various levels that you could deconstruct. And if you are planning for that romantic date, hey, go for Frank.

There is much more at the link, interesting throughout, and again here is Ted’s new book.

Revisiting NSA Spying

In 2013 in light of the Snowden revelations about NSA spying I wrote, Did Obama Spy on Mitt Romney?

Did Obama spy on Mitt Romney? As recently as a few weeks ago if anyone had asked me that question I would have consigned them to a right (or left) wing loony bin. Today, the only loonies are those who think the question unreasonable. Indeed, in one sense the answer is clearly yes. Do I think Obama ordered the NSA to spy on Romney for political gain? No. Some people claim that President Obama didn’t even know about the full extent of NSA spying. Indeed, I imagine that President Obama was almost as surprised as the rest of us when he first discovered that we live in a mass surveillance state in which billions of emails, phone calls, facebook metadata and other data are being collected.

The answer is yes, however, if we mean did the NSA spy on political candidates like Mitt Romney. Did Mitt Romney ever speak with Angela Merkel, whose phone the NSA bugged, or any one of the dozens of her advisers that the NSA was also bugging? Did Romney exchange emails with Mexican President Felipe Calderon? Were any of Romney’s emails, photos, texts or other metadata hoovered up by the NSA’s break-in to the Google and Yahoo communications links? Almost certainly the answer is yes.

As I read the situation, mass government surveillance has now become accepted in America, as in China. This bit remains relevant:

Did the NSA use the information they gathered on Mitt Romney and other political candidates for political purposes? Probably not. Will the next president or the one after that be so virtuous so as to not use this kind of power? I have grave doubts. Men are not angels.

“And what a decade it was”

From Alex X.:

With the decade coming to a close, I would be curious on everyone’s favorite of the decade [gives list of categories]:

Without too much pondering, here is what comes to mind right away:

Film: Uncle Boonmee Who Can Recall His Past Lives, or A Touch of Sin.  Might Winter Sleep by next?  It was probably the best decade ever for foreign movies, the worst decade ever for Hollywood movies (NYT).

Blockbuster/action film: Transformers 4?  Big screen only, live or die by CGI!

Album: Kanye West, My Beautiful Dark Twisted Fantasy.

Single: I don’t see an obvious, non-derivative pick here that really stands out.  Kendrick Lamar’s “Alright” probably is the mainstream choice, but do I ever go over to the stereo to put it on?  Janelle Monae’s “Make Me Feel” is another option, but is it such a big step beyond Prince?  Lorde or Beyonce?  LCD Soundsystem seems more about the entire album, same for Frank Ocean.  Something from Kanye’s Yeezus?  To pull a dark horse option out of the hat, how about Gillian Welch, “The Way It Goes“?  Or Death Grips “Giving Bad People Good Ideas“?  I’ve spent enough time on Twitter that I have to opt for that one.

TV Show: Srugim, Borgen, The Americans.

Single Season: Selections from same, you know which seasons.

Book Fiction: The Ferrante quadrology and Houllebecq’s Submission.

Book Non Fiction: Knausgaard, volumes I and II.

Athlete of the Decade: Stephen Curry or Lebron James.

What are your picks?

Can we spend another $52 trillion without raising middle class taxes?

The question seems like a joke, right?  Yet because so much of our elite media class wants Elizabeth Warren to win, they are contorting themselves into every possible direction to make this one sound coherent.  It is not a question of whether total nominal expenditures on health care go up or down, but rather of thinking through incidence and opportunity cost and where the real burdens of the plan will fall.  Those are the core themes of my Bloomberg column, here is one excerpt:

Another part of the plan is to pay lower prices — 70% lower — for branded prescription drugs. That is supposed to save about $1.7 trillion, but again focus on which opportunities are lost. Lower drug prices will mean fewer new drugs are developed. There is good evidence that pharmaceuticals are among the most cost-effective ways of saving human lives, so the resulting higher mortality and illness might be especially severe.

And the close:

Warren’s proposals, when all is said and done, are best viewed not as a way of paying for her program but as a series of admissions about just how expensive it would be. Whether or not you call those taxes, they are very real burdens — and many of them will end up falling on the middle class.

By the way, here is a good NYT summary of Warren’s financing plan.  Here is a good Maxim Jacobs tweet:

It’s really hard to pick out which part of her plan is most insane?: – Lowering brand drug pricing by 70%? – CMS paying specialists less money – Taxing unrealized capital gains – Claim hiring more IRS agents will raise $2.3 trillion – “Not one penny in middle-class tax increases”

Here is more from Peter Suderman.

The Global Kidney Exchange Programme

In my WSJ review of Al Roth’s excellent book Who Gets What—and Why I wrote about Roth’s proposal to extend the idea of kidney swaps globally:

It’s often the case that a living donor is willing to give a kidney to a loved one, but the loved one can’t accept it because of immunity mismatch. But if a pair of such mismatched donors could be found (call them A and A´ and B and B´), then perhaps a match could be found by a crisscross pairing: Donor A could give to recipient B´ and donor B could give to recipient A´, thus solving the mismatch problem and saving lives.

…Today such multi-way exchanges are becoming common….Mr. Roth, however, wants to go further….why not open U.S. transplants to the world? Imagine that A and A´ are Nigerian while B and B´ are American. Nigeria has virtually no transplant surgery or dialysis available, so in Nigeria patient A’ will die for certain. But if we offered a free transplant to him, and received a kidney for an American patient in return, two lives would be saved.

The plan sounds noble but expensive. Yet remember, Mr. Roth says, “removing an American patient from dialysis saves Medicare a quarter of a million dollars. That’s more than enough to finance two kidney transplants.” So offering a free transplant to the Nigerian patient can save money and lives.

It’s hard to think of a better example of gains from trade (or a better PR coup for the U.S. on the world stage).

Recently, Rees et al., (including Roth) announced the first such global kidney exchange:

We report the 1‐year experience of an initial Filipino pair, whose recipient was transplanted in the United states with an American donor’s kidney at no cost to him. The Filipino donor donated to an American in the United States through a kidney exchange chain. Follow‐up care and medications in the Philippines were supported by funds from the United States. We show that the logistical obstacles in this approach, although considerable, are surmountable.

Naturally, some people aren’t happy because of “ethical” objections. Minerva, Savulescu and Peter Singer write in defense of the program:

Lurking behind all the arguments against the GKE is the assumption that people who are poor are incapable of autonomous choices. So, if they appear to choose to act in ways that benefit not only themselves, but people in HICs, they must have been coerced, exploited, or commodified.

…Poverty does not necessarily make a person unable to choose to donate a kidney to a loved one, nor does it make someone incapable of weighing the pros and cons of an option like that offered by the GKE. Poverty does narrow down the options available to people, and often forces them to settle for an option that is not as good as a wealthy person would choose. That, however, is irrelevant to the ethics of the GKE if that programme provides a better option to patients in LMICs who need a kidney than any other option currently available to them.

…It would be tragic if such misguided objections were to prevent the GKE from realising its potential to reduce suffering and save the lives of rich and poor patients alike.

Hat tip: Frank McCormick.

Harvard sentences to ponder

We show that Harvard encourages applications from many students who effectively have no chance of being admitted, and that this is particularly true for African Americans.

Here is the whole abstract, by Peter Arcidiacono, Josh Kinsler, and Tyler Ransom:

Over the past 20 years, elite colleges in the US have seen dramatic increases in applications. We provide context for part of this trend using detailed data on Harvard University that was unsealed as part of the SFFA v. Harvard lawsuit. We show that Harvard encourages applications from many students who effectively have no chance of being admitted, and that this is particularly true for African Americans. African American applications soared beginning with the Class of 2009, with the increase driven by those with lower SAT scores. Yet there was little change in the share of admits who were African American. We show that this change in applicant behavior resulted in substantial convergence in the overall admissions rates across races yet no change in the large cross-race differences in admissions rates for high-SAT applicants.

And from the paper’s conclusion:

If the goal of recruiting African Americans is not simply to increase the diversity of matriculants, but also to achieve racial balance in the admit pool and/or racial balance in admit rates, then the policy could be deemed a success. As an example, admit rates for African American applicants were twice as large as admit rates for Asian American applicants in 2000, but by 2017 were approximately the same. Why Harvard might careabout the racial distribution of admit rates and applicants is not obvious. What is clear is that each year there are a significant number of African American high school students who have a potentially false impression about their chances of being admitted to Harvard.

Here is the full paper.  And here is a recent paper by Howell, Hurwitz, and Smith, with related results.

The effect of district attourneys on criminal justice outcomes

In the United States, elected district attorneys’ offices prosecute over 85% of all felony cases, but we know little about their effect on local criminal justice outcomes. Using a newly-collected dataset of district attorney elections, I show that Republican district attorneys lead to a 18-21% increase in new prison admissions in the two years following their election, while nonwhite district attorneys lead to a 10% decline. In both cases, there are no significant effects on local crime or arrest rates. These results show that the identity of the local district attorney is an important determinant of incarceration rates.

Here is the paper, by Sam Krumholz, on the job market this year from UCSD, that is not his job market paper, here is his full portfolio, public economics and law and economics, to me one of the more interesting candidates this year.

Monday assorted links

The Prescription Escalator

Ask anyone and they will tell you that their prescription costs are rising. But generic drug prices are falling (also here) and generics are 80-90 percent of all prescriptions. Moreover, although branded drugs are expensive total out-of-pocket costs for the population as a whole are flat or even decreasing as Michael Mandel points out:

[A] May 2019 research report from the Agency for Healthcare Research and Quality reported that average out-of-pocket spending for prescribed medications, among persons who obtained at least one prescribed medication, declined from $327 in 2009 to $238 by 2016, a decrease of 27 percent. Data from the Bureau of Labor Statistics Consumer Expenditure Survey shows that average household spending on prescription drugs fell by 11% between 2013 and 2018.

Moreover, OECD data shows that average out- of-pocket spending on prescribed medicines in the United States ($143 per capita in 2017) is actually lower than countries such as Canada ($144), Korea ($156), Norway ($178), and Switzerland ($215).

So are people simply mistaken about what they are experiencing? Not quite. Mandel uses the metaphor of the prescription escalator to explain the apparent paradox:

It turns out that an escalator is the appropriate model for prescription drug costs for individuals. As people get older, they unwillingly ride the prescription escalator, with their average spending on prescription drugs rising by about 5-6% per year. This figure assumes no change in the underlying price of drugs. Rather, people fill more prescriptions as they age.

In other words, every individual experiences an increase in prescription costs as they age even though for the population as a whole prescription prices are flat or falling–a form of Simpson’s paradox. The driver of higher costs is usage not price. People aged 65-74 have on average 25 (!) prescriptions to fill, more than two and half times as many as people aged 25-34 (about 9 per year).

Understanding the prescription escalator is important because regulating drug prices–aside from being a bad idea–won’t solve the perceived problem.

…even if drug reform efforts were successful and there were no more increases in drug costs, every individual would still face a 5.6% increase each year in drug spending as they got older. That would total 30% after five years, and 70% after ten years, across the board.These are enormous increases.

Indeed, the prescription escalator is a sign of success. If drugs weren’t successful we wouldn’t buy more of them when we were older and sicker and costs wouldn’t rise.