Yomiuri Shimbun via AP

The Washington Post explains how negotiations are going with North Korea over a possible June 12 summit meeting:

U.S.-North Korea differences over denuclearization remain ‘significant,’ South Korean official says

Right. We want North Korea to get rid of its nukes. North Korea wants to keep its nukes. That’s pretty “significant.” It’s also something that would have been instantly clear if President Trump had bothered with the usual sorts of preparation and working-level discussion before he agreed to the meeting in the first place. It’s not as if this is some kind of subtle issue that’s only now popped up out of nowhere.

What a moron.

Ron Sachs/CNP via ZUMA

I see that President Trump has a shiny new excuse for not talking to Robert Mueller:

President Trump’s lawyer Rudolph W. Giuliani said Tuesday that Trump will not agree to an interview with the special counsel until prosecutors allow the president’s legal team to review documents related to the FBI’s use of a source to interact with members of Trump’s 2016 campaign….Giuliani has previously said that Trump has done nothing improper and is eager to talk to Mueller. On Tuesday, though, he said, he doesn’t want to make a call on whether Trump will sit down for an interview “until they decide whether they are going to give us the documents or not.”

This is like a mafia don refusing to talk unless the FBI turns over everything they have on the informants who have already cooperated. This is, I’m pretty sure, not standard procedure.

At this point, I’m genuinely curious why Mueller keeps putting up with this. Why not just subpoena Trump to testify in front of a grand jury and be done with it? Is he trying to show that he made every possible effort to procure voluntary cooperation first? It’s been months now, and it’s pretty obvious that Trump is never going to agree to an interview. What’s the downside of a subpoena?

This is a western bluebird. It’s Marian’s favorite local bird—except for our hummingbirds, of course—and she’s been bugging me to get a good picture of one for a while. On Friday I finally did. The light was good, there was a tree for me to stand behind, and this little guy posed cooperatively for a nice long time. According to Wikipedia, “the bluebird is the quintessential helpful garden bird. Gardeners go to extreme lengths to attract and keep them in the garden for their advantageous properties. Bluebirds are voracious insect consumers, quickly ridding a garden of insect pests.” What’s not to like?

No, Donald, the father of our country is not pleased with you. Not pleased at all.Martin H. Simon/CNP via ZUMA

We all know that Donald Trump has a tortured relationship with the truth. What’s less remarked on is the fact that Trump has basically admitted over and over that he lies:

“People will just believe you. You just tell them and they believe you.”
          To Billy Bush after Bush called him out for inflating his ratings.

“I could stand in the middle of 5th Avenue and shoot somebody and I wouldn’t lose voters.”
          At a rally in Iowa.

“You know why I do it? I do it to discredit you all and demean you all so when you write negative stories about me, no one will believe you.”
          To Lesley Stahl, explaining why he constantly attacks the press.

“I call it truthful hyperbole.”
          From The Art of the Deal.

Trump told one ally this week that he wanted “to brand” the informant a “spy,” believing the more nefarious term would resonate more in the media and with the public.
          Explaining why he falsely insisted that Stefan Halper was a “spy.”

“They may have been phony in the past, but it’s very real now.”
          After his inauguration, admitting that he lied about unemployment figures.

There’s at least one quote I need to add to this list. Late in the campaign, I think, Trump admitted to a reporter that the whole “Mexico will pay for it” thing was solely about getting a rise out of his audience when his rallies flagged a bit. But I can’t find it. Maybe somebody else remembers this.

Alex Tabarrok points to a new paper this morning that should thrill data nerds and chart enthusiasts everywhere. Steven Piantadosi of the University of Rochester has produced a simple equation with one variable that is “not novel mathematically” but can perfectly fit any set of scatterplot points. Here it is:

As long as you pick your value of θ with extreme precision, this equation will provide a fit of any data, no matter how complex. Piantadosi demonstrates how this works by fitting a scatterplot of an elephant:

How about that? Give me any set of points, and I can describe them precisely with a universal, single-valued equation. All I have to do is choose θ really, really carefully.

As it happens, this is more a novelty than anything else, but Piantadosi says it contains a warning for artificial intelligence:

The existence of this model has implications for statistical model comparison, and shows that great care must be taken in machine learning efforts to discover equations from data since some simple models can fit any data set arbitrarily well.

A human mathematician is unlikely to be fooled by this, but a machine-learning algorithm could easily decide that the best fit for a bunch of data is an equation like the one above. After all, it works, doesn’t it?

Here in Irvine, we are observing Memorial Day by fighting over a veterans cemetery.

A few years ago, the legislature approved construction of a veterans cemetery here, and the plan was to build it on a picturesque plot of land near the Great Park, which is being slowly but steadily built atop the old El Toro Marine base that was decommissioned in 1999. Then some stuff happened—not all of it as transparently as one might hope—and the city council agreed with a local developer to swap the cemetery land for a different plot of land that abuts the San Diego Freeway. Uproar ensued.

So now it’s up for a vote. Yes on B is a vote to go ahead with the land swap. It allows the developer to put up offices near the Great Park and moves the cemetery to the freeway site. No on B is a vote to keep the cemetery where it was originally planned. June 5th will decide the matter.

Austan Goolsbee and Peter Klenow have taken a crack at estimating e-inflation: the inflation rate of items bought online compared to those bought offline. They make use of the Adobe Analytics dataset to calculate a Digital Price Index (DPI) and then compare it to a matched set of items from the normal CPI basket of goods. Here’s the result, with ordinary CPI-U overlaid on their chart:

As you can see, they calculate that DPI is quite a bit lower than the matched CPI: about four points lower over the course of nearly four years. However, the fact that the basket of goods happens to be one that’s been deflating over the past four years makes me wonder if it’s really representative of overall inflation. If it is, then overall e-inflation has been about four points lower than CPI-U. If not, then who knows?

This is above my pay grade, but here are a couple of interesting charts from their paper. First of all, there’s one category where buying online apparently doesn’t help. Can you guess what it is?

Good ol’ health care. Apparently prices rise faster online than offline for medical supplies. Yeesh. But in the news-you-can-use category, check out recreational goods:

Every year, like clockwork, there’s a sudden drop in November. So if you’re looking for a good deal on sleeping bags or canoes, buy online in November. You’re welcome.

Do you know where these mangoes are from? Probably not, thanks to your primitive recordkeeping. But if you had blockchain, you'd know in 2.2 seconds!Momen Faiz/NurPhoto via ZUMA

This morning the LA Times greets me with this headline:

Could blockchain have solved the mystery of the romaine lettuce E. coli outbreak?

Blockchain will get us to Mars! It will cure cancer! It will tell us where our lettuce came from! And not just lettuce:

Take mangoes. The increasingly popular fruit grows on small farms scattered across Latin America, and can harbor listeria, a bacterium that kills 260 people per year in the U.S., according to the Centers for Disease Control and Prevention. Two years ago, [Frank] Yiannas told his staff to trace a packet of sliced mango from a Walmart aisle the traditional way. “I looked at my clock and wrote down the time and date, and I timed them,” he said. “It took them six days, 18 hours and 26 minutes.”

But with Walmart’s new IBM blockchain-enabled system, Yiannas tracked the mango in 2.2 seconds!

This is ridiculous. The problem, as the article makes clear, is that right now a lot of this recordkeeping is done on paper. To speed it up you need to computerize everything in a common format, which is what IBM did. In other words, the reason Yiannas could perform a trackback test in 2.2 seconds has nothing to do with blockchain and everything to do with “cooperative partners agreeing on what information to contribute.”

That’s called EDI, and it’s been around since the 70s. To make it work, all you need is (a) agreement from every single person in your supply chain about how to encode the data, and (b) agreement from every single person in your supply chain to computerize their records using the appropriate ANSI standard. Using blockchain as one of the underlying encoding formats is an architectural choice, but there are dozens that probably work just as well. The hard part isn’t the formatting, it’s getting everyone to computerize their records.

But it’s blockchain! Whoop!

There’s a new poll of presidential experts out that ranks our nation’s chief executives, and Lincoln leads the list with an average rating of 95.03 out of 100. William Henry Harrison got a score of 19.02. I decided to rescale the whole thing going from 0 to 100, with Lincoln on top at Harrison at zero, since, let’s face it, Harrison is sort of zero by definition, right? Besides, I think it also more accurately represents the bottom of the range too.

Visionary or charlatan?Andrej Sokolow/DPA via ZUMA

Bret Stephens doesn’t think much of Elon Musk or Tesla. Fair enough. I’m on the fence about both myself. But in Friday’s column he suddenly decides that not only is Musk a charlatan, but so is the whole idea of electric cars:

The terrible idea is that electric cars are the wave of the future, at least for the mass market. Gasoline has advantages in energy density, cost, infrastructure and transportability that electricity doesn’t and won’t for decades. The brilliance is Musk’s Trump-like ability to get people to believe in him and his preposterous promises. Tesla without Musk would be Oz without the Wizard.

Much of the blame for the Tesla fiasco goes to government, which, in the name of green virtue, decided to subsidize the hobbies of millionaires to the tune of a $7,500 federal tax credit per car sold, along with additional state-based rebates….Electric vehicles were supposed to be the car of the future because we were running out of oil — until we weren’t.

What a weird comment. Energy density is indeed an issue with batteries—just ask my Surface 4 about that—but that’s improving and, at any rate, is already sufficient to power even an inexpensive car for over 100 miles. As for the rest, cost is lower than gasoline, precisely because there’s so little infrastructure involved. Electricity is everywhere, and doesn’t need to be transported at all. The subsidies were mostly designed for cars like the Nissan Leaf and Chevy Volt, not the Tesla. And they had nothing to do with running out of oil. After all, it still takes oil and natural gas to generate the electricity in the first place.

There’s almost literally nothing correct in this entire excerpt except for the (implied) fact that electric cars are still expensive compared to gasoline cars, even when you account for lower operating costs. There’s also the problem that it takes a long time to charge the batteries. But that’s a solvable problem, and it’s being solved. Last month ABB announced a 350kW charging station that can provide a 200-mile charge in 12-15 minutes.¹ That’s not much longer than it takes to fill up with gasoline.

Different technologies advance differently. Some take off almost instantly,² while others have to pass over a technology hump before they become truly usable for mass production. Electric cars are in the latter category, and they still haven’t entirely gotten past all their hurdles. Mainly, we need cheaper battery technology and the buildout of high-power charging stations. The former is a tough technical problem, while the latter is a mostly tedious infrastructure problem. But I’m willing to bet we’re not much more than five years away from either one. There’s certainly no reason to think it will take decades.

¹This originally said “about ten minutes.” ABB claims they can do a 200-kilometer charge in eight minutes, so I figured a 200-mile charge would take “about ten minutes.” However, reader RJ wanted to check this out and contacted ABB. They said 12-15 minutes. So I’ve corrected the text.

²Artificial intelligence is an example of this. We’re still in what I’d call the proto-AI stage, and yet it’s pervasive and useful even in its current crude form. It doesn’t have to work to become widely accepted because it already is—and it will simply get more useful as it continues to improve. Other technologies, like early automobiles, took a couple of decades to get to a point where they were useful for ordinary people at mass-market levels.