The latest figures for border apprehensions are out, and the news is either good or bad depending on how you look at it. If you look at the change from last year, apprehensions are way up:

With spring now in the books, apprehensions are up from 14,000 per month last year to 37,000 per month. That’s a big increase, and represents a big increase in attempted border crossings. But if you look just at June and just at absolute numbers, things are looking a little better:

These numbers jump around a lot from month to month, but at least it looks like border crossings aren’t just on an endless upward trend under Trump.

The Straits Times/ZUMA

The latest on North Korea:

Secretary of State Mike Pompeo came away from a two-day visit to North Korea on Saturday without meeting North Korean leader Kim Jong Un or securing a breakthrough in efforts to implement the denuclearization agreement signed by Washington and Pyongyang in Singapore last month….[He said the talks] were “productive” and he claimed progress on “central issues” between the two longtime adversaries.

His rosy outlook was almost immediately rejected by North Korea’s foreign ministry, which called the U.S. attitude to the talks “regrettable” and accused the United States of making unilateral demands for denuclearization. Pompeo just hours earlier said the two sides engaged in “good-faith negotiations.”

Anything else?

More Pompeo: A US-NK meeting is set for on or around July 12 at the interKorean border to discuss the process for the repatriation of Korean War remains (something agreed to at the Singapore summit)

— Elise Hu (@elisewho) July 7, 2018

So North Korea is stalling on even the simplest promise made last month. Needless to say, this gives them more and more time to build out their nuclear program while the United States engages in endless futile talks over trivia.

Just for the sake of context, it’s worth noting that this is why treaty talks usually start at lower levels and only turn into presidential-level meetings toward the end, when the basics have been settled and the top guys can make decisions on a few remaining hard issues. Trump did it the other way around because it made good TV, but the result is that none of the basics have even been discussed, let alone settled, and even if the talks are done in good faith it will take years to do this.

Unfortunately, there’s every reason to think they aren’t being held in good faith. You see, Kim Jong Un likes good TV too.

This is from the Wall Street Journal:

The Trump administration is expected to suspend an Affordable Care Act program that plays a key role in the health law’s insurance markets, a move that could deal a financial blow to many insurers that expect payments. The suspension of some payouts under the program, known as risk adjustment, could come in the wake of a recent decision by a federal judge in New Mexico, who ruled that part of its implementation was flawed and hadn’t been adequately justified by federal regulators, people familiar with the plans said.

First off: this is not the risk corridor program, which was temporary from the start and was recently dealt a blow by an appeals court. This is the risk adjustment program, which is a permanent part of Obamacare designed to prevent insurers from gaming the system by trying to attract low-risk customers. There’s a complicated formula involved that takes money away from insurers with low-risk populations and reimburses reinsurers who have high-risk populations. The net cost of the program is zero.

A couple of years ago an insurer in New Mexico complained that the formula was badly designed. The judge agreed, and remanded the case to the federal government to design a different formula. Under normal circumstances this would not be a big deal: the government would ask for a stay while it rejiggers the formula or perhaps files an appeal. Life would go on, and the worst that would happen is that the formula would get tweaked and some insurers would benefit slightly and others would be hurt. And in theory, this could still happen:

The expected suspension may draw second-guessing from insurers and supporters of the ACA. Timothy S. Jost, an emeritus professor at Washington and Lee University, said it appeared that federal officials might still have other legal options before suspending payments. Mr. Jost said federal regulators could issue a rule formally offering a justification for the risk-adjustment methodology’s use in past years, then ask the judge to consider that.

That’s adorable. But of course that’s not what Trump is going to do. Based on a very limited ruling from a single district judge in a single state, he apparently plans to suspend the entire program. Presumably he will then slow-walk any change to the formula or any appeal of the judge’s decision. It’s yet another way to cause some heartburn for any insurer who has the gall to continue offering policies under Obamacare:

[Suspension] could be a financial blow to those insurers that are expecting payments this fall based on 2017 plans, and potentially for those that expected payments in the fall of 2019 based on their 2018 business. It would rattle those insurers if a suspension occurs, said Larry Levitt, a senior vice president at Kaiser Family Foundation. For insurers expecting payments, abruptly suspending them “would be a big hit to their financial position,” said Deep Banerjee, an analyst with S&P Global Ratings. Estimates of the next round of payouts had already been entered in many insurers’ books as receivables because they are related to 2017 business, he said.

There’s no reason for any of this, which is probably why Trump likes it so much. It’s a message to insurers that he doesn’t care if Obamacare is getting more popular all the time. He’s going to continue messing with them in any way possible, regardless of whether there’s any reason for it. Only if they drop out of Obamacare will they be left alone.

UPDATE: CMS has issued a statement in response to the Journal story. “CMS is seeking a quick resolution to the legal issues raised,” they say, not entirely believably. Basically, they point out that (a) the current formula is the Obama administration’s fault, and (b) even though the New Mexico ruling was handed down in February, CMS is still puttering around asking the court to reconsider its decision. They don’t appear to be in any big hurry, and they’re still promising that guidance on other issues will be issued shortly.

First things first: everything is fine. Everything I said last week is still operative. My cancer load is down substantially and I generally feel pretty good.

And yet. Thanks to the evil dex (postponed to Friday this week thanks to the 4th of July holiday) here I am awake all night once again with nothing to do, so you’re all going to hear me rant a bit. To start with, here’s a new chart I haven’t shown you before:

The basic process of producing antibodies in humans looks like this:

Bone marrow → B cells → Lymph node → Plasma cells →
Antibodies → G, A, M proteins

Bone marrow produces white blood cells called B cells. These travel through the bloodstream to the lymph node, where they’re activated and turn into plasma cells. These in turn produce five kinds of immunoglobulins, better known as antibodies. There are three that we care about: IgG, IgA, and IgM, all of which are large protein molecules. My variety of multiple myeloma happens to be the G type—IgG light kappa if you want the whole enchilada—and as you can see in the chart, it was sky high at first. I had the normal number of healthy G antibodies plus a gigantic number of cancerous G antibodies that eventually would have overwhelmed the healthy ones.

The first round of chemo brought my G antibodies under control, and they dropped even further for a few months when we added dex to my maintenance regimen. Then they stabilized for a while, and around the beginning of the year started rising above the normal range. In May we switched to Darzalex, and the G antibodies plummeted almost immediately back into the normal range. This was a good sign that the cancerous antibodies were being killed off.

So that’s great. But what about my other antibodies? My M antibodies have declined slowly but steadily over the past four years, but not in any dramatic way. My A antibodies were low to begin with and then increased into the normal range, which was good news. But then came the Darzalex, which kills off immunoglobulins pretty indiscriminately:

On Monday I visited my oncologist, who seemed happy with my results so far. Just out of curiosity, though, I asked him what was going on with the A antibodies. Was this anything to worry about? He didn’t say anything much, but seemed not to think it was too big a deal.

So I left happy. Until I got a phone call from the oncology pharmacist telling me that this week’s chemo session was going to be canceled. My platelet level was low, and my doctor wanted to give it a week to recover. That seemed odd, since it’s been low ever since the first week of the new treatment, but OK.

But then I decided to chat a bit, and Melanie admitted that it wasn’t really the platelets that were bothering my doctor. It was the A antibody level. For some reason he didn’t feel like telling me this during our visit, but an antibody is an antibody and you really don’t want any of them to fall significantly below a healthy level. Each of them does different things, and if one of them goes way below its normal range it stops providing protection against certain kinds of infections—in the case of A antibodies, infections related to the mucous membranes and the gut. Because of this, Melanie told me, my chemo schedule was going to be immediately changed to biweekly for one of the drugs in my cocktail and monthly for the Darzalex.

In one sense this is bad news, since it means we’re cutting back on the Darzalex not because it’s done everything it can against my cancerous G antibodies, but because it’s having side effects that could become dangerous. On the bright side, it means the Darzalex is pretty effective. Overall, this is probably modestly negative news, but nothing super bad. After a few more months we’ll know if the A antibodies recover and the cancerous G antibodies continue to decline.

But. It turns out there’s one more thing nobody ever mentioned, and that’s the subject of my rant. My cocktail consists of three drugs, and you all know what the third one is: the evil dex. That particular drug will remain on a weekly schedule through December.

This pisses me off. Not because of the treatment itself, of course: if the dex needs to be weekly, then it needs to be weekly. That’s no one’s fault. But the dex badly compromises my sleep, and it gets worse and worse as time goes by. I don’t really mind the one night a week I stay awake, but a couple of weeks ago it got to the point where the rest of the days were bad too. I was sleeping only about five hours a night and napping a lot during the day. From previous experience, I know that this will deteriorate further, and I’ll be sleeping only 3-4 hours a night before long.

However, I’ve been resistant to restarting the sleep meds. I figured things would ease up when we went biweekly and then ease up some more when we went monthly. But it turns out there was never any reason to hold out. The dex was going to remain weekly the whole time, so I should have restarted the sleep meds after three or four weeks of treatment. And I would have if ANYBODY HAD BOTHERED TELLING ME WHAT THE ACTUAL CHEMO SCHEDULE WAS GOING TO BE.

But nobody did. As usual. My doctor, like every other oncologist I’ve encountered so far, seems to be almost pathologically averse to telling me much of anything, especially if it can be interpreted as bad news of any kind. I actually like my current doctor, who’s nowhere near as squirrely as my first one, but he still doesn’t tell me very much. Sometimes it’s because he doesn’t know, and that’s fine. Obviously there’s a lot of uncertainty in these treatments. But other times it turns out he does know, and only later do I find out the real story. This includes both diagnostic issues, like the decline in the A antibodies, and process issues, like the actual schedule for the drugs I’m taking.

I don’t understand this, and I don’t know what I can do to change it. I know that different patients react very differently to things, and doctors are sometimes cautious because they don’t know for sure what kind of response they might get to even modestly bad news. And yet, I still don’t think this is a good excuse for withholding their unvarnished opinions. That’s their job, after all, and it’s what I expect from them.

I also don’t understand why oncologists¹ are so often oblivious to the mechanics of the health care system they work in. They literally know almost nothing about how appointments are made, how the pharmacy operates, how scheduling works, etc. And when they do know, it often doesn’t occur to them that their patients might like to know this stuff too. My doctor, for example, probably never gave any thought to the possibility that I might care if the schedule for dex was going to be any different from the schedule for the other drugs. I can say, almost literally, that everything useful I’ve learned at a process level about my treatment has come from nurses and pharmacists.

So that’s it. That’s my rant. I have no idea what I can do about any of this, and obviously that’s frustrating. But it also means that I’m going to have a sleepless night once a week for the rest of the year. I’ve already started running out of ideas for nighttime photography excursions, and I sure don’t have enough ideas to fill up 20+ weeks. I better start thinking of other projects to take up my time.

¹I’m limiting this specifically to oncologists. Maybe it applies to all specialists. I don’t know. But my primary care doctor seems generally pretty knowledgeable about the nuts and bolts of how the bureaucracy works.

Brad DeLong posts a Wikipedia page today showing the cost of computing power over time. I adjusted it to show the cost of about 30 petaflops of computing power, which is roughly the consensus estimate of the power of the human brain. Here it is:

Don’t take this too seriously. It’s an extrapolation based on multiplying the power of the fastest/cheapest chips by several hundred, which provides a highly thoretical construct. Actually getting 30 petaflops requires building a huge machine with all these chips running in parallel—plus power supplies, air conditioning, backplanes, support circuitry, etc. etc.—which increases the cost dramatically.

Still, as a sort of thought experiment it shows two things:

• The price of commercially-available computing power has come down about 7x in just four years.
• Even when you account for all the infrastructure, we’re really starting to get close to human-power computers at a relatively low price. This assumes that massively parallel computers turn out to be suitable for AI research, but since the brain is basically a massively parallel biological machine, this seems like a reasonable guess.

And don’t forget software! That’s almost certainly the gating item. My guess is that we’ll have affordable¹ human-power computers in another decade or so, but it will still be two or three decades before we get true AI. Needless to say, this assumes that all the AI boffins figure out how to write the code to do it.

¹By affordable, I don’t mean that you’ll be able to buy one for your desk. I mean that it will be reasonably affordable for just about any AI researcher who wants one. This is important, since software research will skyrocket when human-power computers are available to tens of thousands of researchers, not just the one or two hundred with gigantic budgets.

US District Court Judge Ellen Segal Huvelle.

Last year, a couple of Democratic donors and a former mid-level staffer for the DNC filed a longshot suit against Roger Stone and the Trump campaign. They alleged that the Trump campaign had colluded with Russia to hack the DNC email server and publish the hacked material on Wikileaks. As a result, some of their personal information became public and caused them personal hardship: Social Security numbers were disclosed, they became targets of identity theft, family relationships were strained when one of them was outed as gay, etc. All of this led to “anxiety and distress.”

This was never likely to go anywhere and was barely even reported at the time. It had nothing to do with any Democratic Party organization. It was just three guys who filed a suit. Unsurprisingly, on Friday the suit was dismissed on technical grounds:

U.S. District Court Judge Ellen Huvelle said in a ruling Tuesday evening that the suit’s efforts to tie the Trump campaign and Stone’s alleged actions to the nation’s capital were too flimsy for the case to proceed in a Washington, D.C., court….But Huvelle made clear that her decision was a technical one based on issues of legal jurisdiction and was not a definitive ruling on allegations that the Trump campaign struck an illicit deal with the Russians during the presidential contest.

“It bears emphasizing that this Court’s ruling is not based on a finding that there was no collusion between defendants and Russia during the 2016 presidential election,” Huvelle wrote. “This is the wrong forum for plaintiffs’ lawsuit. The Court takes no position on the merits of plaintiffs’ claims.“

So that’s that. It was a nothingburger case and the judge made no substantive ruling. Naturally that didn’t stop President Trump, who must have heard about it on Fox & Friends or something:

Just won lawsuit filed by the DNC and a bunch of Democrat crazies trying to claim the Trump Campaign (and others), colluded with Russia. They haven’t figured out that this was an excuse for them losing the election!

— Donald J. Trump (@realDonaldTrump) July 6, 2018

Trump didn’t win anything. The suit was not filed by the DNC. There was no ruling one way or the other about collusion with Russia. His entire tweet is a lie from beginning to end. As usual.

Today I have a metaphysical question for you. A few months ago a new cat started wandering around my mother’s house. It was not her cat, but it kept showing up. Then it started coming in and eating out of her food bowl. But it was still not her cat. Then it started eating from her food bowl regularly. And still, it was not her cat.

Then it started allowing my mother to pet it briefly. But still it was not her cat. Then it began sleeping on her bed in the afternoons. Still not her cat. Then it allowed longer petting sessions. Then it became best friends with my mother’s other cat. Then it began hanging out on the piano. Then she gave it a name. And yet, my mother still insists it is not her cat.

Philosophers tell us that defining one’s terms is critical in any rigorous discussion, and certainly edge cases always provoke difficulties. Surely, though, somewhere in there we passed the edge and left it far behind? Perhaps at the point when our furry visitor went from having no name to being “that cat” to being TC and finally Teecee. I submit to you that Teecee is my mother’s cat, and the iron logic of this conclusion depends in no way on her desires. The cat decides, and it has made itself clear.

Anyway, that’s Teecee on the left and Tillamook on the right. They’re both good-size cats, but in this case the camera really has put on a few pounds. They aren’t that big.

YIKES! I got the wrong cats. Here are Teecee and Tillamook:

And here’s the original picture I put up. That’s Luna on the left and Tillamook on the right. Teecee isn’t quite this cuddly yet, but obviously pretty comfortable with the other cats in the house.

Earlier this morning I mentioned that June’s rise in the unemployment rate was due to a big increase in the civilian labor force, mostly caused by people graduating from high school or college. If you’d like to see this in chart form, here it is:

Every year, starting in June and peaking in July, the labor force spikes upward. Then September rolls around, and lots of kids go back to college, thus taking themselves out of the labor force. This is basically meaningless, which is why for most purposes we always use the seasonally adjusted figures, shown by the dotted line. That helps, but some of the spike remains, as it did in June’s numbers this year. Here’s how this affects the unemployment rate:

Unemployment spikes in January, when temporary holiday workers go back home, and in July, when kids are temporarily on summer vacation. The seasonal adjustment smooths out most of this, but not all. That’s one reason among many not to get too agitated over a single month’s numbers, especially in summer.

NOTE: Like most people, I always use seasonally adjusted figures. If you see a chart, it’s seasonally adjusted unless I specifically say otherwise.

Temperatures will reach 95 degrees today…in Siberia. Nick Humphrey explains:

An extreme heat event for this particular region…with high temperatures of greater than 40 degrees F above recent normals….I’ve looked over the European model and there appears to be general agreement over the intensity and timing of this extreme event. It is absolutely incredible and really one of the most intense heat events I’ve ever seen for so far north….more and more sea ice disappears earlier in the season, leaving more dark blue ocean to absorb more daytime sunlight….This is known as Arctic Amplification….causing an abrupt weakening of the polar jet stream.

2018 has unfortunately been a prime example of global warming’s effect on the jet stream. And northern Siberia has been getting blowtorched by heat that refuses to quit because of an ongoing blocked pattern favorable for intense heat.

Would you like to see this in map form? Of course you would:

Today it’s heat, tomorrow it might be extreme cold. Climate change amplifies both, causing an increasing number of extreme weather events. Dan Spinelli has more from Humphrey here:

Once you reach certain tipping points, the Earth takes over and amplifies what we’ve done to make things go faster. These increasing heat events, these extreme weather events, that’s all a process of the Earth trying to speed itself up to get to a new stable state. Is that stable state going to be one that’s suitable for humanity?

Get back to me in a hundred years and I’ll let you know the answer to that question.

The American economy gained 213,000 jobs last month. We need 90,000 new jobs just to keep up with population growth, which means that net job growth clocked in at 123,000 jobs. The headline unemployment rate rose to 4.0 percent, largely because a whopping 600,000 people entered the labor force, of which 500,000 were unemployed. This is most likely due to high school and college kids graduating, so it’s nothing to worry about.

Wages of production and nonsupervisory workers were up at an annualized rate of 2.2 percent. Unfortunately, the latest CPI report shows inflation running at an annualized rate of 2.5 percent, which means workers got a pay cut last month. Overall, the employment situation seems to still be fine, but workers just can’t seem to string together any decent wage gains. The last six years have been a pretty good recovery for the rich, but pretty ho-hum for the rest of us.