We still aren’t close to true artificial intelligence, but fast, accurate machine vision is one of the stepping stones to getting there:

Food manufacturers are combining advances in laser vision with artificial-intelligence software so that automated arms can carry out more-complex tasks, such as slicing chicken cutlets precisely or inspecting toppings on machine-made pizzas. At a sausage factory, more-powerful cameras and quicker processors enable robots to detect the twisted point between two cylindrical wieners fast enough that they can be cut apart at the rate of 200 a minute.

….A high-speed system may have a process time of 10 to 30 milliseconds, or about 100 times as fast as a human, said Bob Hosler, chief operating officer of the U.S. subsidiary of Osaka, Japan-based Keyence Corp. , one of the biggest companies in the vision-products field.

….While vision sensors are good at scanning images for what’s missing, robotic eyes face a wall in inspecting objects from multiple angles, according to engineers at Kyoto, Japan-based Omron Corp. Their proposed solution: big data. To teach a sensor to distinguish a chocolate chip from a burned bit in a cookie, for example, Omrom is using AI to analyze thousands of inspection results. That sort of software will be crucial as robots increasingly permeate the economy.

Personally, I think that distinguishing a chocolate chip from other things that are pretending to be chocolate chips is a key sign of progress. This kind of deceit needs to be rooted out and destroyed with the kind of ruthlessness that only a robot can bring to the job.

More to the point, however, here’s what’s happening to the price of processed meats:

In fairness, most of this decline is just making up for a sudden increase during the Great Recession. Nonetheless, I expect this price trend to continue over the next decade as food processing plants slowly become fully automated, keeping only a tiny shell of humans around to make sure the robots are doing their jobs. This will be great for consumers, but not so great for all the people who currently make their living in food processing plants.

Today is a slo-o-o-o-o-w day. In addition, I experimented yesterday to see if I could force myself to stay awake all day, and I mostly did. However, there’s no fooling the evil dex: it just slammed me into a coma today instead.

But like I said, nothing much is going on, so no harm done. Instead, here is something to amuse you: the inflation-adjusted prices of dog and cat food over the past 30 years.

Your assignment is to decide which line is which. One of them is more expensive than it was in 1985. The other is cheaper. But which is it? Dog food or cat food? You must include your reasoning in your answer.

UPDATE: The orange line is dog food. Its price has increased more than cat food, though I have no idea why.

I’ve had this lying around for a while but never used it because it didn’t turn out as well as I’d hoped it would. But here it is anyway: a triptych of the Los Angeles skyline taken during sunset from Angeles Crest Highway on the way home from Mt. Wilson. Better weather (for clarity) and a better time of year (for the direction of the sunset) would have helped, but sometimes you have to take what you can get.

Demonstrators with United We Dream and others rally in the Hart Senate Office Building on Tuesday to call on Congress to protect DACA recipients.Tom Williams/AP

This has been sitting in a browser tab for a while, and it’s about time I either closed it or used it. So let’s use it:

Just before the beginning of Labor Day weekend on Friday night, a federal judge wrote that the Deferred Action for Childhood Arrivals program was illegal, and would likely fail to pass constitutional muster in the future. And yet, he ruled that the program could continue for the time being.

Andrew S. Hanen of the Federal District Court in Texas said that President Obama exceeded his constitutional authority in 2012 when he created DACA, which protects undocumented immigrants brought to the U.S. by their parents from deportation. But, Hanen reasoned, ending the initiative abruptly after six years would create an unreasonable and unhelpful level of chaos. He compared the act of doing so to unscrambling an egg.

“Here, the egg has been scrambled,” Hanen wrote. “To try to put it back in the shell with only a preliminary injunction record, and perhaps at great risk to many, does not make sense nor serve the best interests of this country.”

I only have one purpose here: to demonstrate the difficulty of repealing executive orders put in place by a previous administration. It is very much not just a “stroke of a pen” that’s required, but a lot of people don’t seem to get this.

There are basically two ways that an executive order can take root: time and time. In the first case, it’s the time taken to produce a detailed plan supported by scientific judgment and public hearings. Once that’s done and an EO is finalized, it can’t simply be tossed in the ash can on a presidential whim. That’s typically considered “arbitrary and capricious” and courts won’t allow it.

In the second case, if an EO has been around for a while, judges will rule the same way Andrew Hanen did. He was skeptical of DACA, but after six years the public has a right to expect that it’s the current law of the land unless it goes through a thorough review by the courts. Until then, the president can’t just ask for a preliminary injunction and then sit back and watch the chaos unfold.

Generally speaking, it takes a lot of effort to get an executive order fully established: it requires real work, real policymaking, and real public consultation. Once it’s in place—especially after it’s been in place for a few years—it’s considered well grounded and justified, and can’t be arbitrarily repealed by a new president who happens not to like it. If you want to repeal it, you have to go through the same process of policymaking and public consultation to demonstrate that the original process got something wrong. That takes a while.

Oh, and generally speaking, the policymaking has to be real, not made-up. That is, you have to do real research and real cost-benefit analysis and real science. Some presidents are better at that than others.

One of the key jobs statistics to keep an eye on is the employment ratio for prime-age workers. By looking only at workers between the age of 25-54, it eliminates issues of students still in school or changes in retirement rates due to an aging population. What it shows is only how many prime-age workers are engaged in market jobs. Here it is since the mid-80s:

EPOP has always been variable, going down during recessions and then back up during recoveries, but generally speaking it increased steadily for 50 years following World War II as more women entered the work force, peaking around 2000. However, during the dotcom recovery EPOP never got back to its previous peak, and as you can see, during the current recovery we haven’t even gotten pack to that peak. Will we ever? Or are we now seeing a steady downward trend over long periods of time? And why?

Nobody knows for sure, and there are lots of theories floating around. For our purposes today, however, the question is whether EPOP is continuing to increase right now. Neil Irwin points out that EPOP has been flat since February, which suggests new workers are no longer being enticed off the sidelines and into the job market anymore even with wages finally going up. This in turn suggests that the recovery is starting to slow down.

Dean Baker says this is just a coincidence: you only see this flatness if you carefully compare February to August, ignoring all the other months and the longer-term data. He’s right, and this gives me an interesting way of demonstrating this graphically. Here is EPOP for the past couple of years:

What I’ve done is run two default trend lines through the data. The black line is just a linear regression. But what if we use a polynomial fit, which would show us if the data is better represented by a simple curve? That’s the red line. As you can see, there’s only the slightest curve, though it is showing a slightly increased growth rate over the past 12 months.¹

My point here isn’t really to say anything about the jobs market, just to show an interesting way of graphically illustrating whether there’s much of a change from the linear trend. In this case there isn’t, which means the February-August flatness is almost certainly just a coincidence.

¹For you data nerds out there, the R²  is 0.9499 for the linear trend and 0.9537 for the quadratic curve. Since you’d expect the quadratic to have at least a slightly better fit in all cases, this tiny difference suggests there’s nothing here. This is a linear growth rate so far, and that’s the best way to capture it.

Say hello to Bodie, a 9-week-old cockapoo. He was acquired by our friends Dave and Eileen a few days ago, and this is his official Hollywood glamor shot. If only this were the 1930s, he’d be a shoo-in to play Shirley Temple’s beloved dog, given to her as a birthday present by Daddy Warbucks.

And here he is in a full body shot, all the better for showing off his svelte figure to the guardians of the casting couch.

Barack Obama was in office for eight years and never managed to convince all his fellow Democrats of this, but now he’s giving it another try:

You cannot sit back and wait for a savior. You can’t opt out because you don’t feel sufficiently inspired by this or that particular candidate. This is not a rock concert. This is not Coachella. We don’t need a messiah. All we need are decent, honest, hard-working people who are accountable and who have America’s best interests at heart. And they’ll step up and they’ll join our government, and they will make things better if they have support.

Say what you will about Republicans all voting for Donald Trump even though he’s obviously a grandstanding moron, but when it finally came down to Election Day they put him in office anyway. Obama is telling Democrats to be more like that: when the opposition party has become what the GOP has become, your job is to vote for the party, not the candidate. After all, at this point if you still think there’s something more important than putting the party of Donald Trump solidly out of business, there’s really not much more to say to you. What more could it take to convince you?

Poor Bernie Sanders has fallen victim to the hack gap. A few days ago he proposed the Stop BEZOS Act, which would require large companies to effectively reimburse the federal government for any welfare benefits used by its workers. Now, make no mistake: as policy, this is a pretty dumb idea. If you want the details, CBPP has you covered here.

But apparently Sanders and his allies aren’t taking criticism of his plan well:

This demagogic side to BernieLand is not attractive.

Politicians need to be able to process negative feedback from experts and analysts. After all, CBPP’s view of this is hardly an outlier — it’s just an unsound proposal. https://t.co/aQc3x3mAXx

— Matthew Yglesias (@mattyglesias) September 6, 2018

HuffPost rounded up a few other critics:

Several progressive policy types ― such as Dean Baker of the Center for Economic and Policy Research and Mike Konzcal of the Roosevelt Institute ― sounded sour notes on Twitter.  “It’s ham-fisted,” Betsey Stevenson, a labor economist at the University of Michigan who served on the Obama administration’s Council of Economic Advisers, told HuffPost.

Can you imagine this happening in the Republican Party? Think tanks and politicians would all chime in to support this great plan. The RNC would post YouTubes. The con men would send out fundraising appeals. The most intellectually honest of them would just stay quiet. This is because they know it’s not policy that matters to voters, it’s hating the right people—and if Jeff Bezos is on the list, then whatever plan most obviously hurts him is a good one. It sends the right message to your voters, and anyway, it’s not like it’s ever going to go anywhere.

But Democrats? If it’s lousy policy, they just can’t help themselves. They have to write white papers and appear on TV and explain in excruciating detail why stopping Bezos is a bad idea. So the voters never hear about it, Bernie is pissed, and once again the most interesting idea Democrats have is to expand the EITC in some obscure way.

Which is probably great, although Democrats tend to carry this idea to extremes sometimes. But if you’re going to be a Democrat, it’s just part of the deal. The Democratic wonk class really, really cares about policy that actually works and actually helps people in a tolerably efficient way. They demand to know where the funding is going to come from, even though they know this is a lose-lose proposition. Basically, if you’re a Democrat, you have to accept that there’s a limit to how dumb an idea the party will support. If this doesn’t work for you, you’ll have to join the GOP. They have no such restraints.

Samer Bouidani/DPA via ZUMA

It looks like the Syrian/Russian attack on Idlib isn’t far off:

The Russian, Turkish and Iranian presidents convened Friday in Tehran in a high-stakes and apparently tense summit that promised to shape one of the final battles of the Syrian conflict and potentially determine the outcome of the war….Iran and Russia back Syrian President Bashar al-Assad, while Turkey supports the Syrian opposition.

…..But Putin on Friday rejected the Turkish leader’s call to halt the fighting. And this week, Russian warplanes carried out strikes on militant positions in at least one city in Idlib, a British-based war monitoring group said. It was unclear Friday whether the three leaders had agreed to tamp down the violence or whether a military assault was imminent.

Meanwhile:

Back in March, President Trump seemed to announce a new Syria policy without notifying anyone else in government. “We’re coming out of Syria, like, very soon,” he said. “Let the other people take care of it now. Very soon — very soon we’re coming out.”

….On Thursday James Jeffrey, a retired senior Foreign Service officer who was recently named Secretary of State Mike Pompeo’s “representative for Syria engagement,” told reporters there’s been a change of plans. “The new policy is we’re no longer pulling out by the end of the year,” he said. As the Washington Post explains, the Trump administration has adopted a broader mission that could keep troops in Syria indefinitely.

….Jeffrey said the U.S. is taking a tougher stance and will not tolerate an all-out attack on the region.

Great. Russia and Iran are almost certainly going to mount an all-out attack on Idlib. The US now says it will not “tolerate” this and plans to keep troops in Syria forever. I guess Trump learned nothing from Obama’s bobble on red lines in Syria.

As writeups begin to appear about the August jobs report, I’ve noticed a fair number of people describing it as “strong.” I wish reporters would stop doing this when the BLS announces fairly ordinary numbers like the 201,000 new jobs we got last month.¹ This isn’t terrible or anything, but jobs have been growing at about this rate for so long that it seems as if we’ve finally all been socialized to think of anything over 200,000 or so as strong. But it’s really not. If I had my druthers, the starting point for “strong” would be 300,000 new jobs—and we’ve hit that number only five times in the past five years.

So let’s not get carried away. Donald Trump will probably describe this as the greatest jobs report ever, but the rest of us should stick to more judicious descriptions. 200,000 new jobs is fairly good, but that’s it. We can do better, and we should.

¹They should especially knock it off when the labor force shrinks and the number of people working drops by 423,000. That just isn’t a great jobs report.