Friday, June 28, 2019

Driverless Vehicles and Driving Jobs: 4th Annual Forecast

Autonomous vehicle technology implementation and commentary since our July 11, 2018 issue has been discouraging.  What effect has that had on its projected proliferation dates, and on future taxi and truck driving positions?

First, two recent articles on the state of the field.  Slate’s June 13th “How Close Are We to Self-Driving Cars, Really?” stunned me with a subtitle citing Chris Urmson, “who helped pioneer the technology at Google,” claiming “it could be 50 years before we see them everywhere.”  2069!  In the first paragraph, we got Urmson, interviewed by April Glaser, giving it “5 to 10 years” for “robots cruising down the road in a handful of cities and towns across the country” (2024-2029), and no less than 30 (2049) for them to be “everywhere.”  That’s a far cry from most attitudes we saw as recently as last year.  Urmson gave sober views, mostly a compendium of those in print in 2017, and added that maps for autonomous vehicles will need, instead of “where the Safeway is,” exact locations of stoplights, lanes, and right-of-way rules, and that at least some driverless vehicles should be cone-shaped “so you can see all the way around (them) really nearby.”  He had unfounded assumptions about people using much more mass transit, but even those fit in with the general idea I got, that we should be wondering if Urmson had been sleeping for two solid years. 

The other was Lawrence Ulrich’s June 20th New York Times “Driverless Cars May Be Coming, but Let’s Not Get Carried Away.”  Was he saying that’s what the same publication did 20 months ago when it published an entire magazine dedicated to the view that they were certain to happen soon?  After appropriately starting with Tesla’s Elon Musk’s ravings that autonomous taxis will be ubiquitous in 2020, Ulrich moved to sources saying that “none of us have any idea,” that “driver-free transport will begin with a trickle, not a flood” (even in cities and parts of same with great emphasis and full sanction?), named one prominent research director saying there is now a “trough of disillusionment,” and that “making a robocar so foolproof that consumers and automakers can trust it with their lives, including one-in-a-billion situations” (necessary even with those 30,000 American-driver-caused annual deaths?) is much harder than demonstrating the likes of ability to make emergency stops.  Ulrich then described Cadillac’s Super Cruise semiautonomous-software option, a solid Level 2 product now available on at least one model, which monitors driver alertness but allows pedal and steering-wheel-free driving “on major highways.” 

Otherwise, what has happened and, more importantly, what has not happened over the past year?  The two events are above – Super Cruise, and pessimistic commentators getting attention.  The following, though, did not occur, or if so were not publicized.  There was no more driverless implementation in ordinary, wide-open areas.  There was no substantial rollout of automated shuttles, even in tightly controlled settings.  There was no debate or pushback between those at the front of the field and national, state, or city regulators.  There was no widespread use of such vehicles even in the most accommodating and suitable areas, such as Phoenix.  There was no evidence of more aggressive implementation in already sanctioned parts of cities, such as San Francisco or Singapore, despite geofencing providing invisible but firm barriers.  There was no significant public relations effort to persuade or cut the fears of ordinary people on the technology.  I heard nothing about related activity in other countries.  And, despite all of these non-happenings, there was no sign of anything like a solid wall of individual resistance or any discussed consensus of regulatory disapproval.

Some things, though, stayed the same.  We still had Musk all by himself with hyperoptimism.  We had overemphasis on the one fatal accident.  We had consortia, alliances, and division of labor.  We had vaporware peddlers offering vague and unreasonable promises without product.  And we still had technical progress, amount strangely unpublicized.

Taking all this, accepting doubt milder than that from our interviewees above, and realizing that though Super Cruise means good things for Level 2 implementation in particular we are still in real trouble all over, we get:

As before, you can see the level definitions at, though the document here has been updated.  Expect more from Work’s New Age as this issue evolves. 

Friday, June 21, 2019

Robots and Artificial Intelligence: Four Viewpoints, Three Useful

Driverless cars are not the only area where less of substance is being communicated than a year or two ago.  This combined field, which is now hard to split into its two parts and within the decade may in most observers’ minds be fully merged, has over the past seven months only got four articles, beyond technical updates, on my desk.  What do they have to say?

The oldest, Salon’s November 23rd “Robots are making us less human, too: “Certain things essential to the democratic fabric erode,”” didn’t follow its title, but provided ideas worthy of discussion.  It was an interview of television director Maxim Pozdorovkin, whose work “The Truth About Killer Robots” ran on HBO November 26th.  The show considered “the ethical dimension of using robots in our everyday life,” with concerns based on the worst these automata have given us, from spectacular industrial accidents to one employer after their implementation going from 3,200 workers to 800, and on to workers’ general losses, as in Kurt Vonnegut’s Player Piano, of “self-satisfaction” and “a sense of becoming better.”  Pozdorovkin also claimed the shrinking likelihood of doing tasks such as banking and hotel check-ins with humans “undercuts the social fabric.”  These are real if sometimes here overstated problems – how can we best deal with them?

An intriguing characteristic of good artificial intelligence systems is that they throw away what we consider best practices and invent their own, which are often superior.  That was the underlying main point of “One Giant Step for a Chess-Playing Machine” (Steven Strogatz, The New York Times, December 26th.)  The piece discussed AlphaZero, described as “a deep-learning algorithm” as opposed to “the world’s strongest chess programs” which had a larger base of technical knowledge but lacked insight into unprogrammable “basic principles.”  AlphaZero, after it “played against itself millions of times and learned from its mistakes,” “crushed… the reigning computer world champion” in a hundred-game match in which it, though having 72 draws, won the remaining 28 and was thus undefeated.  That could only have been because it was developing its own ways of winning.  Of course, chess is only one opportunity for such self-directed thinking machines, of which AlphaZero constituted “humankind’s first glimpse of an awesome new kind of intelligence”; Strogatz mentioned “the great unsolved problems of science and medicine, such as cancer and consciousness; the riddles of the immune system, the mysteries of the genome.”  Scary, and we may be there before we know it. 

On the diametrically opposite side, we had “’AI could send us back to the stone age’: In conversation with the End Of The World” (Olivia Tambini, TechRadar, January 31).  This piece is really a summary of previously published work, some in this blog, about the dangers of “general intelligence” capability, such as AlphaZero above freed to solve the world’s problems.  We again, though with different words, got the Terminator autonomous goal-seeking problem, along with the results of an interview with author and philosopher Nick Bostrom, concerns similar to those named in the landmark now-19-years-old Bill Joy Wired paper “Why the Future Doesn’t Need Us,” the black-box nature of high-level machine knowledge development, and the need to somehow program “benevolence.”  Nothing especially new here, but one of the world’s most suitable topics for repeating, re-repeating, re-re-repeating, and so on.

Last was Joanie Courtney’s April 5th Fox Business “The robots are here: New, unheard-of job titles signal growing occupations in digital age.”  There are not enough of those I have always said, and nothing here, especially Courtney’s efforts to blame employers’ insufficiently paying practices on workers, changed my view.  It is true that we should “encourage among future workers… the ability to keep learning and adapting,” though that is not a skill but a meta-skill, and that there is a gap between ordinary people becoming computer-theory experts and them responding to eras ending with the folded arms of 1980s autoworkers, but the “simulation training” Courtney advocates needs to be much more than that. 

Next week, on to autonomous vehicles and my annual projections. 

Friday, June 14, 2019

The Economist’s “Great Jobs Boom” – How Accurate Is That?

Three weeks ago, this venerable publication had quite a cover.  It had a headline “the great jobs boom,” and a cartoon showing dozens of people at work, from a travel hawker and a bus driver to a plant seller and a pizza deliverer.  The two connected articles’ shared thesis was that, despite some noted flaws, these are tremendous times for employment opportunities, that, although “a recession will kill it off… it deserves a little appreciation,” and that, in a positive sense, “the rich world’s jobs market could have more surprises in store.”  The pieces made numerous points, the most important of which I sorted into four categories:  pro-boom, anti-boom, good and underrated, and bad and oversold. 

Supporting the boom interpretation were three main ideas.  By an undisclosed definition, the gig economy, per the articles, only covers about 1% of American jobs.  Many government-released numbers, such as seasonally adjusted and unadjusted unemployment rates, net job growth, and number of people working, are better than they have been since Nixon was president.  Over the past ten years, we have had massively accumulating prosperity progress, and most recent months it has continued. 

On the other side, two points were negative.  Much of the 2010s jobs gains have been from women’s still continuing to join the labor force, which must reach a peak sometime.  Even The Economist admitted here that “middle skilled jobs are becoming harder to find,” contradicting the idea of huge growth almost by itself.  And, outside the articles, the American Job Shortage Number or AJSN still shows latent demand for 15.6 million additional positions, which, though off a third from the Great Recession’s top, seems only a solid improvement.

As well as the positive points above, several other things the unbilled authors mentioned point toward employment gains being more sustainable than we would otherwise think.  The first, “in the ten years to 2016 the cost of filling a vacancy fell by 80% in real terms,” attributed lower hiring costs to the ability of employers as well as potential workers to cheaply use job websites.  Second, what many have perceived as lagging average pay rates have helped.  Third, general progress, which I had thought would level off in 2017 or 2018, has continued, and next year we may well see sub-3.0% unemployment rates with average 200,000 monthly jobs gains.  Fourth, we have thus far had little business deregulation, which will probably change and allow even more positions.  Fifth, numbers found by the authors suggest that, as well as gig engagements, the overall count of low-paying jobs is not as high as what might be called a consensus view, and almost anyone under 80 knows that bad ones, actually in decline with higher low-end wages, are nothing new.  And sixth, the result of the most stunning statistic in the piece, that “in 2018, the employment rate among people of working age was the highest ever in Britain, Canada, Germany, Australia, and 22 other OECD countries,” could someday also, at least potentially, include central North America.

Not all underrated and less-known factors, though, favor our employment situation.  Minimum wage increases, though phased in and generally well-focused, have done hidden damage by cutting the number of jobs created, and scheduled raises are just getting started.  The Phillips curve, which previously showed a relationship between pay and employment, will continue being inoperative, as easily duplicatable products such as software and still-high latent demand for jobs mean productivity and low joblessness are no longer determinants.  The lack of immediate widespread automation danger may be gulling observers into thinking the robots will never come – but, with the next recession, we will know otherwise.  The colorful example contrasting 30%-unemployment Malaga’s clean roads and “buzzing” restaurants with 2.6% San Francisco’s “rough sleepers and empty lots” underscores the gap between front-line rates and how many more people would actually work if given the chance. 

As you can see, we have a mixed bag.  To resolve the problem, let’s go to a definition of “boom,” as provided in this case by Merriam-Webster: “a rapid expansion or increase:  such as… c:  a rapid widespread expansion of economic activity… d:  an upsurge in activity, interest, or popularity.”  We’re close here, but I don’t think this one qualifies, for two reasons.  First, it hasn’t been rapid.  Second, it just isn’t strong enough.  With over 15 million people still on the sidelines and a well-acknowledged shortage of positions suiting most workers’ ability and credentials, 2000s North Dakota it isn’t.  We are in relatively excellent times which as far as we know haven’t peaked yet, but after ten years it isn’t even “widespread,” and it’s hardly difficult to find people, especially among the 1.3 million who have been officially unemployed for over half a year, who know from experience that we don’t have a true boom.  So we can appreciate the help these times have given millions of jobholders, but let’s not get carried away.  Our jobs crisis is not over, and, low unemployment rates notwithstanding, sadly shows no signs of ending.  Eye-catching or not, that is the truth.

Friday, June 7, 2019

May’s Employment Data: An Off Month for Once, with American Job Shortage Number (AJSN) Up 350,000 to 15.6 Million

Expectations for this morning’s Bureau of Labor Statistics Employment Situation Summary, if the identical-to-April Fox News projection of 185,000 net new nonfarm positions is a good indicator, were much the same as last time.  However, the results weren’t.  We gained only 75,000, and, while the other numbers maintained strength they didn’t improve as a group.  The front-line seasonally adjusted unemployment rate stayed at 3.6%, with the unadjusted one up 0.1% to 3.4%.  We had 5.9 million unemployed persons, with those out for 27 weeks or longer up 100,000 to 1.3 million.  The count of those working part-time for economic reasons, or seeking full-time positions while keeping smaller-hours ones, fell 300,000 to 4.4 million, reversing last month’s 200,000 worsening and then some.  The two measures of how common it is for Americans to actually be working, the labor force participation rate and the employment-population ratio, stayed the same at 62.8% and 60.6%.  Private nonfarm payroll wages matched the last result, gaining 6 cents per hour, a bit more than the latest, 2.0%, annual inflation rate, to reach $27.83.

The American Job Shortage Number or AJSN, the measure showing how many additional positions could be quickly filled if getting one were as easy as getting a pizza, gained 357,000, as follows:

Compared with April, changes in those unemployed and discouraged exactly cancelled out, leaving almost the entire variation to those saying they wanted to work but did not look for it in the past year, up over 400,000 with an AJSN effect of 325,000.  The number of people claiming no interest in jobs fell almost one million – that only cut the AJSN by 50,000, but showed us once again that membership for many in that category is fluid and negotiable.  That was almost exactly offset by the 164,000 rise in those wanting to work but being temporarily unavailable.  Year-over-year, the AJSN improved 340,000, with two-thirds of that from lower official unemployment and the rest from shrinkage in five other groups.  The share of the AJSN from unemployment as the BLS defines it reached another long-term low with 31.7%.

How did this month look?  Not terrible but hardly good.  Let’s call it sluggish.  The two participation ratios, one of which worsened 0.2% in April, did not bounce back and are still, fine employment times notwithstanding, about half a recession from reaching new post-1977 lows.  We can take the poor, well-below-population-increase-requiring net jobs gain, and the other metrics here are still generally looking good, but those are due to previous months, not May 2019.  Accordingly, the turtle stayed just where he was.