Labor Day 2040: What Happens When Robots Do All the Work?
Labor Day 2040: What Happens When Robots Do All the Work?
The Holiday Weekend From Hell: 2018 When Humans Still
Drove Cars and Flew The Planes
By Clive Irving 09.02.18 10:45 PM ET
Labor Day weekend, 2040.
With millions of Americans again in mass motion it’s a
good time to recall those quaint times when we thought we needed human beings
to steer us safely to our destinations.
Yes, no vehicle could move without a human in charge of
it. Just think about that. Even to get out of the driveway a car had to have
someone at the wheel.
That meant that the car was controlled by people of any
age, from 16 to 90! Imagine your grandfather, myopic and rheumatic, actually
taking your kids on the road. How could that possibly happen?
Don’t even think about that daughter checking her makeup
in the rear view mirror every five seconds.
All that empty air space above your street, too. There
was not a single Uber Lifter to be seen—no eVTOL taxi to get you to the airport
in five minutes, buzzing happily aloft leaving no emissions at 200 mph as the
freeway below is backed up for miles with driverless cars.
Levitation! It’s so easy. How did they do without it?
And, get this. When they got to the airport the airplanes
actually had two guys sitting in a cockpit up front. Two guys! (Maybe a woman
or two among them.) As though the airplane couldn’t fly itself.
It seems so madly illogical to us now that the case for
pilotless airplanes had to wait so long to be accepted.
In fact, a careful scrutiny of the past shows that the
turning point in that argument came, obscurely, in April, 2018.
In Washington, D.C. the House of Representatives was
voting on the FAA Reauthorization Act of 2018—basically to fund the agency. At
the last minute, and with hardly anyone noticing, a provision was slipped into
the Act to prepare the ground for having only one pilot in the cockpit of
cargo-only flights.
Naturally, the first people to protest the move were the
airline pilots—who screamed “this dangerous provision must be removed.”
The president of the Airline Pilots Association
International, representing 60,000 pilots in the U.S. and Canada, claimed that
his union had worked hard to improve aviation safety since 1931, and that it
would never accept having only one pilot in the cockpit because having two
remained essential for safety.
A different way of putting that argument would be that
the union had resisted every effort to modernize cockpits since 1931 if it
involved making any aircrew redundant.
That state of mind did, indeed, go back to the 1930s. As
international airline routes across the continents and oceans were pioneered,
the technology on the flight deck was primitive.
When Pan Am began the first regular transatlantic flights
between New York and Marseilles in 1939 the airplane used was the Boeing 314
flying boat. Twenty-two passengers paid one-way fares of $375—the equivalent of
$12,000 in 2040. As the term flying-boat implies, the accommodations and
service were modeled on great ocean liners, with sleeping berths and even a
luxury suite.
The cockpit, on an upper deck, was equally expansive.
Five people worked up there—a captain and first officer and, behind them in
what was termed an “operations room” were the navigator, who had seven-foot
long chart table, a radio operator and a flight engineer who monitored all the
systems.
When jets arrived in the late 1950s the technology had
advanced enough for the radio operator and navigator to be dispensed with, but
pilots insisted that they still required a flight engineer and so all the new
jets retained a third seat in the cockpit that faced a bank of dials and
switches.
That finally changed in the late 1960s. There was a
landmark dispute between the airlines and the pilots when Boeing introduced its
smallest jet, the 737, in which the flight engineer’s role was removed.
Opposition by the pilots was overcome, but for Boeing’s biggest jet, the 747,
the flight engineer’s place persisted until 1989, when a new iteration of the
jet arrived with technology that made the flight engineer inarguably redundant.
And that, remarkably, was where things remained, with the
need for two pilots unquestioned, for decades. Even the arrival of the
twenty-first century and far greater levels of cockpit automation did not alter
that.
The technology that feels so natural to us in 2040 really
had its origins in the second decade of our century when people realized that,
for most of the time, airline pilots were left with little to do while the
computers flew the airplane from takeoff to landing.
But the pilots still insisted that they were
indispensable—despite the fact that FAA statistics showed that around 80
percent of aviation accidents and close calls were caused by pilot error. In
fact, technology on the flight deck had so effectively reduced the opportunity
for the human factor to screw up that in 2017 not one person was killed flying
on commercial airline jets in the U.S. (In the same year around 40,000
Americans died in highway crashes. So much for the idea of people driving
cars.)
At the same time the wage bargaining position of pilots
was strengthening because the airlines were actually running short of pilots.
In the U.S. the major airlines had to replace more than 18,000 pilots over a
few years because of the mandatory retirement age of 65, and training schools
could not keep up with the demand.
As though by divine right all the human fallibilities
were still sitting up there in the cockpit: fatigue, inattention caused by long
periods of having nothing to do, medical or psychological conditions, and the
inability of the brain to respond with the same speed and situational acuity
that a computer can.
Fighting a rearguard action, the pilots cited cases
where, they argued, the action of pilots had saved an airplane and its
passengers when the computers could not have.
In at least two of those cases they had a point.
In January 2009 Captain “Sully” Sullenberger saved the
lives of 150 passengers by making an emergency landing on the Hudson River. And
in November, 2010 what would have been one of the world’s worst air disasters
was averted when Captain Richard de Crespigny of the Australian airline Qantas
managed to get a giant Airbus A380 that had been badly crippled by an exploding
engine back to earth in Singapore, saving 469 people.
The pilots in both instances were flying Airbus airplanes
with “fly-by-wire” controls and what was then state-of-the-art cockpit
automation. Sullenberger saved his airplane by choosing the Hudson as his
nearest landing point, a split-second calculation that his computers could
never have made. Crespigny’s computers, faced with 120 major systems failures,
automatically shut down 99 percent of the airplane’s electrical systems.
Fortunately there were three off-duty pilots on the A380
in addition to Crespigny and his first officer and it needed the skills of all
five to get to the runway—they had the brains while their computers had become
imbecilic.
Those two incredible feats of airmanship happened within
two years of each other. But it’s important to realize that the Airbus
automation system in use then was basically conceived in the 1980s and even by
2010 remained at a level that today would seem as archaic to us, in 2040, as
Facebook or an iPhone.
But the problem remained that pilots were not ready to concede
that the computers were able to learn from potential disasters. Each emergency,
whether survived or not, produced a new algorithm that could react faster than
any human. Step by step, the human brain was being outwitted. Algorithms are,
of course, the ultimate learning curve—and algorithms, unlike humans, never
forget what they have learned.
“When programs pass into code and code passes into
algorithms and then algorithms start to create new algorithms, it gets farther
and farther from human agency” admitted Ellen Ullman, a pioneering programmer
in 2018.
Given all of this history it is ironic that, in the end,
the public acceptance of pilotless flight did not have to wait for the airline
pilots to accept their redundancy. Uber forced the issue.
It wasn’t easy. When Uber began trials of its Lifter
service in Dallas and Los Angeles in 2020 the FAA insisted that the four-seats
should include one occupied by a pilot. The result was farcical. It was rather
like the scene in Amadeus when Mozart mounts an opera in which the dance
sequences are performed without music because a court authority banned dance
music from the theater, only for the king to be baffled by the result.
The Lifter pilots sat staring at a computer display but
had nothing to do—the flights were programmed and performed flawlessly. In
2025, when Uber’s international network began flying thousands of passengers
every day, the pilots were gone.
And by then Lifters were sharing urban air space with
thousands of Amazon delivery drones. All this traffic was safely controlled by
NASA-developed automated protocols that ensured safe separation in designated
lanes between ground level and 2,000 feet.
By that time all cargo flights across the world were
being flown with only one pilot on the flight deck of the big jets. As the
far-sighted 2018 FAA provision had called for, these pilots were supported by a
new generation of flight management computers that were the forerunners of
today’s robotic systems.
And so, in 2025, pilots finally disappeared from all
airline flight decks.
In order to get there, two pejorative terms had to be
erased from our vocabulary.
The first was Artificial Intelligence. It was the
venerable president of Steve Jobs University in Palo Alto, Sir Jony Ive, who
brought sense to the problem, in the same way that he frequently channeled
Jobs’ aesthetic discipline to every Apple product.
“Artificial implies fake” said Sir Jony in his landmark
2028 Jobs Lecture in Bejing, “… and there is nothing fake about it. So I
propose that in future we embrace the far more apt term, Superior
Intelligence.”
It was, perhaps, just coincidence that SI repeated the
first two letters of Siri, now our universal and indispensable wrist implant
known as iTalk.
The second great step was to stop talking of “robots” as
though they are a sub-human species.
Robots evolve. The human machine does not in its own
lifetime. If we are to be honest we must demonstrate that we know the limits of
what we can do.
Very little cognitive has changed since we lived in
caves. The last time we had any really useful systems update was around 540 BC,
thanks to Pythagoras.
We are not very good at keeping up high levels of
concentration and maximum alertness in all situations. That is why we created
machines that are far better at it than we are. It took too long for us to
admit this. Give the robots the kind of self-respect and dignity that we demand
in our new infinite leisure.
That said, I am disclaiming any responsibility for
everything that is written above. I have consulted many sources in building
this argument, but who am I to know if they make any sense? I am not Clive
Irving. I am a clone assigned to posthumously represent that ancient relic.
Comments
Post a Comment