There are three stories, and Bloomberg misses all three of them.

by w3woody

Why the Robot Takeover of the Economy Is Proceeding Slowly

  • Replacing salespeople with software is tough sell for one CEO
  • Automation shows limits at U.S. factories making cars, motors

… Hundreds of companies are trying to disrupt the way we consume, work, or move. The economy’s growth potential could be higher if smart machines could turbocharge how humans go about their tasks. Higher productivity, or output per hour, would boost corporate profits and may help U.S. workers finally get a pay rise.

That economic nirvana just isn’t happening yet.

Well, sure.

Sadly, however, the Bloomberg article seems to boil down the problem to one of sales and to lack of flexibility.

I think the reason why everyone seems to miss why “robots” (I hate that term) are not taking over the economy and displacing workers is because, for some reason, over the past half decade, we completely misunderstand what a “robot” is. And the three potential stories here have been overlooked.


I think the blind spot comes because most writers of crap like this think of a “robot” as some sort of magical device which will partially or completely replace workers, and they think of productivity software as being the intellectual equivalent: an all-software magical device which can partially or completely replace workers.

But that’s a really crappy definition, because it sweeps under a “magic” rug what is actually going on.

I’d prefer to think of “robots” as one of a variety of tools, fixtures and jigs which help humans or replace repetitive work done by humans.

When thought of this way, nail guns (which can drive a nail with the touch of a button, replacing the repetitive banging of a hammer) is a tool which replaces workers, as it allows a single framer on a home building job site to do the work of several framers when framing a house. A circular saw allows fewer guys on a job site since one guy can now saw 2×4 boards faster than a handful of guys armed with hand saws.

We see the same thing in other industries as well. Red Digital Cameras have been replacing traditional film-based motion picture cameras–reducing the need for someone to run film to a processor for “dailies” (the daily development of the film to see the results the next day), and reducing the need for film developers in the process. Fire trucks have replaced the need for bucket brigades–lines of people carrying buckets of water to put out a fire. Modern software IDEs have made programmers more productive than in the days of text editors and make files. And so forth.

And winding backwards, we see the tools these things replacing were themselves replacements of older tools. The history of the nail gun and of the hammer can trace its way to the day when primitive man tied a rock to a stick to give him greater leverage–replacing hammering things by holding a rock.

So the first story is this:

The story of tools replacing humans by making people more productive goes back to the invention of the first tool during the stone age.

So thinking that “robots” have changed this story because robots are somehow “magical” misses the larger story, which is one of progress as we learn better ways to do the things we want to do.


This, I think, carries us to the second story.

The hype around robotics suggests a world where humans have little input in manufacturing. Talk to BMW managers, however, and it’s all about getting the right mix of humans and machines in a world where customization and complexity are big challenges.

Managers are constantly on the lookout for new ways to insert more automation. One recent addition: a small “co-bot,” working next to humans that rolls protective foil on a door frame. Having a machine do this simple task several hundred times a day saves time and wear on human hands.

Let’s be clear what this “co-bot” is: it’s a tool. Like a specialized nail gun, it is a simple fixture which helps a person unroll protective foil onto a door frame.

And let’s be clear what the overall problem here is for BMW. It is not that robots could not be constructed which can pick the correct custom stereo system or install the right trim. It’s that the automation necessary to do these things cost money–and often, especially when the number of combinations are great enough, it is far cheaper to hire a person to install and quality check the installation, than to program a complex series of automated tools and fixtures to do the same thing.

The same problem can be seen with the other stories here:

“The biggest bottleneck to machine learning is trust,” he said. As a result, finding the “hero CEO” who will tell their shareholders they are trimming a sales team to rely on a black box is difficult.

Melding big data with manufacturing is the next step for hundreds of companies, and it is challenging, said Bryan Tantzen, head of manufacturing and industry solutions at Cisco, the networking-technology giant.

“You have to connect these machines to transform them,” he says. There are obstacles. Not all machines are loaded with sensors.

But behind the story of a lack of trust and a lack of sensors is something far more fundamental: the cost/benefit analysis.

How much will it cost–in terms of time, in terms of effort, in terms of changes to the existing processes–and how much will it benefit the organization, in terms of increased productivity.

Risk, by the way, is also a cost: risk can be translated directly into monetary terms by calculating the chances that something will go south, and how much it will cost to fix the problem. Some risks can essentially “bet the company”–at which point the cost of your risky change could effectively mean bankruptcy and ruin.

So the second story is the cost/benefit analysis: sometimes the cost of buying the tool, adopting the tool to your business, incorporating the tool into your processes is just too high.

Robots are not some sort of free magical ingredient that will allow a manufacturing plant to reach the utopian ideal of replacing a million square feet full of workstations full of employees into a black box where raw materials go in, and finished products go out, to be sold by lead software which automatically calls retailers to ask what they want.

They are tools–and some tools are just wrong for the job.

And sometimes, like the foil unrolling tool at the BMW plant in South Carolina, the right tool is simple, relatively inexpensive, and specially designed for the job.


There is a third story here that is not even hinted by this article, but hinted by other articles.

And that is, as automation replaces humans helping to drive down the cost of products, sometimes corporations instead keep humans on and provide better products, rather than drive the price of their products to zero.

Look at movie making. Innovations such as Red Cameras have made it possible for a “pro-sumer” (a non-professional consumer wanting professional quality products) to build a film rig which can shoot professional-quality video. In theory cameras like Red allow second- and third-shooting units to be much smaller; it permits professional movies to be assembled with fewer people working on them.

Yet while we certainly are seeing the rise of films shot “on the cheap”: Clerks was shot for around $25,000, Primer for around $7,000 and Paranormal Activity for around $15,000–what we see more and more are smaller budget films (and independent films) using some very sophisticated effects that would have been impossible just a decade before.

Films like Ex Machina, shot on a $15 million dollar budget (in inflation-adjusted terms, perhaps 1/4th that used by Star Wars, which was itself shot on the cheap using sophisticated effects for its time) featuring some amazing special effects. And special effects have creeped into nearly every film, being used to augment establishing scenes or to add realism.

It’s cheaper to make a film, but instead, we’ve chosen to make more visually sophisticated films.

And we see this same story repeated over and over again. Technological improvements on how we make gasoline-powered car engines have allowed us to make cars with greater mileage, but instead we make more powerful cars. (I remember when I first learned to drive in the 1980’s when 85 horsepower was a lot for a small family car, and when Camaro muscle car with a V8 got 120 horsepower. Today, a Fiat 500 comes standard with a 100 horsepower engine, and the lowest trim level of a modern Toyota Camry family car has 178 horsepower and a 0-60 time that is 3.3 seconds faster than a 1980’s Ford Mustang equipped with a V8.

That, despite fuel economy rising from an average of 16 MPG to a fleet average of 25.9 MPG.

We see this everywhere. Our cell phones are easily far more powerful than the most powerful supercomputers three decades before, and are essentially portable mini-computers connected to the world wide web that allow us to watch movies, take pictures and play games. Our cars, beyond gaining horsepower, have also gained an array of gadgets from bluetooth hands-free speaker phones to backup cameras.

Even housing, which is often subject to market forces that discourage innovation, have become far bigger and far better than the housing our grandparents and great-grandparents lived in a half century before. In the United States we’ve gone from living in less than 200 square feet per person in the late 1800’s to nearly 850 square feet per person of livable space today–850 square feet full of electric outlets, air conditioning, washing machines and dishwashers and flat screen TVs and stereos that in the early 1900’s would seem like magic.

Our third story: in a world of infinite wants, when given the ability to increase productivity, we often use that increase to make better and more sophisticated products, rather than create the same old products cheaper.

So we may pay the same price for the things we bought a decade before, but we get so much more value for the same price, as the same people who made simple products a decade before use automation to create more sophisticated products today.


This story that Bloomberg wants to tell, of a sort of nirvana where humans have been replaced by robots, completely misses the trends that have been taking place.

First, this is not a new story, nor is it a story about robots. It’s a story about the increases of productivity which have been going on since the stone age, which allow the average person to live in larger homes than the wealthiest families lived in just two centuries before, with greater access to everything from entertainment to medical care to knowledge than could possibly be imagined just half a century ago.

Second, it is a story about the thoughtful adoption of automation made only when the benefits justifies the cost–adoption which follows the same patterns that caused carpenters to replace their trusty hammers with nail guns and filmmakers to replace their trusty film cameras with digital cameras. If the cost makes no sense, berating and name calling (which the first Bloomberg tail talking about Infer, demanding that what their tool needs is a “hero CEO” and implying CEOs who don’t adopt their tool are backwards and untrusting) will not help your cause.

Third, it is a story about companies who, with the increased productivity, can either give us cheaper products or give us products that do more of what we want. And in that story we don’t see a world flooded with cheap underpowered cars and cheaper cell phones–but products that make us far happier as they do more of the things we want.

It is a shame Bloomberg (as well as many others) get caught up in this story about workers losing jobs to magical robots that can do anything, the story pushed by a Silicon Valley elite who have lost themselves in their own echo-chambers.

Because the real story is so much more interesting.

Advertisements