feature article
Subscribe Now


The Real Rise of the Machines

We knew this was coming all along.

From Mary Shelley’s “Frankenstein” to Isaac Asimov’s “I, Robot”, to James Cameron’s “The Terminator” – our science fiction writers have been warning us for two centuries that our engineering masterpieces would one day rise up to defeat us. With no emotional capacity for remorse or regret, no human souls to temper their relentless pursuit of their goals, and no fear to give them pause, these perfect villains would multi-mindedly put an end to our nice little party where humans were the top of Earth’s food chain. 

The real revolution has already started, and the humans are losing. But the science fiction writers got it terribly wrong. The machines didn’t come at us with lasers, guns, or weapons of mass destruction. They’re not patrolling the streets mowing us down one at a time while we make a frantic and futile run for cover. No, our machines are much smarter than that. They’re systematically de-constructing our society quietly, from within, and we’ve barely even noticed. They’re hitting us, not with bullets, bombs, or energy-beams, but with pinkslips. Technology is slowly and stealthily putting humans out of work.

And it’s all our fault.

If we’re doing our engineering jobs right, technology is steadily making humans more efficient. With the aid of the things we engineers are building, humans need fewer man-hours to accomplish the same work. Or – with the same man-hours, we can do much more work. The choice is ours. The consequence of the former, however, is that people don’t have to work as hard. Sounds sweet, doesn’t it? Instead of slaving away at desks for upwards of forty hours per week, people can simply show up Monday morning, look over the previous week’s results, dispatch their personal work robots with the week’s assignments, and head to the beach until the following Monday. It’s like being on vacation all year round, or – to put it another way – “unemployed”. 

In a recent AP article, Bernard Condon and Paul Wiseman make a sobering case that technology is wiping out an entire range of middle-class jobs in the service sector.

Condon and Wiseman argue – with lots of convincing economic data – that millions of service-sector jobs, from travel agents to secretaries to accountants – have been eliminated by technology. And they’re never coming back. And it’s just going to get worse.

It’s always been popular knowledge that manual labor jobs like manufacturing line work were at risk for robotic replacement. Robots are cheaper, more reliable, and faster than their human counterparts. They never call in sick, get unhappy with benefits or work conditions, or sue you for job-related injuries. Manufacturing, we understood, suffers a double-whammy. Technology replaces some of the workers, and the human work that remains is outsourced to the developing country with the lowest labor costs. We all understood this – and maybe comforted ourselves with the idea that it would all eventually equalize itself with the transition to a global economy and equalized standard-of-living expectations around the world. 

What we weren’t talking or thinking much about was the replacement of employees paid for thinking. But, with the latest generation of technology, that is exactly what is happening. Despite college degrees, many middle-class white-collar employees have been put out of work, just because the world works more efficiently now. If these are smart, resourceful, ambitious people, the stock solution would be for them to “retrain” for new careers. But, retrain for what?

When robots and computers can do everything, the only jobs left will be designing and programming robots and computers. 

As engineers, we might take some comfort in the prediction that our occupations are the cockroaches of the technocalypse – the only ones that will be left standing after the devastating wave of increased productivity sweeps the Earth, obliterates the global economy, and sends the rest of humanity to stand in breadlines manned by robotic do-gooders.

It won’t go down that way, of course. 

The economic equation has a feedback loop (Engineering firing squad ready? Aim your differential equations… Fire!) Obviously, for the fully-automated, zero-employee widgetron factory to stay in business, it needs customers. For now, at least, those customers have to be people – people with purchasing power. That means those people need jobs to earn the money to buy their widgetrons. If humanity as a whole is underemployed, widgetrons will have to get cheaper. That’s OK, because widgetrons don’t cost much to build – not in terms of man-hours, at least. How much cheaper will they get? Almost free, as it turns out…

We already have a huge real-world example of this effect. Software, and in the extreme case – mobile apps – have been in this economic wormhole for several years now. Software is one example of the fully-automated factory. It takes the same number of man-hours to produce, manufacture, and distribute a mobile phone app to one person as it does to deliver that same capability to a billion people. Incremental unit cost is basically zero. The entire cost is non-recurring engineering (NRE). As a result, you can buy a mobile phone app that performs a function that would have once required equipment costing hundreds of dollars – for 99 cents. Now, instead of that capability being available only to people with the resources to invest hundreds in the stand-alone device, that capability is available to anyone who can afford a cup of coffee. Economically speaking, it’s deflation in the extreme. 

This is good news, right? Instead of advanced technology being the purview of only the rich and privileged, it can be available to the masses – at a price they can afford. Take the example of smartphones. Because of the massive economy of scale required to make something as advanced as a smartphone in the first place, we have to sell millions of copies of each one in order to come close to break-even. That means we have to make them cheap. Since we made them cheap – just about everybody in the civilized world can afford one. As a result, a remarkable thing has happened. Unlike just about every previous type of product that has existed, the wealthiest people in the world are walking around using exactly the same kind of smartphone that you see in the hands of starving students. On the consumer side, the economics of automated mass-production of extreme technology can virtually erase social class barriers.

What about the supply side, though? All those humans still need something to do with themselves for forty-some hours per week, right? For all of us engineers, it’s easy. Go out and design more computers and robots. Other occupations will fall at varying rates – the number of people required to supervise the robots will gradually decrease as the robots require less supervision. The waters of unemployment will rise – systematically swamping ground with ever-higher levels of human education and expertise. Want to stay dry? Educate yourself and create – every day of your life. Robots and computers will not replace human creativity any time soon. Engineers, artists, musicians – people whose occupations begin with some unique understanding and perspective on the human condition – will remain employed. 

Looking at things another way, there are two categories of jobs – those that scale employment with the number of people served, and those that do not. The number of hairdressers required to dress the hair of a million people is a thousand times more than are required to dress the hair of a thousand people. Hairdressing employment scales with the size of the served market.

Jobs like engineering do not scale employment. Designing the clippers to clip the hair of a million people requires the same number of engineers as designing clippers to clip the hair of a thousand… or one. If that engineer can now create clippers that analyze the face of the customer, apply the perfect hairstyle, and eliminate the hairdresser from the equation altogether, he has created something with the value of all the hairdressers it replaced. If he can then engineer those clippers to be so inexpensive that anyone could afford them, he’s taken the whole hairdressing problem off the table. Thanks to him – all humans now have dressed hair. We can move on to the next burning social issue.

(OK, sorry hairdressers. That was a joke, and I acknowledge that hairdressing is an art – and thus not subject to robot replacement under the terms of my earlier paragraph.) However, this does bring up an important point. Technology does not necessarily solve problems in the order of our hierarchy of needs. If it did, food, housing, and healthcare could be downloaded for 99 cents per month and smartphones would be expensive and reserved for the rich. If engineers could magically solve problems in their order of importance, the social picture would be much simpler. For example, in the Western world, we’ve just about got the water problem solved (for now, anyway.) Anybody with the resources to afford shelter also can afford water. In many places, drinking water is available for free on the street in fountains. If other basic needs could follow in the footsteps of water, we’d be debating the flavor of frosting on the cake here.

Yes, technology eliminates jobs. The goal of technology is to make people’s lives easier and better. As engineers, we want to help people achieve the same goals with less work and discomfort. We want to empower humans to accomplish more than ever – with greater safety, less effort, and less discomfort. Because of the inherent feedback in the system, the economic issues will work themselves out. We’ve seen this effect time and time again. The blacksmith profession was annihilated by the rise of the automobile. Humanity adapted and survived. It turns out that shoeing horses wasn’t really all that fun or fulfilling anyway. With the transition will come pain, of course. Evolutionary change in any complex system takes its toll. Those who are most averse to change will suffer the most from that evolution. Those who create that change stand to gain the most. 

Be part of that change. 

Keep on engineering.

6 thoughts on “Technocalypse!”

  1. Is technology the harbinger of doom, or the gateway to paradise? As engineers, we work to solve problems in new and inventive ways, but does incrementally solving problems make the world better?

  2. The more things change, the more they stay the same. (I cannot accurately remember the French quotation.)

    We have gone through all this before. In 1900, 50+% of the jobs in the country were on the farm. Automation in the form of the tractor and farm machinery changed this. By 1999, Less than 4% of the jobs were associated with farming, with only 2% of the jobs actually on the farm. And we are making so much more food today that we pay farmers to keep land out of production.

    One of the components of the Great Depression was the shift jobs from the farm to … something else: the factory, in this case. Dire predictions about jobs being lost forever were made. Instead, we created more and better paying jobs, leading to a large middle class.

    In 1950, something like 39% of the jobs were in manufacturing. Today it is 12% and falling. Automated machinery – NC controlled lathes and mills, computer controlled assembly lines and robots – has displaced and is displacing the original workers.

    Again, labor intensive factory jobs are displaced by machines. Likewise, labor intensive office jobs – typists, file clerks, etc. – are displaced by computers.

    Many old jobs be taken by machines. Will all the jobs be taken? Only if humans are ever satisfied with what they have. Answer: not likely! There is always a higher shelf in the candy store. Humans always want more and better than they have or grew up with. Given food and shelter, we want experiences, entertainment, knowledge, increased social contact, and on forever.

    This is the driver of the ongoing shift to new jobs that require human knowledge: finding a need and filling it, meaning people-to-people sales, personal service and invention to serve local needs and desires.

    It ultimately takes humans to understand human needs and desires, and to find new ways to satisfy them. Remember: machines want nothing. All value comes down to the value of human time, the time I spend to get what I want or the time someone spends to help me so I do not have to spend my own time getting what I want.

    We will go through uncomfortable times – again – learning how to work in the new world, where some things that once were expensive are now as cheap. But work we will. Because our reach exceeds our grasp, and we want more.

  3. An interesting issue that is very quickly becoming more current! However, I don’t see why the engineering job should be more protected than other educated jobs. What would prevent the machines running from programs and blueprints written by machines? The only “job” left would be that of the consumer, asking for what they want.

    The big difference in computer automation vs factory automation is that computer automation can potentially replace all jobs. Not just the existing jobs, but also all the new jobs we come up with once we are freed from the current jobs. This is a complete game-changer compared to what happened in the industrial revolution. It essentially says that there really will be no more jobs in the future.

    We might (but only perhaps) still have the arts, but those have never payed very well except for a small minute fraction of the performers; the rest have had to keep their day job.

    For those of you who doubt that a computer will ever do everything that a human can, I suggest you take a look at the Human Brain Project. This is a ten-year research project, funded by the EU to the tune of a billion Euros! The goal is to build a complete simulation of a human brain; not just a neural network, but all the brain chemistry as well (this is why even the arts might be a threatened profession). An explicit goal for the simulation is to have enough accuracy to use it for testing new medical drugs and their effect on the mind. At the start, it will require one of the top supercomputers in the world, but the Moore’s Law lag between the top supercomputer and a standard desktop is only about 15 years.

    There is another important issue related to automation: Automation is not free. To automate, you need to buy the machine that replaces the human worker. Essentially, you are replacing the salary cost with a financial cost, and as long as the financial cost is lower than the salary cost, it is a good deal. The cost is still there, nevertheless, so the product is not free. However, since no one is paying salaries any longer, just interest on loans, there will be no consumer that can afford the product (apart from bankers).

  4. @Stefan, You have arrived at the feedback loop I described. If there is no consumer that can afford the product, there is no need to create the machine that can make it. The control mechanism should kick in before that. Dwyland says everything comes down to the value of human time, which is correct except for the concept of ownership. The person who owns the goose that lays the golden egg is given immense control – but that hinges upon gold having value and ownership being an important cultural concept.

  5. I am not sure if the feedback loop will necessarily stabilize things. Admittedly, still heavily idealized, but at least adding the factor of multiple companies, you get this: Some companies automate first, others later, depending on the quality of leadership. The ones that automate first will have no employees themselves, but will have the employees of other companies to sell to. The ones that automate later will make less money and in the end go bankrupt. Since none of the surviving companies want to re-hire expensive humans once they have gotten rid of them, there is no obvious way out of the hole.

Leave a Reply

featured blogs
Aug 16, 2018
Learn about the challenges and solutions for integrating and verification PCIe(r) Gen4 into an Arm-Based Server SoC. Listen to this relatively short webinar by Arm and Cadence, as they describe the collaboration and results, including methodology and technology for speeding i...
Aug 16, 2018
All of the little details were squared up when the check-plots came out for "final" review. Those same preliminary files were shared with the fab and assembly units and, of course, the vendors have c...
Aug 15, 2018
VITA 57.4 FMC+ Standard As an ANSI/VITA member, Samtec supports the release of the new ANSI/VITA 57.4-2018 FPGA Mezzanine Card Plus Standard. VITA 57.4, also referred to as FMC+, expands upon the I/O capabilities defined in ANSI/VITA 57.1 FMC by adding two new connectors that...
Aug 14, 2018
I worked at HP in Ft. Collins, Colorado back in the 1970s. It was a heady experience. We were designing and building early, pre-PC desktop computers and we owned the market back then. The division I worked for eventually migrated to 32-bit workstations, chased from the deskto...
Jul 30, 2018
As discussed in part 1 of this blog post, each instance of an Achronix Speedcore eFPGA in your ASIC or SoC design must be configured after the system powers up because Speedcore eFPGAs employ nonvolatile SRAM technology to store its configuration bits. The time required to pr...