White Collar Job Replacement

RobotIn a survey late last year Gartner Research asked numerous CEOs questions about how automation might affect their companies and workers. Sixty percent of those thought it is a fantasy to think that computers will replace middle class jobs, even within a fifteen year window. Their reaction is not surprising in that it’s a normal human reaction to think that what we do cannot be replaced by machines.

But a lot of my reading suggests that these CEOs are probably wrong. I have often written in this blog how technology advances have been accelerating. Everybody is familiar with Moore’s law that talks about the doubling of integrated computing power roughly every eighteen months. And just a few weeks ago I talked about the speed at which the cost of using data centers has dropped. But these same kind of improvements can be seen in many of the other critical building blocks of computing from processing speeds, memory storage capacity, energy efficiency, download speeds, etc., which have all been improving at exponential rates for some time.

There are a few other factors that will also be pushing job automation. I recently wrote about how supercomputing is getting much cheaper and is also poised to make further breakthroughs in affordability and power with the advent of desktop-sized supercomputing. Plus IBM and others have been making huge strides in plain language interface with computers like Watson and other computer systems. We are also on the verge of a huge expansion of the digital network through the deployment of billions of sensors spread throughout our environment as part of the Internet of Things.

We tend to forget that the new digital age is still relatively new. I remember buying my first home computer, a TRS-80 in the late 70s. I remember getting on-line with CompuServe and other services in the early 80s. I don’t think I was able to get off dial-up access until the mid-90s. We have come a very long way in a very short period of time. Some major industries have been completely killed by the digital age, such as Eastman Kodak, and much of the traditional music industry.

But there is still a huge amount of legacy thinking in the world. Companies tend to keep doing things the way they have always done them until they are forced to do otherwise. There are already a number of jobs that computers could replace if companies really wanted to do so

I have written before about my friend Danny who owns a traditional CPA firm. His clients are almost all ethnic and his business is often conducted in something other than English. But he offers the traditional accounting functions – processing payroll, keeping books, preparing taxes – that all accounting firms do.

Danny understands the power of computers and the cloud and he has already transformed his business. For example, he has created software that prepares accounting summaries directly from his clients’ bank statements. This has eliminated a lot of manual effort and also eliminated human error in transcribing what his clients spend. And he says he will soon be able to prepare draft ledgers and tax returns from these records without need for human intervention.

He knows that some human action is needed in the accounting process, and he refers to this as the customer service function. His clients want to know that somebody they trust is looking out for them to make sure that they get the best deal with taxes. He is always going to need accountants to do that function. And it’s likely that he is going to need accountants for many decades to deal with the IRS and other taxing authorities since they seem to always be a few decades behind with their computer systems

Danny is in the process of transforming an accounting office so that a handful of accountants can service thousands of customers. Since he is far ahead of his competition he is going to be both more profitable and more competitive than his peers.

And Danny is just one small business in one tiny niche of one industry. He is headed towards a more efficient future because he understands why computerization is inevitable and he wants to be out in front of the curve. All other CPA offices will eventually be dragged to this same place or they will perish due to being too expensive and inefficient. One of the biggest consequence of the automation of accounting functions will be the need for fewer accountants and one can expect nationwide for there to eventually be considerably fewer accountants. This same trend could eliminate white collar jobs in scores of other industries, wherever those jobs deal with performing repetitive tasks.

The truth is that that even today there are very few companies around who should not, or could not be undergoing the same transition as Danny. So I think the CEOs who don’t foresee middle class job automation within fifteen years are wrong. They might be foretelling a future where their own companies don’t take the lead on such automation. But if they are not first, then the forty percent of CEOs who see this coming are likely to leave these companies in the dust.

4 thoughts on “White Collar Job Replacement

  1. Doug as you well know, there are so many incredible opportunities available to us in this industry right now and far more than at any other time or place in the past. The billions of dollars spent on maintaining the status quo is now being stretched so hard that it can’t hold back technology ready to spring forth and crush and rebuild the current business model.

    The basic human psychology of peer driven motivation is and will always be the same. People still love the same Roman Circus from thousands of years ago that now shows up as blinking lights on a screen. The only thing we’re fighting over is how do we get those blinking lights to the people that gives them most perceived pleasure and joy from the customers viewpoint wins.

    In the future maybe the smallest operator in the business will be equal (better make that better than) the largest operator by being…

  2. Most people, including executives, don’t understand the amazing power behind exponential growth, when things get significantly better every year or every eighteen months. If cloud computing costs are cutting in half every eighteen months, then in three years they are a fourth as expensive as today and in six years they are a sixteenth as expensive.

    It’s one thing to make a judgment that something is not economically feasible today, but in a world of Moore’s law and the related other changes in technology, it might be 64 times better nine year from now either in affordability or in terms of the raw processing power that is available for the same price.

    One of the first jobs I ever had was working at Control Data and we were the second maker at the time in supercomputers along with Cray. What people fail to recognize is that there is as much computational power in their smartphone as there was in the Cray-1 Supercomputer of the 70s. And we are not at the end game and the technology is just going to keep marching forward.

    At this point all it really is going to take to transform jobs is supercomputing that operates in plain English. IBM is already leasing out Watson by the hour and in a decade that same power is going to be available to companies on a desktop inside their business. Watson not only beat the world’s best Jeopardy players, but is better than any doctor in diagnosing diseases. When a machine can diagnose diseases there are a whole lot of other white color jobs that it is going to be able easily do as well.

    And you are right. There will be small nimble entrepreneurs that probably will take advantage of this before the big companies. And those might become our big companies of the future.

  3. Doug, the TRS-80 sure brings back memories. I picked up one of the earliest TRS-80 model 4P’s. The P was for portable – yeah right! All in all, it was a good machine and taught me many things about computers, compilers and machine code…

Leave a Reply