By Pat Dixon, PE, PMP
President of DPAS, (DPAS-INC.com)
I learned at an early age that I am not a skilled prognosticator. My ability to predict is more dependent on mathematical models than intuition. Despite this limitation, I will make a prediction in this article; human labor will not be entirely replaced by machines.
The pertinence of this matter is apparent as Industry 4.0 becomes a more common subject in popular culture. The concern that human workers are being replaced by machines and automation has led to proposals such as Universal Basic Income and other policies, which intend to ensure humans will continue to maintain subsistence without a means to generate income. The rise of Artificial Intelligence has raised well publicized concerns about the threat that intelligent machines present, such as the ongoing debate between Elon Musk and Mark Zuckerberg. Since our profession is one that provides and/or uses automation, we are inherently in the same debate and are prime decision makers in the outcome.
Among those in this domain, Ray Kurzweil stands out as a leading authority on technology and its impact on our world. In 2005 he wrote "The Singularity is Near: When Humans Transcend Biology". This book is recognized as a seminal work on this topic and is commonly referenced by those addressing potential threats of automation. In his book, Kurzweil predicts that by 2045 machines will become capable of the full range of human intelligence, which could mean humans would be obsolete as a labor source. This point in time is called "The Singularity".
This can sound rather scary, especially to those outside of our profession. How do we respond when frightened people ask us why we are building and installing automation systems for our production facilities?
A short answer is that these systems make us safer, cleaner, and more sustainable. I believe this can be clearly demonstrated, but such a response does not appease the fears of the future Singularity. Below, I will present a logical argument that may suffice.
I begin with a premise: our world has and always will have problems. We are imperfect, mistakes happen, and there need to be ways to avoid and fix mistakes. While nature may not make mistakes, it does present earthquakes, viruses, fires, and other calamities. Resources are finite and human demand for them can be competitive and lead to scarcity. In a wide variety of ways, our world is full of problems.
The logical corollary is that there will always be demand for solutions. There are opportunities for entrepreneurs to solve these problems and the market will always be open for business.
The next premise is that not every problem has a pre-defined solution. We can toss a boatload of date into a machine and get answers, but some problems don't have answers that can be derived from an algorithm. There are machine algorithms today that can adapt and improve, but they are generally used for a single purpose and not able to generalize intelligence. At the point of the Singularity, machines can learn as we do and figure things out without being programmed to solve a specific problem. In optimization there is the concept of local solutions. This is the case where an algorithm can achieve what it believes is an optimal solution, but the entire range of solutions has not been evaluated. In our multi-dimensional world, the solution space to some of our problems can seem intractable to evaluate. Therefore, we cannot assume the correct or best solution to a problem can be programmed. Whether human or robot, the problem solver needs to be creative. Being creative means to create; to make something new that did not exist before. Being able to go outside the box to consider different approaches that haven't been considered before is part of the creative process. Without this, the solution may elude the problem solver.
The final premise will be the most difficult one to prove and is critical to the argument. The premise is that humans are the ideal creative machine. In other words, if you wanted to create the most creative machine, you would have a baby. In "Algorithms to Live By", authors Christian and Griffiths show that many of the algorithms we use in industrial automation and elsewhere do not work until some human creativity is incorporated. Artificial Intelligence attempts to model our biological brains and capture our creativity, but as most models are asymptotic it may continually approach but never quite reach what it is attempting to model. We are the optimal creative machine.
To make the argument more concise:
- Corollary 1: There will always be a demand for solutions to problems
- Premise 1: There will always be problems to solve
- Premise 2: Some problems require creativity
- Premise 3: Humans are the ideal creative machine
Thesis (Corollary 1 + Premise 2 + Premise 3) = Machines will not entirely replace human labor
The empirical evidence shows that as technology has evolved, some skills have been displaced but other grow in demand and the overall unemployment does not grow. There was a time when everyone needed a horse. A horse provided transportation, power, and other needs. An industry was built around the horse; people produced wagons, saddles, horseshoes, and tamed these animals. Along came cars. Today there is still demand for skills in the horse industry, but it is a small fraction of what is was. That did not lead to mass unemployment. News industries required new skills and workers met that demand. Today, the demand for people in our field is huge. It is hard to imagine what the skillset of the future will be, but humans seem well suited to fill the demand. While Singularity machines will have a growing role and will likely offload much of our problem solving, history suggests the human role will be indispensable.
Taken together, the empirical evidence and logical argument suggest humans have a bright future in the labor force.
