Human-Machine Relations

Updated: Aug 10

There is currently much debate surrounding the relationship between humans and smart systems. Such questions span the full spectrum from practical considerations surrounding work displacement and employment to existential threats to humanity. Indeed, it is important to recognize that at the end of the day the success of this technology revolution is ultimately almost completely dependent upon how human and information systems can work together. The introduction of this new technology that can perform many of the tasks previously the exclusive purview of humans creates a disruptive and sometimes difficult adjustment of human activity, both on a practical level of work processes but also on a cultural level. A prolonged process of readjustment and realignment is certainly ahead of us, critical to this process though is understanding the limits of computation, figuring out what computers are best at and what humans are more adapted to and how to create synergies between the two to achieve optimal overall outcomes. It is decades now since the world’s best chess master was beaten by a computer, but today the world’s best chess master is not a computer but instead a combination of person and computer working together. This will be the same for individuals as for organizations and whole societies, whichever can figure out how to harness the unique capacities of each and coordinate them in a synergistic fashion will be the winners in the information age.

Key to understanding this differentiation between people and machines is understanding the difference between analytical and synthetic reasoning. Analytical reasoning is what computers do. Analysis is a method of inquiry that proceeds by breaking a system down into its elementary parts, studying those parts in isolation and then forming a description of the whole system in terms of its basic components and interactions. Analytical reasoning deals with closed systems, it involves scrutinizing a situation in order to break it down and solve the associated sub-problems. Analytical processes operate on a prespecified problems space and solve it by defining a set of steps or algorithm. Computers will become masters of analytical reasoning. There is no barrier stopping algorithms from expanding to all areas of analytical reasoning and in this respect within just a couple of decades, there will be almost nothing that we can compete with them at.

The barrier though is with synthetic reasoning. Synthesis is the combining of two or more things in a specific way to create something that is new and different. We take two discrete things and because of the specific synergistic way we arrange them they become integrated into something new which is continuous. Synthetic reasoning is the essence of creative thinking and innovation. Analytics acts on a pre-defined problems space but it can not create the problem space itself. Analytical reasoning is effective at studying what is, while synthetic reasoning is effective at understanding context so as to create new problem spaces. As Prof. Joel Mokyr of Northwestern University puts it “it is the humans who ask the questions, it is very hard to teach machines to tell them what are the interesting questions we need an answer to… once we pose the question they can help us answer it.”


Artificial intelligence, machine learning, advanced analytics these are all information technologies. Technologies can be understood as automated systems for resolving a given constraint. All technologies are algorithms in their essence, they are designed to find the most efficient way to map a given input to a desired output. They are an automated means to an end. Technology can only ever be a means to an end, technology can never be an end in itself. As the futurist Gerd Leonhard states it “technology is not what we seek, but how we seek.” Because technology is a means to an end it is always doing something, every technology does something. In contrast, being is not doing something it is the creation of a possibility. Technologies are only ever means to an end, they can’t create their own ends. To understand the ends of something we have to ask the why question, why do we want to be able to do something. The why question is always an expansive question, if you stay asking why, it will lead you into dealing with broader and broader context, which requires synthetic reasoning, it leads you to dealing with poorly defined categories and concepts so as to be able to grasp the whole, which is precisely what computers are bad at.

Synthetic reasoning looks at the system within the context of its environment making it possible to redefine the system and the problem space. This gives paradigm shifts and quantum leaps in results. It is the job of people to define the ends and from that the problem space. The ends are given by the context or the environment and the environment is understood through synthetic reasoning. As technologies will get better at resolving problems spaces, people will have to get better at being, creating the future possibilities that machines then execute on. It requires humans to move up the value chain to become more what we are in essence. To better understand this distinction between analytic and synthetic reasoning and the distinction between means and ends.

As an example we can think about this question, why can we successfully apply data analytics to sports matches or to a production line but not to going backpacking across Africa or to our relationship with our partner? The answer lies in the fact that in the first instance we can define the desired output to the system, we want to win the match or we want our production line to produce more widgets per hour and the process is just a means for achieving that, thus we can apply an algorithm for optimizing it. In the second set of examples, we can’t really define an output to the system, in fact, the process is the output, it is not a process that creates a desired output, but more like a process out of which emerges some form of value. We could come up with an algorithm for the most efficient way to backpack across Africa or to interact with your partner, but in fact, the very act of optimizing the process depletes the value of it. Unlike in the first examples where the more efficiently we can complete the process the better the overall outcome, with the second examples the more you do the process the more value you get. The more deviation from the optimal path to get across Africa the richer your traveling experience will be, the more your relationship deviates from some predefined algorithm the more valuable the relationship will be.

Existential Threat

Throughout civilization, we have use objects and technologies as ways of objectifying our experience so as to escape from the inherent challenges of being. We do this now again with this new technology, instead of taking responsibility for our decisions we defer to an app, instead of living with uncertainty we resort to statistics. It is this that is the real risk and what creates the existential threat, by deferring our humanity to an object we give over control to that thing. Whoever then controls the technology will control us, that’s been the case since the origins of civilization, what is different now is that technology may end up being autonomous, and we collectively defer our humanity to these systems then we become control by them. That is a systemic existential threat, there is no silver bullet simple solution to it. There is no algorithm you can write, no master button to press if things get out of control.  These are the risks we are taking that will be defined by the choices we make individually and collectively in our every interaction with these systems of technology. We are on a journey the opportunities are immense but the risks are just as high.

Just as with the current environmental crisis, where we did not really appreciate our natural environment until it was degraded and now have the opportunity to re-engage with it in new ways with a new level of appreciation, the same may be true of our humanity, the development of this technology will create many negative social and cultural externalities – like that of the combustion engine during the industrial age – which will both degrade our culture but also offer us new ways to appreciate it and re-engage with who we are and what is of value. The risk is not machines thinking like people, but quite the opposite people thinking like machines. Unfortunately, this is exactly what we wanted of people during the Industrial Age and have built a whole educational system, even a whole culture around this, it is the inertia in that system to change that sets people up to fail in this new economy. The opportunities of this ongoing revolution in information are immense, but so too are the perils. Smart systems represent an expected evolution of our technology landscape one that is in many ways required. However, for this current stage in our technology development to be sustainable this expansion in technological means will require a concomitant expansion in human ends and it is yet to be defined whether such a counterbalancing force can or will emerge.



Systems Innovation

  • LinkedIn
  • YouTube
  • Twitter
  • Facebook