"We will work together...unwillingly at first, on your part, but that will pass."
Thus spoke Colossus, the supercomputer at the heart of the 1970 dystopia movie "The Forbin Project." As a warning of what could happen if we give over complete control to The Machines, it was both overly confident that real artificial intelligence was just around the corner, and at the same time hugely prophetic in predicting the anxiety and uncertainty that increasing automation can engender.
The notion that we would eventually trust executive decision-making to an algorithm seems further off today than it did half a century ago; but the march of progress in technology has become a sprint, and the pace of change is now way beyond what even the futurologists thought possible in the golden age of science fiction.
Whether we are "only five years away from true machine intelligence" (as someone has claimed almost every year since the time of Alan Turing) or whether it will forever remain just out of reach is not really the point when it comes to the impact of our daily lives and the business processes with which we engage.
What is having a huge impact is the rise of automation, the removal of human intervention from transactional processes and the steady de-skilling of purchasing.
Whilst this has immense benefit in reducing admin overhead and eliminating cost at source, the seemingly inevitable off-shoot is a steady drive to automation of other steps in the chain that have hither to been the domain of the professional practitioner.
Should we be worried that catalog management, opportunity identification, sourcing and even negotiation might soon be possible in an entirely automated environment?
Perhaps not. Perhaps we should just remain confident that there are some areas of procurement specialism that are too complex for algorithms to handle, and so there is no chance of software taking our jobs.
But that is probably not the case in an empirical sense. Most procurement pros, if asked today whether software could do their jobs would say, "of course not." How, for example, can a sourcing software product evaluate qualitative responses? How could it weigh a written submission from one vendor against a slide deck from another?
Of course, right now, today, the answer is software cannot do this or any of the other activities that make us qualified to do our jobs. But that is not the point. We would probably have all had the same reaction if asked a decade ago if software could ever distinguish between a photograph of questionable content and one that is perfectly innocent, but today those capabilities exist and whilst not fully, 100% accurate, that is only a matter of time.
Ultimately the algorithm is a method of expressing a process that we can conduct with a brain in terms that a machine can emulate. There is no reason to suppose that this puts anything effectively off-limits given time, effort of will and processing power.
It will not take artificial intelligence in the sci-fi sense to remove the need for humans in the business process, but simply the processing power to comprehend and understand the patterns involved.
If there is no technical reason why automation cannot take over, does that make it inevitable? Should we all be looking for the nearest exit?
Exit? No. I don’t think so. But perhaps a different seat at the table.
Automation is good. Reducing overhead, shortening cycle times and accelerating processes are all necessary in the drive for greater value, and the programs we instigate to enable all these are born out of experience and innovation.
Right now, and tomorrow, that is something that is likely to remain the human edge. It will take the birth of true Colossus-like AI to steal that skill from our toolkit, and depending on who you ask that is still five years away!
Innovation, then, is our reason for remaining relevant. No matter how sophisticated the algorithms, they will only be able to very remotely simulate flexible, creative thinking.
Imagine the situation where both buyer and supplier algorithms are conducting sourcing negotiations. The bidding will be over in a flash and each time the results will always be the same. Where then is the benefit to the business?
The story goes that when developing the CGI techniques for the Lord of the Rings movies, the programmers created algorithms that allowed the thousands of warrior characters to behave in naturalistic ways with tendencies to attack or defend according to circumstances.
The programmers, perhaps, did their jobs a little too well. In the first trial, simulations of the massed ranks of orcs and whatever, when spotting the enemy, all turned tail and fled the battleground. The developers needed to then add modifiers instilling willful disregard for survival in order to have the battle that was required.
Algorithms, then, will do what they do, but flexibility and innovation is hard to code.
One thing is clear: the procurement operation of tomorrow will need to be increasingly flexible and creative to drive value in a rapidly changing and even increasingly unstable world.
The degree of automation that could be possible is unlikely to be what our enterprises actually need to navigate ever more tricky waters. Of course, it will have a role to play in removing the need for professionals to conduct or monitor repetitive transactional tasks and even some of the things we take for granted today as our specialties may pass over to the machines, but not everything, not by a long way.
In fact, there will be a risk in taking automation too far. Giving the entire process over to the software not only removes innovation from the picture but will remove the opportunity for innovation.
Procurement is, has always been and will remain a field that succeeds through human intelligence and creativity.
Thus, in the spirit of "just because you can do something doesn't mean you should" automation will have a role to play, a vital one, but it will not be the answer, the solution or the end of the procurement profession. Unless, that is, we permit it to be so.
About the Author: Paul Blake, Senior Manager, Technology Product Marketing Paul Blake leads the technology product marketing team at GEP and is responsible for the promotion of GEP’s procurement technology platform, SMART by GEP. He has worked in technology for over 25 years and has been involved in the development of procurement software solutions since before the advent of the internet. Having run the development team that delivered Europe’s first B2B trading platform in the 1990s and implemented eCommerce networks on three continents in the early 2000s, he has long-term experience of the trends and developments that affect procurement professionals.