Admit it! Everyone is talking about it and there are no doubts that AI is here to stay. So, what is gonna happen? First of all, calm down. AI is a great software achievement and at the same time, a massive opportunity.
Initially, let’s put aside all type of fantasies and science fiction. Overall, Artificial Intelligence is just code developed by people, yes humanoids! There are many cases where AI is making our lives much easier; self-driving cars, speech recognition, expert aviation systems, advanced healthcare decisions and so much more.
Apart from this, let’s focus on workforce. How will AI impact in future Supply Chain jobs? The following is a compilation of facts on why AI will not take your job …
▪ Time-saving does not mean work-saving
AI and Automation can amazingly optimize traditional Supply Chains. On the contrary, this does not mean less work. After all doing something faster it doesn’t implies to not have to do it, you’re just going to be time-efficient.
Additionally, an important part of Supply Chain workload is taken from ‘edge’ situations. Unexpected cases and scenarios which are hard to computarize and predict; even though if they were implemented, an expert decision-maker would be always needed.
▪ Supply Chain is not a standard business
Supply Chain is not a line of business. It’s a department or a dedicated section inside of an specific business. Compared with Aviation, Hi-tech and some others industries, Supply Chain standards are very poor. For instance, the way you measure Stockouts or Customer Service Level probably is not the same than your competitors, right?
This is a critical complexity. Supply Chains are well known for having thousand of different IT solutions in the market, but just a few of them (SAP, Oracle) use standard variables. Consequences? Harder implementations and almost impossible standarizations.
▪ Automation increase quality of time at work
Automation makes your life easier and it has a positive impact on jobs – who enjoys working on repititive and boring tasks? The average job tenure today is actually similar to how it was in 50s and 60s. Think about it. Evolution from Ford’s Mass production to the current Robo-assembly lines has incredibly maximize productivity! Nevertheless, thousands of workers and engineers are still needed.
▪ Machine control is only the tip of the icerberg
To put it more simply, “garbage in, garbage out”. High-quality data is mission-critical. Give the wrong information to algorithms comes up with useless outcomes. Clearly Supply Chains have still a lot of work to do with data quality. The majority of the time is spent in ‘cleansing’ operations and consolidating data to feed a predictive system.
– check out this post to understand the importance of cleansing in Forecasting
Let’s take the model of stock markets; if algorithms where intelligent enough to predict the next momentum, for sure their owners will be rich.
However, the reality is quite different. Stock algorithms are owned by giant corporations (which are the only capable to afford the costs of maintaince), they are helpful to avoid losses and make pushes against the market by generating thousands of orders per second. But still algorithms can not predict next economic crisis.
Many of the information the use needs to be ‘filtered’ and cleansed by humans … don’t forget that in the end, the next big market decision is going to be taken by humanoids boards and CEOs.
▪ AI do not understand humans
Machines will never be able to fully understand us. And yes, you perfectly know the reason. Paradoxically, computers do not have feelings, however somehow they try to process them. AI will never logically accept why we trust one person more than other, because based in facts and data analysis we should not trust any other human.
Just be sure that your AI is not in charge of your next Supplier selection or your next right-hand man. For sure, it will give you the most accurate ranking sort by hundreds of criterias, but the final decision should be yours.
Last but not least, I recommend you to have a look to the interesting Sascha Eder’s article about How Can We Eliminate Bias In Our Algorithms?