- 16th April 2018
- Posted by: Manolis
Supply chains are hugely important in our world of physical goods. Companies invest a lot of money and effort to continually optimize and improve them, and that’s not likely to ever stop. Save 5% on transportation costs? That’s a big deal.
As artificial intelligence begins to reshape our digital world via AI-infused goods, services, and processes, it will require its own supply chain: mind-boggling quantities of data. Mountains of information must be continually gathered, processed, refined, and reassembled to produce the desired intelligence and decision-making.
This is not a new insight; our prior fascination with big data and analytics taught us much the same thing. Only with AI, we’re talking orders of magnitude more information. Good, data-enriched AI will be infinitely preferable to poor, data-starved AI.
If we accept that the march toward an AI-enabled world is upon us, as business leaders we have an important decision to make: Do we construct our own digital supply chains to feed our AI, or be the consumer of someone else’s supply chain?
Like most things, the answer depends.
Table Stakes or Differentiator?
Let’s start with an application every organization uses—ERP, the software that manages such processes as accounting, procurement, project management, and manufacturing. These systems are the backbone of most organizations, and built-in AI functionality is poised to make a huge impact across many core business processes, like demand planning.
The good news is that there’s a great deal of commonality in any given industry, a certain set of modern best practices. People usually don’t go about creating unique, creative solutions for, say, order-to-cash. Cloud software vendors can exploit these similarities by offering easy-to-consume, AI-infused ERP applications, combining modest amounts of internal data with powerful models pre-trained on vast amounts of external data.
So when it comes to build versus buy for AI-enabled core business processes, the answer for many will likely be “buy.”
Now let’s shift the focus to how AI might make your organization unique in the eyes of your customers. Better, intelligence-infused products? Smarter and more personalized services?
We can learn something from our experience with mobile applications. In the first wave of such apps, they were nothing more than pretty skins over existing applications. We quickly learned that we had to reengineer the entire user experience to be successful.
We will likely see the same with AI-infused user experiences. If anyone can easily do a Siri-like voice interface, where’s the competitive advantage? We’ll need to get much more involved in sourcing relevant data, choosing the right algorithms, and presenting it all in a unique and compelling way.
But we won’t need to start with a blank piece of paper, thankfully.
Many aspects of the digital supply chain demanded for AI are available today in easy-to-consume form, thanks to cloud computing. Data enrichment as a service. Data pipeline automation. Self-training models. Conversational interfaces. Security and management. All ready to be put to work.
Yes, you’re building something unique, but doing so with a big head start. You’ll be adding your unique value, but not reinventing the wheel.
Technology is rapidly transforming the world around us. New technologies give us strong incentive to look at familiar problems in new ways.
I expect that just five years from now, many of us will care more about the digital supply chains that feed our AI than the physical ones that feed our warehouses.