As a supply chain professional, it can be difficult to cut through the noise surrounding artificial intelligence. The hype around ChatGPT puts leaders under pressure to incorporate more AI into their setup than ever before. But they must remain cautious - there is no point incorporating AI just for the sake of AI. The purpose is to drive tangibly better business outcomes.
ChatGPT represents a new accessible era of AI, with a human-like ability to converse with people. Not surprisingly, vendors across the board are rushing to augment user experiences with generative AI because the barrier to entry is low, users like it, and the marketing benefits are high. Commoditized tools make it easy to spin up a prototype in a matter of days or a few weeks. What was once “a breakthrough” is quickly becoming table stakes, and before long, conversational user interfaces will be commonplace.
That said, AI is and will continue to be a major differentiator that separates winners from losers in the coming years – just not in the way that one would presume, given all the current hype.
ChatGPT’s ability to answer any imaginable online question has captured our imagination and gives the appearance of the ability to reason. However, it’s important to recognize this does not make it intelligent. Rather, it is a master at correlation, processing masses of data, and giving answers with a high mathematical probability of being right.
Furthermore, generative AI is just one of four artificial intelligence methods available alongside supervised, unsupervised, and reinforcement machine learning. Unlike their flashy ChatGPT cousin, these traditional AI methods are the quiet workhorses of modern supply chain management, working in the background to create new value beyond what people can do.
Navigating the AI landscape – one size does not fit all
When it comes to the different types of AI, one size does not fit all. It’s not that one method is better than another – but about using the right tool for the job at hand. Just as you wouldn’t use a hammer on a bolt or a spanner on a screw, it’s important leaders see through the hype to choose the right AI method(s) for each business challenge.
So, what are the different types of AI technologies, how can supply chain leaders deploy these effectively, and where should they prioritize now?
Supervised learning finds patterns across disparate datasets. These systems are trained to recognize what “good” looks like and learn over time to accurately predict outcomes. Unsupervised learning discovers hidden clusters – many of which are not obvious to people – without any training or guidance. Reinforcement learning explores different options through trial and error, learning which actions to take based on the best outcome. Finally, generative AI uses large language models to interpret masses of unstructured data and generate new content with similar characteristics, often featuring the human chat-like interaction that made ChatGPT famous.
Quiet AI workhorses of supply chain
For decades, supervised and unsupervised AI have been deployed at scale within advanced supply chain management applications in functional domains of planning, logistics, and channel management. Forecasting finished goods, supervised AI recognizes patterns across many real-time demand signals to predict what customers will actually order instead of simply what you hope they will order. This lets companies build the right products and stock them in the right location at the right time, the first time, to better serve clients at the lowest cost and capture growth opportunities.
To support new product launches, unsupervised AI automatically identifies items with similar characteristics and inherits their properties to ensure availability, capture market share, and raise innovation returns while limiting financial liability.
In channel data management, supervised and unsupervised learning cleanse dataflows from thousands of distributors, resellers, and retailers to automatically correct errors and augment missing fields, transforming noisy data into decision-grade information.
In logistics, supervised learning finds patterns hidden in a myriad of transport data, such as routes, loads, and equipment types, to predict and compare freight rates to industry averages. AI is also used to predict future transportation capacity requirements by lane and mode, allowing shippers to proactively identify gaps in capacity and secure transport with preferred carriers at the lowest cost.
In each example, traditional AI runs quietly in the background, doing its job without human intervention or fanfare, unlocking new value beyond what people can do today. These are the unsung AI workhorses of supply chain. It takes all four methods to build a successful AI strategy, so don’t let the conversational allure of generative AI lull you into a one-size-fits-all approach.
AI is meaningless without the right data
Two other prerequisites for success are data and an embedded AI strategy.
Simply put, it doesn’t matter how good your AI is if you don’t have the right data. AI requires data and lots of it. Not just any data but information relevant to the decisions and actions being taken. For making, moving, or selling goods, the most valuable data comes from your extended supply chain. These contain demand and supply signals that influence growth opportunities or risks – such as shifting consumer behavior, materials constraints, logistics capacity shortages, and changing trade regulations.
This includes data from your internal operations and along the entire value chain of multiple tiers of partners, suppliers, distribution, transportation, customs, and border crossings. Gaining access to this information relies on a network that can connect the dots and lower the barriers of entry, getting leaders and AI systems the data they need to make better decisions. The more data points you have aggregated, the better any AI solution will perform.
Embed in day-to-day decisions to run the business
In recent years, AI has become a commodity, so unlocking value is no longer about having the best technology but how it’s applied to the business challenge. Specific to supply chain, the greatest value comes from embedding AI within the core tools used to make day-to-day decisions to make, move, and sell goods. Otherwise, it ends up as an overlay and more of an afterthought – instead of driving behavior at the time of decision.
“Yes, and” approach to success
In short, embrace the conversational aspects of generative AI, but don’t stop there. Build a strategy that considers all four methods and uses the right tools for the right job. Focus on data, data, data – especially from your value chain – and embed AI in your day-to-day decisions. AI will revolutionize supply chain management; leaders just need to know where and how to use it to unlock new value for their organization.
