“Ugh. That is so AI.” My 14-year-old daughter is decidedly not an artificial intelligence fan—not surprising since her social media feed serves up glossy, yet shallow, generative AI images and videos and her classmates use ChatGPT to cheat on class assignments. For her, calling something AI has now become shorthand for anything that is fake, vapid, and soulless.
We often talk about hype cycles around new technology. Sometimes the buzz about a new technology can reach a fevered pitch only to be followed by a similarly intense backlash as the dreams of what the tech can do meet the harsh reality of having to make it work in the real world. The analyst group Gartner calls this “the trough of disillusionment.”
Is there a clearer sign that a technology has hit the trough of disillusionment than when it becomes a teenage insult?
Based on the general public’s reaction to the large number of AI ads during the Super Bowl, my daughter is not alone in her derision of the technology. But is the increasingly strong backlash against AI seen in the general public and broader media reflective of how business is viewing the technology? Afterall using AI to calculate the best route for a shipment in real time has a far different feeling than using it to cheat on your math test or to create fake photos on social media.
According to Caleb Thomson, an analyst at Gartner who specializes in supply chain strategy and planning, supply chain managers have always had “a healthy level of skepticism” when it comes to AI. “And that skepticism may be growing somewhat,” he says.
In Thomson’s opinion, that skepticism is well deserved. To be clear, he says that there are an increasing number of fully mature AI-enabled supply chain solutions. But these solutions typically fall into the more traditional version of AI—what used to be more commonly known as advanced analytics or machine learning. Where these solutions work best is for improving decision quality for short-term, narrowly defined optimization opportunities, such as improving task scheduling in a production process.
Where things get a little dicey is when you move into the realm of what’s known as agentic AI—where an AI “agent” is supposed to be able to make a decision by itself that is as at least as good as a skilled human operator in a shorter time frame. “Outside of the most narrowly bounded decisions, that isn’t really a clear possibility for most organizations today,” he says.
Part of supply chain’s skepticism may also be a defense mechanism against some AI naivety and “fear of missing out” coming from the corner office. As Thomson points out, often those outside of supply chain can underappreciate the level of complexity and uncertainty inherent on a daily basis in all supply chains. And at one time or another, most of us have fallen for a technology vendor promising turnkey advanced capabilities that may not have been deployed in a single account—let alone one like your own. Thomson advises that if you suspect that your operation may not be ready for AI, there’s probably a good reason for your apprehension.
But that deserved skepticism doesn’t mean that we should reject the technology out of hand. Rather, start thinking ahead about how you can improve data quality and existing workflows and systems so that they are ready for AI when the time comes.
The message here? Supply chain managers should continue to do what they do best. Be realistic. Be practical. Engage with the technology—don’t just embrace it or reject it at face value. Ask questions. What can the technology do? What can it not do? What data do we need to have internally to do it? And what boring mundane things do we need to be doing now—like practicing good data hygiene or training our employees—to prepare for using AI in the future?
And if you need someone to push back on the AI naivety coming from the corner office, let me know, my teenager is always available for a dose of reality.
Source link
