Inaction is a colossal risk in the industrial AI world
Industry is staring down a truly disruptive and transformative technology – industrial artificial intelligence (industrial AI). This innovation promises to solve numerous manufacturing challenges associated with automation, communication, training, research and more.
Understandably, some industry leaders highlight the risks associated with AI’s unpredictability, hallucinations and its black box nature. History has proven with other industrial technologies, however, that risks to adoption can become opportunity, while the risks of inaction can be existential.
Few of these examples show as many parallels towards the hesitation around industrial AI as the story of the simple ice cube. Here is how the ice trade played out and what that history teaches business leaders about industrial AI.
The rhyming history of the ice industry
Frederic Tudor started the “natural ice” industry in the early 1800’s.[1] After a few false starts, countless sceptics and three trips to debtor’s prison, the business eventually gave him riches and, more importantly, a global monopoly.
The company did not shy away from innovative technologies. By 1826, Tudor’s new foreman Nathaniel Wyeth optimized the harvesting of ice from wintery waterways using horses, plows, grids, conveyor belts and assembly lines.
But after Tudor’s death, in 1864, the company stagnated. By the 1880’s steam driven technologies emerged that froze ice at an industrial scale. Those running Tudor’s business were stubbornly against these technologies. As a result, new ice companies emerged and thrived as they embraced this mechanical means to make ice year-round.[2]
Champions of natural ice demonized “mechanical ice.” They claimed it melted faster and was of lower quality due to refrigerant contamination. But the dangers, costs and disease associated with harvesting natural ice proved more damaging. Mechanical ice prevailed.
Then, another shift, people started buying in-home refrigeration units. They could make ice themselves. In the face of this change, ice manufacturers refused to learn the lessons they taught the natural ice industry. Instead of modernizing, they started their own demonization campaign.
History repeated. Big ice declared home refrigeration as a menace, prone to accidents, leaks and other health effects.
Convivence won in the end; big ice was a puddle by the 1940’s.
What the ice cube’s past can teach about industrial AI
First, fighting disruptive technologies, like industrial AI, is an uphill battle. Companies that only hold onto the past miss opportunities to make their business better, faster or more reliable, convenient and affordable.
Another lesson is that innovative technologies do not need perfection. Like the hallucination and black box detractions of early AI, some early refrigeration complaints had more than an ounce of truth. Refrigerant leaks, and other risks, hurt people, the ozone layer and the environment. But as technology progressed, people solved the issues espoused by critics.
Today, home refrigeration has advanced so far that it is an afterthought. Industrial AI’s applications offer so much promise; it is easy to see it having a similar future.
Like with refrigeration, industry must adapt for industrial AI
Industrial AI will forever change industry. Just like electricity, the refrigeration cycle, assembly lines, automation and the internet.
Business leaders are right to see the risks associated with Industrial AI. Like the technologies listed above, early iterations will have flaws. In the wrong situations, AI can hallucinate, lie and fall into logic loops. AI can also, like many other technologies, be used to nefarious ends.
These risk assessments are valid and, in fact, necessary for industrial AI technology to grow. When risks force industry leaders to inaction, that is the issue.
The key is to turn risk into opportunity. Find ways to:
- Address those risks,
- Keep innovating and
- Make better products.
Natural ice leaders could have adopted refrigeration technologies. Use their money and expertise to improve mechanical ice faster and better than their start-up competitors.
Similarly, mechanical ice leaders knew more about refrigeration than most. They could have used that knowledge to improve miniaturization and solve the issues before the competition perfected in-home refrigeration.
The key is that they did not act in the face of disruptive technology.
Industrial AI technology is disruptive. It is not a fad. It is far too useful and convenient to disappear. Critics are right, however, early iterations come with risk. But just like refrigeration, industrial AI will get better. If industry leaders hesitate to find those solutions, then smaller, hungrier startups will.
Or as Bill Hahn, director of Solutions Consulting at Siemens digital Industries Software once put it: “What do you learn from inaction? Absolutely nothing.”
So, how can long-standing industries modernize with industrial AI? How do they get early buy-in and overcome the risks? For answers listen to the podcast: Industrial AI jeopardy: the truth about inaction.
[1] McRobbie L. R, “The Surprisingly Cool History of Ice.” Mental Floss, 10 Feb. 2016, The Surprisingly Cool History of Ice. Accessed 11 August 2026.
[2] Duffy, A & Ling, R. “Cold Comfort: Lessons for the Twenty-First-Century Newspaper Industry from the Twentieth-Century Ice Industry.” University of Michigan, Media industries. Volume 4, Issue 2, 2017, https://doi.org/10.3998/mij.15031809.0004.202. Accessed 11 August 2026.
