Legacy manufacturers are lagging behind in AI. They have plenty of tools to choose from, but they’re held back by employees, culture, and a lack of pressure to act.  

In this episode of The Disruption Is Now, host Greg Matusky talks to Ed Marsh, a manufacturing consultant and host of the Industrial Growth Institute podcast, about how manufacturers can shift the tide toward AI adoption. 

From lone sales reps experimenting with AI to the idea of adding a non-voting AI “board member,” this conversation digs into the real barriers and opportunities for AI in manufacturing.

Watch now: 

Key takeaways: 

 

AI adoption is wildly uneven, even within the same company

AI maturity doesn’t vary by company, but by individual. 

“I’m seeing that wide range of maturity within companies,” said Marsh. “So in other words, one salesperson or one marketing person might be adept and agile and committed to finding use cases and many of their colleagues may be dismissive or even mock them about doing it.”

Unlike past tech rollouts like CRM systems, which were implemented top-down across entire organizations, AI adoption is happening ad hoc, driven by curiosity, not corporate mandates. The result is a patchwork of maturity, with a few super‑users quietly building skills and many others stuck in old habits.

That inconsistency makes internal communication and change management much harder — and much more important.

Manufacturers should try to build AI tools before they buy

Most companies assume they need to buy off-the-shelf AI tools. But Marsh argues that these tools hide assumptions behind their proprietary models and companies should experiment internally first with low-code and no-code platforms. In many cases, teams can build effective internal tools in just a few hours. 

He uses the example of refining ideal customer profiles (ICPs) using qualitative research and then teaching AI to prioritize leads based on that data. 

“I think there’s value in that process,” said Marsh. “Not only to solve for the use case of filtering the ICP, but also because if the company builds that muscle internally, I think they’re gonna understand much better how they can adopt [AI] in other areas of the business.”

Manufacturing execs aren’t feeling the pressure to adopt AI (yet)

Despite all the buzz, many middle-market manufacturers aren’t feeling pressure from their boards, their customers, or even their competitors to adopt AI. 

Marsh says that’s a missed opportunity. “For my clients, for instance, one of the things that I recommend is that they establish a KPI for their senior leadership team every time they have a staff meeting on a weekly or a bi-weekly tempo, that everyone’s expected to bring examples of use cases of how they’ve tried to use it and what results they found and whether there’s inefficiency or lessons learned and share those with the group.”

Until AI adoption becomes a routine expectation, most companies will stay stuck in pilot mode.

Boards may soon need AI directors

One of the most unexpected ideas from the episode: Some boards are exploring the addition of non-voting AI directors. 

They would act more like scenario engines, running simulations, surfacing blind spots, and stress-testing decisions. Marsh sees this as an evolution of AI’s role — as a thinking partner that helps leadership teams make better strategic calls. 

Key moments: 

  • Why some workers embrace AI and others ridicule it (2:16) 
  • Curiosity is the real differentiator (4:47) 
  • AI prep for discovery meetings can save thousands per lead (6:10)
  • Build your own lead‑scoring tool (9:43) 
  • How building a crisis planning app revealed AI’s strengths and weaknesses (11:20)
  • Should AI sit on your board? (15:30)
  • The zero‑click challenge (18:40) 
  • What’s the biggest opportunity in AI for a legacy factory? (26:45) 
  • Is manufacturing back? (30:32) .

Q&A with Ed Marsh

Q: Why do some people think they’re AI-proficient when they’re not?

A: “If they use Clippy or some kind of a quick tool to help them digest a long email, they think they’re AI proficient,” said Marsh. “We need to view everything as a maturity model. Not everyone has to be in the far right-hand column … but they ought to understand where they are across that distribution of maturity.”

Q: Should manufacturers build or buy AI tools?

A: Marsh encourages building simple tools internally. “If this is something that somebody can sit down and in two hours build and then iterate… a company can legitimately build this internally and have complete documentation and insight to how it’s working. There’s value in that process.”

Q: Is there any pressure from boards or execs to adopt AI?

A: No, but there should be. “I would say there’s an absence of pressure that is startling,” said Marsh. “I don’t see it coming from the board, who I think certainly should be having those conversations. I don’t see it coming from the senior leadership in the C-suite and the executive team. I don’t see it coming from operations.”

Q: How is AI helping boards of directors?

A: Some boards are experimenting with AI as a non-voting director — a kind of scenario engine. “Not a voting director, but one that you can run decisions and risk analysis through and say, ‘Hey, what are we missing?’ Straw man this for us.”

Q: What’s the biggest AI opportunity for a legacy manufacturer?

A: “Return on equity,” said Marsh. “You may not want to sell … But I think that in order to run a business well, you need to be thinking about strategic options all the time. The highest priority for me would be to figure out how to use AI to help me on a regular basis, call it quarterly or semi-annual or annual, essentially look at strategic options, including a return on equity.”

Q: Is manufacturing coming back to the U.S.? Will it create jobs?

A: “Manufacturing output has never fallen,” said Marsh. “Manufacturing employment has — because there’s an increase in efficiency and productivity and automation.” 

He believes reshoring manufacturing is virtuous and good for communities, but notes the jobs coming back won’t look like the ones that left. Robotics and AI will handle more dangerous, repetitive work, while humans will focus on higher-level tasks. 

“If we create an infrastructure of manufacturing and all the businesses that go into that,” he said, “then there’s an ecosystem that develops around it that I think is worthwhile.”