Russ Bukowski, President of Mastercam.
Manufacturing leaders can’t navigate today’s technological landscape without encountering artificial intelligence at every turn. From medical devices to aerospace systems, AI features are being integrated into virtually everything.
But I believe the question isn’t whether AI is coming to manufacturing; it’s whether the current surge delivers genuine value or simply represents the latest wave of over-hyped features that nobody actually requested.
After spending considerable time speaking with manufacturing leaders, partners and industry experts, I can say that AI dominates nearly every business conversation today. However, when I dig deeper into these discussions, particularly with manufacturing executives, I find that “AI” has become a catch-all term describing everything from basic automation to sophisticated search capabilities—and yes, occasionally, genuine machine intelligence.
The Seven Levels Of Manufacturing AI
Industry experts generally recognize between five and seven levels of AI. I’ve found that understanding this spectrum can help manufacturers separate marketing hype from practical implementation opportunities and have meaningful conversations about implementation methods, development costs and timelines, and appropriate market positioning for different AI categories.
Level 1: Rule-Based Logic
This represents the simplest form of AI, using predetermined decision-making models based on specific inputs. Many existing manufacturing software solutions already operate at this level, making decisions based on established rules and user parameters.
Level 2: Basic Machine Learning
At this level, AI incorporates statistical models to provide recommendations based on historical datasets. In manufacturing, this might include optimized feed and speed recommendations based on proven results from previous operations. While these models typically improve over time with additional data, they can also carry risks based on approximation and data quality.
Level 3: Pattern Recognition
Pattern recognition leverages deep neural networks to identify similarities in complex datasets. Think voice recognition or image identification but applied to manufacturing processes. This technology could potentially revolutionize quality control through automated defect detection or by enabling sophisticated feature recognition in computer-aided design systems.
Level 4: Large Language Models (LLMs)
LLMs have captured the public imagination through platforms such as ChatGPT and Microsoft Copilot. While essentially acting as extensions of pattern-recognition technology, LLMs deserve separate consideration due to their ability to handle unstructured input and generate human-like responses.
However, before employing LLMs in your manufacturing workflows, it’s important to carefully evaluate the risks, including potential misinformation, intellectual property concerns and significant computational requirements that could impact your environmental sustainability.
Level 5: Deterministic Optimization
This level of AI moves beyond human-like behavior to actually exceed human baseline capabilities. It utilizes explicit problem models incorporating both underlying physics and performance metrics. Rather than relying on training data, deterministic optimization tools search for optimal solutions using available algorithms. Based on my experience in the manufacturing software field, I believe this could have the potential to reshape manufacturing processes through micro-level optimization.
Level 6: Sequential Decision Problems
Sequential decision problems extend deterministic optimization by considering future implications of current decisions. This reinforcement learning approach runs multiple “what-if” scenarios to optimize not just individual problems but also entire chains of interconnected challenges. Predictive maintenance would be one example of where this type of AI methodology could benefit manufacturers.
Level 7: High Intelligence
This level remains science fiction for now, representing AI that uses creativity, logic and reasoning to solve problems using emotion and experience, like humans do. Perhaps it’s fortunate that this level remains theoretical for the time being.
What Manufacturers Really Want: Practical Automation
AI often comes wrapped in complex language and high expectations. But in my experience, the real value for manufacturing executives is straightforward: Using intelligent manufacturing can save time, control costs and keep production moving with less friction.
However, right now, those goals are being met with many challenges.
For one, the skills gap continues to widen. Experienced technicians are leaving the workforce faster than new talent is entering. A Deloitte study warns that if we don’t close the talent gap, the U.S. could be short by almost 2 million manufacturing workers by 2033.
Also, material costs are climbing. Inflation, supply chain strain and global competition have increased the cost of raw inputs, tooling and consumables. Product designs are evolving, and this growing complexity increases the demands on quality assurance and production agility.
In addition, global uncertainty is creating challenges across supply chains. Many manufacturing and supply chain leaders worry about rising trade tensions, according to the 10th Annual State of Manufacturing and Supply Chain Report.
The concept of “AI” in manufacturing often represents any automated solution that effectively addresses these challenges. And I have found that the nature of the solution’s underlying technology—whether it is technically “AI” or not—typically matters less than the practical results.
Reframing Existing Capabilities
When it comes to manufacturing software, look for “user-friendliness” over complexity. For example, a software platform that repackages an overwhelming array of technical options as “intelligent selection” or “AI-assisted workflows” can help guide decision-making based on context and historical success patterns without forcing users to master dozens of complex parameters.
This isn’t misleading; it’s simplified, presenting powerful existing technology in ways that highlight its capabilities and accessibility. I recommend looking for the same reframing approach in various manufacturing software capabilities, so you can more easily recognize the immediate value.
The Strategic Opportunity
AI isn’t a silver bullet that will solve every production challenge, nor will it replace skilled workers en masse. Instead, AI represents another powerful tool to augment workflows and improve productivity.
I believe the key insight for manufacturing leaders is this: We’re positioned to define what customers believe AI should be. Rather than chasing AI purity or getting caught up in general technology hype, focus on customer outcomes. If a solution looks intelligent, behaves intelligently and delivers intelligent results, customers will likely perceive it as AI, regardless of the underlying technology stack.
Moving Forward
If you’re considering AI tools, I recommend focusing on balancing customer demands with practical business realities. Success in the AI era isn’t limited to companies with sophisticated algorithms or large data science teams—it also comes from understanding customer needs, identifying practical automation opportunities, and delivering solutions that provide measurable business value, regardless of whether the underlying technology technically qualifies as “artificial intelligence.”
The rise of AI adoption in manufacturing is real, but it’s being driven by business outcomes, not technological complexity. So focus on the former while letting the latter evolve naturally.
Forbes Business Council is the foremost growth and networking organization for business owners and leaders. Do I qualify?