Artificial intelligence is a tool reshaping every sector, from finance to manufacturing. In a presentation at the 2025 Foundry Leadership Summit held Sept. 22–24 in Farmington, Pennsylvania, CEO, CTO, innovator and futurist Kevin Surace explored the accelerating impact of AI on business, innovation, and productivity. Surace explored how foundries and companies in the metal casting sector can apply this new tool to unleash its benefits. 

AI Accuracy and Legal Boundaries

Surace opened by demystifying concerns about AI “hallucinations,” or inaccurate results. Surace believes the hallucination rate in modern models is currently around 1%, a dramatic improvement from 20% just two years ago. This reliability means businesses can increasingly trust AI-generated outputs. However, Surace also cautioned that results must still be verified.

Turning to legal issues, Surace clarified the recent court ruling on fair use, which affirmed that AI training on publicly available materials is legal. “AI learning from published materials is protected as fair use. That’s no longer an argument,” said Surace.  However, the AI model trainer must pay at least as a single user to access the data. He also emphasized that if models use pirated or paywalled data, “the AI model owner may be liable for stealing information.”

For manufacturers and foundries using AI to optimize operations or analyze technical papers, this ruling is significant. It means engineers can safely use AI tools to summarize trade journals, patents, or research papers—as long as proper access fees are paid.

AI as a Tool for Ideation

To demonstrate how AI can be used as a tool for ideation, Surace asked an AI model, “How can I win foundry business from competitors with value over price?”

In seconds, AI provided several strategies and tactics that foundries could explore, from research and positioning to showcasing differentiation, building trust, and competitive tactics. According to Surace, it’s not about whether the information is right or wrong, it’s about using AI to find options you might not have imagined on your own. 

Foundry managers should think of AI as a brainstorming partner who suggests marketing strategies, efficiency improvements, or workforce development ideas. It doesn’t replace human experience but accelerates it. 

AI and the Modern Foundry

Surace highlighted several real-world applications within metal casting. Using a YouTube video titled “Spotlight on the Modern Metal Casting Industry,” he demonstrated how AI can summarize complex audiovisual materials. 

He also showcased how AI tools can generate training and onboarding materials for foundry employees—without cameras or production costs. After showing a training video created with AI using just a few prompts, Surace pointed out the savings. 

“I didn’t have to film anyone. That $10,000 in film costs—I don’t need that anymore. It’s all AI-generated.” For foundries struggling with skilled labor shortages, AI-generated training could help accelerate onboarding while maintaining consistency and quality.
Another striking demonstration showed how AI can analyze foundry videos, determining what was being made, the materials used and the approximate temperature. This capability has the potential to revolutionize process monitoring and quality assurance in foundries, allowing AI systems to detect defects or unsafe conditions through real-time video analysis.

Transforming Sales, HR, and Productivity

Surace emphasized that AI is transforming not only production but also business functions that support manufacturing. “AI is transforming sales,” he said. Tools can now generate personalized outreach messages, analyze LinkedIn profiles, or even send custom video messages to potential customers. Instead of having sales managers train salespeople, he suggests providing your customer’s frequently asked questions into an AI model such as Second Nature, and have new salespeople test themselves against AI-generated customers to improve before they meet with real customers. 

In human resources, AI is revolutionizing recruitment. “Resume screening using AI is becoming the standard,” Surace noted. “Everyone gets interviewed by an AI first.” Every applicant can be thoroughly vetted before they ever meet a human.

For foundries, where the work is physically demanding and turnover can be high, AI pre-screening can ensure candidates are fully aware of job expectations—saving HR teams time and improving retention.

AI for Engineering and Process Optimization

AI’s analytical capabilities will be key for engineers who can now use AI to analyze production data. Instant insights can be found when AI analyzes production logs, defect rates, or energy usage data—identifying patterns that would take days to uncover manually.

Surace also showed how AI could invent and simulate processes. Surace asked the AI tool Claude Science to build a digital twin of a U.S. government onboarding process, to simulate the process, spot inefficiencies, propose optimizations and run them with various scenarios. The AI model identified inefficiencies and took a 379 minutes process down to 70 minutes—an 82% improvement.
Applied to foundries, digital twins could simulate furnace operations, casting cycles, or maintenance procedures, spotting inefficiencies before they cause costly downtime.

Inventing the Future: Using AI for Radical Invention

Perhaps the most intriguing segment was Surace’s call to use AI for radical invention. “Let’s invent something radical in metal casting,” he said. “I don’t know what would be radical … how about self-healing cast alloys via embedded microcapsules?”
AI then generated a detailed description of how such alloys could seal microcracks automatically. While purely theoretical, Surace’s point was that AI can help companies ideate on something brand new that no one’s ever done before. With this capability, even small manufacturers can ideate like research labs and explore futuristic possibilities without multimillion-dollar R&D budgets.

AI and the Workforce: Every Job Is Now an AI Job

Surace believes AI proficiency is important to every position in the workforce. “Every job now is an AI job,” he declared. “If you’re not doing something every hour with this technology, your competitor is—and they’re going to be smarter and faster than you.”
He discussed how colleges must adapt: “No employer is going to pay anyone in the future to write anything by themselves. Just like you don’t pay someone to do math without Excel—it’s ridiculous.”

For foundries, this means tomorrow’s metallurgists, machinists, and engineers must learn to work with AI tools, not against them. Whether analyzing casting defects or managing production scheduling, the integration of AI is becoming as essential as knowing how to operate a furnace.

Putting AI to Use Now and in the Future

Surace encouraged attendees to experiment daily with AI. “Maybe it’s just uploading spreadsheets and getting analysis of the spreadsheet,” he said. 

With a new study from Writer and Workplace Intelligence reporting that 31% of employees admit they are sabotaging gen AI strategy, Surace offered several tips to ensure successful adoption.  

Rather than trying to do a broad rollout across the company, Surace suggested that companies start with a small tiger team. 
“Incentivize them to make it work,” he said.  In addition, he advised companies to focus on productivity improvements and evaluate real ROI. 

For manufacturing, Surace noted that IoT data is critical to success. “You need a lot of data,” said Surace. IOT helps create that data and AI can suggest the changes that can bring improvement. 

When asked how companies can keep their data private, Surace explained that most models have privatized paid versions that the models cannot train or learn on. Information stays within your own domain. 

He also discussed future developments such as humanoid robots, which by 2030 could be rented for $20–$30 per hour, capable of learning jobs from human workers through reinforcement learning. Reinforcement learning is a type of machine learning where an AI agent learns to make decisions by performing actions in an environment and receiving feedback in the form of rewards or penalties. “They’ll work 24/7,” he said. “They’re not million-dollar robots anymore.”

In an industry where labor shortages are a constant challenge, such technology could complement human workers—handling repetitive or dangerous tasks while skilled technicians focus on quality and innovation.

AI as an Amplifier of Human Potential

Surace ended with a reminder that AI is not about replacing people—it’s about amplifying human potential. “I always look at AI as just another teammate,” he said. “A really powerful, really smart, really inexpensive teammate.”

Surace’s message to foundry leaders is both a challenge and an invitation. The future of metal casting isn’t just molten metal and molds—it’s data, algorithms, and turning what was once thought impossible into reality. As Surace put it, “What can’t I do better with AI? Because if you ask me that, it’s hard to find things I can’t do better with Gen AI.”