Share content on LinkedIn Share content on YouTube
Peter Ulintz Peter Ulintz
PMA Technical Consultant

Artificial Intelligence Will Not Replace Humans

October 23, 2024
0
Comments

…but manufacturers embracing artificial intelligence (AI) will replace those who don’t.

This column’s title and the words above are the tagline of a book that I recently read by Dr. Markus Guerster, Artificial Intelligence Will Revolutionize Manufacturing. The book focuses on understanding AI and the strategic advantage of adopting advanced-manufacturing systems where AI does not replace human expertise, but augments it.

Today’s manufacturing plants contain a blend of manual, semi-automatic and fully automated processes where tremendous amounts of data flow through PLCs that capture every detail of an operation using sensors and actuators. A portion of this data may be displayed on a control panel for the operator to use in decision making. Other data may be stored elsewhere for later analysis. The problem for manufacturers: It has become so easy and cost-effective to collect and store massive amounts of process data that extracting and leveraging any actionable insight from it can prove impossible.

The emergence of AI has created a unique opportunity for industries to transform. Guerster asserts that the tools and systems necessary for this evolution are within reach. It’s an ideal time to transition from the conventional factory settings of today to the intelligent, data-driven factories of tomorrow.

The History of AI

Not a new phenomenon, AI arguably boasts a history that stretches back to the early 20th century. 

In 1936, Alan Turing published a paper, On Computable Numbers with an Application to the Entscheidungsproblem. The Entscheidungsproblem—literally, decision problem—originally was posed by German mathematician David Hilbert in 1928. Turing proved that a hypothetical computing device could perform any conceivable mathematical computation were it represented as an algorithm. The device eventually became known as the Turing machine. During World War II, Turing helped the British government at Bletchley Park pioneer the technology to decrypt Nazi Germany’s secret Enigma code.

In 1956, American computer scientist John McCarthy and three colleagues coined the term “artificial intelligence” in their proposal for the famous Dartmouth conference in the summer of 1956. This conference essentially birthed AI as a field.

Despite initial excitement, funding for AI dried up and interest from the private and public sectors waned. Research persisted but often in academic settings away from the limelight. During one of these “quiet years,” many fundamental AI aspects emerged, among them machine learning, neural networks and natural language processing.

The Resurgence of AI

According to Guerster, “the advent of the Internet and the explosion of data” coupled with “the availability of large datasets and advances in computational power” gave AI new life. Machine algorithms, fed with data and run on increasingly powerful computers, began to perform tasks previously thought impossible.

In November 2022, ChatGPT, developed by Open AI, took the world by storm. This technology not only understands or processes human language, it also can generate it, opening new possibilities for human-computer interaction.

At the core of modern AI: machine learning, where machines improve and learn over time without any additional programming required. For manufacturers, this marks the transformation from asking machines if they can think to teaching them how to learn. Guerster identifies three steps in the learning process: supervised learning (requiring labeled data and human intervention), unsupervised data (no labeled data, as the machine understands pattern/structure) and reinforced learning (rewards-based learning, where the machine learns how to act in a certain environment). Deep learning, the next evolution, takes principles from machine learning and applies them in a complex, layered approach akin to the human brain’s neural networks. It’s about teaching machines to discern not only patterns but also nuances and context of the data that they process.

Human Interaction with AI

The primary role of AI in manufacturing likely will consist of analysis and discernment of massive amounts of process data—quickly analyzed, sorted, prioritized and contextualized into recommendations executed by humans. Keeping humans in the loop is crucial, allowing them to make final decisions based on AI’s suggestions. Over time and as the technology improves, AI may evolve to a point where it autonomously executes actions, closing the control loop with the machine. However, it is difficult to imagine AI replacing humans anytime soon.

Guerster envisions factories where AI does not replace human expertise but amplifies it. Here, innovation is the currency and those who invest in AI’s transformative power become the industry’s new leaders. MF

Industry-Related Terms: Core, Point, Run
View Glossary of Metalforming Terms

Technologies: Management

Comments

Must be logged in to post a comment.
There are no comments posted.

Subscribe to the Newsletter

Start receiving newsletters.