ChatGPT and Manufacturing

Podcast Series: Finish Strong® – Fulcrum ConsultingWorks Inc.

Manufacturing has played a major role over time in advancing automation, computerization, digitization, and more. Our industries are amazingly different now than they were 15 years ago, much less 30.

In December of 2022 the company OpenAI released an online product called ChatGPT. Unless you’re living under a rock, you’ve at least heard of it. You may not understand what it does, and very few understand how it works. It is, however taking the world by storm.

The question for you is, does ChatGPT offer value to manufacturers?

The short answer is: No, at least not yet.

But the longer answer is: The literally awesome advances in Artificial Intelligence (AI) reflected in the technical underpinnings of the product lay the groundwork for a very different future for all of us.

Importantly it should reassure us that parallel efforts building AI capabilities to solve other larger problems are advancing quickly also.

So why do I believe ChatGPT does not help us manufacturers now? Because it suffers from a problem that all AI will need to overcome: What is the truth used as the basis of its advancing logic and reasoning?

A recent LinkedIn post bragged about the great answer received from ChatGPT when asked to distinguish lean from six-sigma. The problem is, the answer was significantly inaccurate.

How can that happen? Alleged experts have to guide the AI tool’s initial learning in understanding what is true and what is not true. As it trains itself on logic/reasoning from that information, it receives continued guidance, though less and less as its training is judged to be high quality.

This product is not just a fancy Google search engine. In fact it relies heavily on the quality of the information it considers, while Google merely spits back links to what it sees as relevant web content.

How does a manufacturing leader, who obviously has a plate full already, consider the ever evolving technologies arising daily?

In larger organizations with better trained technical experts, she can build in regular updates from those people regarding what is happening and why she should care. Reading, asking trusted peers, following respected sources on YouTube or other social media, and attending technical conferences for leaders are a few methods to consider. Smaller company leaders may choose not to invest resources in this research, strategically behaving as “late adopters.”

Any advance others are leveraging to increase competencies should not be a surprise to you, as the leader of your organization. Know terms and a one paragraph summary of definition, and what makes it special, including strengths and limitations.

You can insource or outsource development of that information, but having it is not optional. Well, it is optional, but not for those building companies to endure.

Invest 5 minutes setting up your ChatGPT account, realize that it handles statements or directives better than it does questions, and try it out for a few topics of importance to you.

If you see capabilities that can help you now, invest more resources; if not, move off of this advance for now.

You can’t implement every technology advance, nor should you. Some of the non-tech advances may be of greater value. You may already have all the change you can swallow.

Please don’t be an ostrich, though the temptation may be great.

The post ChatGPT and Manufacturing first appeared on Fulcrum ConsultingWorks Inc..