MASTER the AI Prompt: Crafting Effective Queries for ChatGPT & Claude

Welcome to the Industry 4.0 Club's fireside chat. The club's purpose is to engage the power of diverse worldwide talent to accelerate the global evolution to Industry 4.0. We accomplish this by hosting and supporting respectful, thoughtful, and interdisciplinary open conversations and creating quality content on wide-ranging Industry 4.0 topics to expand the reach of our community. My name is Mike Ungar, and I am the CEO of the Industry 4.0 Club. I have over 40 years of experience with most of that with Michelin in manufacturing. I'm joined today by my co-host, Mike Yost. Mike, would you introduce yourself, please?

Absolutely. Thank you, Mike, and welcome everybody. As Mike said, my name is Mike Yost. I am the volunteer chairman of the Industry 4.0 Club. I've been in industry about 35 years, about half of it on the industrial solution provider side, and the last 15 years or so on the industrial nonprofit side. So I've been working for a long time to help manufacturers figure out the role and the value and the fit of modern technologies in their operations. And so continue that by having conversations like we're having here today on today's fireside chat. So I'll take just a moment and introduce our guest today. We're happy to be speaking with GenAI Coach and Industry 4.0 Club founder, Jason Bean. Jason's going to be leading a workshop at the MD&M East event in New York City on May 20th, and the title of the workshop is Master the AI Prompt: Crafting Effective Queries for ChatGPT and Claude. So we'll talk just a little bit about that and what Jason's hoping to accomplish. But first, by way of introduction, Jason has more than 20 years of experience in automation and manufacturing industry. He's working with machine builders, end users, systems integrators, as well as OEMs. His passion is to help companies cut through the noise around industrial IoT and to develop— help them develop their digital transformation strategies. His goal is to help leverage the right people, processes, and industry technology partnerships to engage employees and transform their business to compete in the new age of manufacturing. Jason's also one of our co-founders of the club, so known him for several years. And he's also the founder of Empath Consulting Inc., where he focuses on autism awareness training and HR strategies to successfully support all employees, especially those on the autism spectrum. So great cause from a great person. So we're really anxious to get into the conversation with, with Jason. So welcome, Jason.

Thank you, Mike. And Mike, it's great to be back on the show yet again.

Yeah, thanks, Jason. So Jason, you know, You're going to be leading this workshop in New York. And I guess, what interest— how did you first become interested in GenAI? Can you talk a bit about that?

So I first got exposed to it probably about 6 months before it released to the public through a company I was working alongside that had developed a solution. And at that time, they weren't aware what they really had created. And didn't really realize the power of the tool that they made. They later partnered with OpenAI due to synergies and what they offered and partnered with Microsoft. So when ChatGPT first initially released back in November 2022, I jumped on it right away because I saw the similarities between the platforms. Um, and then I just, as with anybody with autism, you find something you're interested in and you just dive in and don't worry about sticking your toe in. You just jump in and drown and figure it out. And so, so I've probably got well in excess of, I'd say, 20,000 hours plus of just playing with The chatbots alone. Wow.

Well, and Jason, you've been running these workshops both in person and virtually for some time now. What are some of the types of typical questions that you're getting from the people that are showing up in these workshops?

Well, the typical things that people really just aren't realizing that these chatbots actually have a place at work. And they can be used at work. You know, most in the workshops, you know, they're either getting introduced to them for their first time, or they've been using them very superficially just to help write an email and polish it up.

Mm-hmm. Right.

And there's so much more that can be done with these tools, especially when you turn them on yourself, on your position, and on your company to actually start Identifying places where, you know, you may have strength that you've overlooked or places you haven't even thought to look that you need to actually improve.

I've had a chance to be in a couple of the sessions you've led. And I'm always quite surprised about the diversity or the range of knowledge in the room, right? So you get the really basic questions a little bit, like which, which GenAI system should I use, right? Because I know there's differences between them, all the way to, you know, helping people figure out which prompts to do. What have you seen most about it? Are you seeing anything particular that people kind of dive into between, between that range?

I would say the most common place I'm seeing people use it Once they start understanding it is developing SOPs, developing SOPs, and as a training tool. You know, they've, they've moved from using it for their emails and small little weak analysis to actually developing stuff that, you know, becomes useful to the manufacturing floor.

When you mentioned training, that's an interesting question. So I think about the jagged nature of it. And so when someone goes to use it as a training tool, how do you avoid it taking on maybe what I would call some bad practices versus the right practices?

Oh, that is a constant battle with it when you're, when you're using it as a training tool. It needs to be audited on a fairly regular basis, and you really need to be careful in how you train these chatbots and what documents you load into them and the format in which they're worded, because these chatbots will adopt a bias based on who trains it. And one of the easiest ways to avoid that is not to have one person train it, but have a number of people train it and test it.

Does that tie back, Jason, then to the title of your workshop about mastering the AI prompt?

It does. I developed my own framework for working with it. For which is kind of where the mastering comes from. It's a it's a way of interacting with it that you really don't get taught through a YouTube video. To my knowledge, I'm the only one that shows people to use it the way that I use it and how to prompt with it. I'm not saying the prompt engineering that's on. YouTube is bad. But there, when it comes to very specific tasks, that for manufacturing or business, there are steps that if that you can take before you even get to the prompt engineering that greatly improve the output of the model and make it far more relevant to your business.

Can you give— so without giving away your entire workshop, Can you give one little hint about one of those things that you just mentioned?

Da da da da da. I could say, let's say when I do the workshop, what I do is I generally open it up with a prompt exercise that walks them through showing how shallow the initial responses really are if you just focus on a Typical prompt engineering. One of the things I like to tell people is, you know, consider these models to be like about 100,000 miles wide and an inch deep in knowledge. And if you don't know how to get the depth out of it, you'll always stay in the shallow end of this, of these tools.

And how are you seeing I guess maybe 2 dimensions to Mike's point about the different levels of experience that you get with the folks in the class. Where does trust come in? Do newcomers coming in, are they trusting these tools and jumping right in, or are they hesitant until they learn some of the tools that you're talking about? Does that change as people get more comfortable writing the right prompts and experimenting with that depth?

Trust of these tools, you see both. You see those that trust it too much and those that don't trust it at all.

Yeah.

You need to be somewhere in between. Right? You can't trust the tool to make a work decision for you because it doesn't carry the responsibility of that decision once it gets carried out.

Right.

You know, that ownership still lies with the individual responsible for that decision. So I talk a lot about, you know, you need to weigh what comes out against your lived experience within that role and within the company. You've got to use your domain knowledge to justify that what comes out is actually for the benefit. Because these tools don't have— they don't have a moral compass. They can imitate one if you train it to. They don't have the emotional intelligence. They may not see the impact of a decision which looks good and will produce results, but they may not see what that does to a team environment. 6 months down the road. That's where the human interaction comes in and the human interaction has to always bounce that recommendation off their experience.

Is there one thing— I can open my computer every day and read something about GenAI. Is there something that people aren't talking about much that Maybe people should be talking about more when it comes to GenAI?

I think using it as a self-reflection tool. We were so used to new technology coming out and trying to find a spot to, to fit it, to fit it where it accomplishes a task and lightens the load for us. But when we turn If we take— this is a very powerful tool to help identify weaknesses and strengths within yourself if you turn it on yourself. And we don't see that being talked about really, that, you know, that this can be a strong tool for personal development, and it can ramp that personal development time up significantly. But at the same time, through improper use or improper understanding, it can take you in the completely opposite direction at the same rate. One of the other things we're not really looking at when we look at these tools is we've been living through rapid technology advancement now for quite some time. It just always seems technology gets better and better. We've always looked at where does this tool go to improve and make things happen? And I'm getting a lot more requests now to talk about the change management aspect of it and how do we handle it.

Right.

And what I'm seeing is change management has become a Band-Aid solution almost. We talk about it and we put a Band-Aid on and we— something else changes, we put another Band-Aid on, we put another Band-Aid on. And technology has ramped up fracturing, not just ourselves and our relationships, but also our, you know, our workplaces. You know, everything starts to get more siloed, more specialized, and every time you, you break off, you leave a little piece behind. Uh, not saying that the fracturing itself is bad. You've got to do that to be able to accomplish, to do very specific tasks. But with these tools, it can ramp that up significantly faster. And you may see smaller and smaller segments of highly specialized. And we keep— and every time you do that, you're going to be leaving something else out that was originally there. And we can use these tools to actually refocus that so that way we retain the identity of what the company started to be, or what the department was meant to be, or, you know, what the role was meant to be without fracturing the role. Like, there are so many people now I know that, you know, because of technology, they're now doing 4 different roles at the same time. And each role they treat as an individual silo, and they don't cross. They don't bring parts of themselves to all those roles. They just bring the highly specified task that they need.

Well, thanks for sharing that, Jason. Just want to remind everybody that we're talking with Jason Bean, who will be speaking at the MD&M East event in New York City on May 20th, 2025, leading a workshop on GenAI. And we'll give instructions at the end here of how to, how to find that event and register for the workshop. Definitely encourage you to check it out. Jason, just wondering about different types of platforms, right? We hear a lot of different types of platforms. Can you elaborate on some of the platforms that are out there, and are they particularly strong in one area or fit for purpose in a different area?

So I do go over this in the workshop, and I go over What I personally recommend based off my knowledge of the models, and I do show some of the backend settings that you can see. So you've got, you know, the one everybody knows is ChatGPT. Everyone knows that name, knows that model. You have Anthropic's Claude, you've got, you got xAI, you've got Google Gemini, you've got Microsoft Copilot, you've got DeepSeek. And there are others out there, but those are probably the main ones that most people know and hear about. They all basically do the same thing. You can get either one of them to do what the others do, but they do it slightly differently. Where the difference comes is in the back end of the models. And my personal preference is to use ChatGPT because of the controls that they give you on the back end and the settings. They, they give you a lot more. Um, you can— you choose to turn on or off using your information to train the model. You have the ability to set, you know, some custom instructions, load in custom knowledge bases that you know, can be usable across any chat where you're not going to a new chat and you have to load everything in all over again. So from a functionality point, I find ChatGPT currently right now to be the best model for chatbots if you're actually using it to do, do specific work tasks. It's not my favorite. My favorite is actually Claude. But I don't have that same functionality or customization ability in Claude that I get in ChatGPT.

Yeah.

Well, thanks for sharing that. And that'll be great insights you have that you can share with the participants. I'm sorry, Mike, didn't mean to talk over you.

No, that's interesting. I get a kick out of that. You know, it's— there's preference and then there's functionality. So that's really neat that you shared that, the difference between those for yourself. So, what are you hoping that people take away from the workshop that you're going to be giving in New York City? What would be some of the things you'd hope people would learn? I know we already have a pretty large group signed up, but, and we could take more.

I hope they walk away with a better understanding of, you know, what the models really are and open their eyes to You know the possibilities where they can use because somebody can give you examples of where to use it, but until you've played with it and start coming familiar with it, you're not really gonna know where you can use it, where it brings value to you and your team.

Right?

I I can use the example of you know a lot of people are using it for training, but you know maybe your company. Has a solid training program and that works really well for them, and you'll find that this really doesn't add any value to what you've already built, right? Um, only you can really determine where this model brings value to your organization, right? If you're taking somebody else's advice on where to put it, you're probably missing something.

Right.

Because you don't operate the same way somebody else does. You don't have the same internal domain knowledge or tribal knowledge that, you know, another company has. And that's another area where these tools are powerful is capturing that tribal knowledge. You can develop chats that are intentionally designed to capture that tribal knowledge from the senior employees before, you know, they walk out the door in retirement or, or head on vacation, right?

You're—

it's a really great tool at capturing that because people are more likely to have a conversation with a chatbot than to sit there in front of HR or their manager and list off a litany of what they do during the day. Mm-hmm.

So just one thought that popped in, Jason, you know, when you, when you talk about how folks are making these kind of decisions. Are you seeing that these conversations about GenAI are coming from like senior leadership down, they're being driven by business leaders to go figure out the strategy, or are they just like, since it's so pervasive and so accessible, do you see pretty much everybody everywhere is just saying, hey, I got to figure this out for me and my role?

I see everybody trying to figure out how they can use it in their role.

Okay.

But in the workshops, it is predominantly senior management.

Okay.

Right. You know, these tools can take any size of company, whether it's a 1-person company or a 1,000-person company, and amplify what you do.

Right, right.

They're— they are great at re— uh, lifting, you know, unnecessary burden from your workers. But, you know, you've got to make that decision as what that unnecessary burden is that needs to be lifted. Right, right.

Yeah, that's great insight. I just— it's one of those things we've seen this in, in, you know, manufacturing for some time. I, I said I've been here 35 years, and there have been people that have been interested in grabbing technology and solving problems. And they, depending on the organization, sometimes, you know, engineers might be grabbing something and running faster than, say, you know, the group or the corporate IT staff is moving. And just the pervasiveness and the accessibility of these tools nowadays You know, that it presents a challenge, both an opportunity for people to grab it and take advantage of it and propose new problems to solve and new approaches, but also the challenge that now you have everybody architecting all sorts of new things.

And this is a tool that where a policy of we will not use GenAI or you are not allowed to use GenAI is not a policy. Sorry to say, it is so easily accessible by everybody. If you have to use it to stay relevant, whether you like it or not, you got to learn it. Yeah, it's, it's like the internet. It's like the cell phone. They're, they're there.

Sort of feels like that personally and professionally, to be honest.

Well, that's really great. So Jason, as we begin to wrap up, any final thoughts for our audience today?

If you haven't started, you're not behind. That's a simple truth. There are so many that have not even started to try these tools out. Don't be afraid of them. You know, dive in. There are no—I don't think there are any experts in this tool as of yet. There are people with higher skill levels.

But are there experts?

I don't think so. The tools are advancing too quickly. You know, we're seeing new models released at a rapid pace, and every time a new model releases, it can change how you interact with the tool. So, you know, just get yours at the bare minimum, at least get yourself a free subscription and start playing with it.

That's good advice. No, excellent. Well, Hey, I want to thank Jason for joining us. Thanks, Jason. It's been great to just kind of talk a bit about what's going to happen at your workshop at MD&M East on May 20th at the Javits Center in New York City. We hope you're going to decide to join, Jason. As I mentioned, we already have a good group signed up, but there's room for more. So here's where you go to register. If you go to this website and you click on the Education and Events tab, and you'll see the, uh, the, the note for ChatGPT and AI Prompt Mastery Workshop. So that's where you go to sign up. Um, so please take a moment, check out the website, and, and join us and join Jason at the event. The club would like to encourage you to take action, uh, to advance the deployment of Industry 4.0. Read an article, attend a workshop, or even develop a smart manufacturing roadmap for your plant or company, or challenge your company just to move forward in the deployment of Industry 4.0. The club would like to know what you're, what you're doing, what action you're taking. So let us know either through our website or our LinkedIn tab. The club also routinely hosts conversations like these on LinkedIn and YouTube, and if you'd like to participate in one of those conversations, again, reach out to us on our website or our LinkedIn page and let us know you'd be interested. And with that, again, thank you, Jason. Thanks, Mike, for your time today.

Of course.

And we look forward to seeing you guys in New York City, uh, in just a few weeks.

See you soon. Take care. Thanks. Bye-bye.