The Fabulous Learning Nerds

Scott Schuette (00:03.118)
Hey everybody, welcome to another fantastic episode of Brand New Year with your fabulous learning nerds. I'm Scott Trudy, your host, and with me, the co-host with the most, you love him, Dan Coonrod, everybody.

Scott Schuette (00:18.744)
Dan!

Daniel (00:20.756)
Scott, what's up man?

Scott Schuette (00:22.658)
You know, I'm doing okay. I'm doing okay. Hey, quick shout out if I could, quick shout out if I could. There you go. ID Lance, we had Parker and his team on a few episodes back. Yeah, I reached out because I needed some help and I reached out and they did amazing work. So folks, if you need a freelance designer, they're a great resource. did fantastic work.

Daniel (00:35.036)
Yeah!

Scott Schuette (00:51.04)
It equally as good as the work that you do, Dan. So really, really awesome stuff. How are you doing, sir?

Daniel (01:00.0)
Fair to Midland. Just busy, busy holiday season. Things are kind of finally catching up and cooling down. There it is. There it is. What about you? How are you doing?

Scott Schuette (01:09.964)
Yeah, yeah, All right. I'm doing OK. Santa gave me COVID, which is not fun. But I don't have it anymore.

Daniel (01:17.123)
Daniel (01:22.75)
Worse than a bag of coal.

Scott Schuette (01:23.146)
That's right. Yeah, yeah, yeah. Did I ever tell you where the whole, by the way, just as a nerdy thing for people, did I tell you where the whole bag of coal comes from? No. You want to take a guess as to where it comes from?

Daniel (01:35.496)
Now.

Daniel (01:42.1)
I grew up in central Pennsylvania, right smack dab in the heart of coal country. And I have no, no idea. have no idea.

Scott Schuette (01:54.678)
Okay. All right. Well then you would be wrong. But anyway, that being said, it comes from Krampus, our good friend Krampus. So Krampus kidnaps children that are bad and throws them into a wicker basket that he carries with them and feeds them hot coal, shoves a hot coal in their mouth. And that's been tamed down from the Nordic tradition. And that's why you get coal on Christmas if you've been naughty.

Daniel (02:03.028)
Krampus!

Daniel (02:24.468)
What's the other Christmas tradition? There's a giant, there's like a Norwegian giant cat that will eat you if you haven't got new clothes. The Yule Cat. Yeah, that's my favorite.

Scott Schuette (02:31.308)
The Yule Cat. That's right. The Yule Cat. Yes.

Samuel VanTassel (02:31.935)
Yeah, they're your cat.

I think it's Jollacolter? No,

Scott Schuette (02:38.028)
Yeah, that voice you hear by the way is our man behind the curtain. We should probably introduce him before more people go, what the hell is that? And Sam's in the house everybody.

Daniel (02:39.346)
Yeah.

Samuel VanTassel (02:45.258)
always here.

Scott Schuette (02:55.33)
Sam!

Samuel VanTassel (02:57.505)
How's everyone going?

Scott Schuette (02:59.502)
Good, good. So you, how's your house?

Daniel (03:01.172)
Yeah.

Samuel VanTassel (03:03.681)
house is doing good. I very much enjoy having a house. I have a... Oh, what do I do? Well, I'm gonna sh... Yes. As a first time home...

Scott Schuette (03:08.133)
you do?

Scott Schuette (03:11.81)
No, no, no, I say, you do. Like, that's great.

Daniel (03:15.892)
Ha.

Scott Schuette (03:17.644)
More often than not, feel like Tom Hanks in the money pit. That's me.

Daniel (03:21.948)
Hahaha!

Samuel VanTassel (03:22.465)
As a first time homeowner, will say this, like the first month I like had the temperature up to 75. It was nice and toasty and great. Then I got my gas bill and now I'm just keeping it like around a 68.

Scott Schuette (03:36.781)
Wow.

Daniel (03:36.98)
That's still too warm. Still too warm.

Scott Schuette (03:39.694)
Yeah, yeah, that'll happen, but that's totally okay. 75 in the winter.

Okay, yeah, that's interesting. I guess so, I guess so, for sure. Hey everybody, we're gonna go ahead and cut to the chase. It's okay, we get some groovy stuff to talk about. We're getting super excited about this week's discussion. So let's go ahead and jump into our topic of the week.

Daniel (03:47.796)
Look at you, big spender. Once.

Samuel VanTassel (03:51.392)
once.

Daniel (03:57.054)
Let's do it.

Scott Schuette (04:12.462)
Okay, we're going to talk about leveraging tech to create content. A to B. Or is it A to Z? A to Z. A to Z this time, right? So we're going to start at the beginning and get to the end, like leveraging content. Lots of groovy things that can help you create content. Where do I start? How do I think about it? Dan, help us out. Where are we starting?

Daniel (04:22.087)
A to Z.

Daniel (04:37.502)
So we talked about this a couple days ago, we talked about this pre-show, but I was thinking we talk a lot about AI tools and AI usage and you know, for good or for bad, these tools exist now and kind of like burying your head in the sand and either waiting for them to collapse, pop or go away, it seems more and more like a fool's errand. We on this show have talked a lot about how great they can be, how

what cool things you can do. And we've talked about a bunch of different tools here and there, off and on. And I was talking to Scott and I asked Scott to challenge me to basically pick a topic and we were going to build a presentation on it today. We're going to build it during this podcast and we're going to share what gets made at the end. Good, bad and different. I don't even know what the topic is yet. Scott's going to surprise me with it and we're going to do it real.

In 30 minutes, I'm going to walk through all the steps, talk through everything, while Scott and Sam chime in with questions and suggestions and stuff like that.

Scott Schuette (05:42.255)
Okay. All right. Well, let's start. I need a topic and I'm not going to even monkey around really. I'm going to just go ahead and I'm going to ask ChatGTP what a topic could be, right? Well, why not? mean, so what are we leveraging? It's the topic's going to be, you know what? I already kind of did this with the, were looking for some ideas, but why don't we go ahead and let's see.

Daniel (05:53.575)
my god, top to bottom, alright.

Scott Schuette (06:15.502)
Here's a discussion the Civil War was neither civil nor a war. Discuss. So there we go.

Daniel (06:26.556)
Is that our topic?

Scott Schuette (06:28.052)
No, no, no. That's what you guys are going to be doing while I search for a topic. you know what? It came up with, since we're using AI, why don't we have our topic is why most AI training fails. What do you guys think of that? OK. All right.

Daniel (06:35.028)
What's up?

Daniel (06:43.257)
I love it. I love it. I love it. All right, so you're the my show here

what objectives you want from this course.

Scott Schuette (06:53.294)
That is fantastic. I don't know what I would want because I just got the why most AI training fails. To me, if somebody came up to me and asked me for this, I'd be like, for sure. I would need to go and research that. Where should I go to research that?

Daniel (07:12.274)
Well, I tell you if if.

If you're not sure what objectives you want to go from, why don't we just go for the jump? Normally I would tell somebody not to just build, but since we're on the timer, why don't we just get started and we start to build and we see where we go. We see what comes out.

Scott Schuette (07:31.286)
Right. OK. So I went back to my agent, went back to my ChatGTP agent. By the way, who's your agent, Daniel?

Daniel (07:38.998)
I'm not gonna say it, because I'm gonna be pulling up that stuff while we start building. I'll talk to you about it. You guys will all see.

Scott Schuette (07:43.923)
okay. All right. Sam, do you have a preferred agent?

Samuel VanTassel (07:49.05)
I can't say that I do.

Scott Schuette (07:51.841)
you need to get one. Folks, I guess that'd be step one, get an agent. Like if you don't have an agent that you, I don't care who you're using, but just have one and then have that be a preferred one. I have two. I use Chet, ChetTP and I use Gemini. I use them for different things. I'll probably get three or four before things are done, just because certain tools are better at stuff. But I will say that the more you pour into one, the more it learns what you want and what you need and who you are what you like and all the other groovy stuff and the closer it's gonna get to the right answers, right?

If I'm thinking about why most AI training fails, like for me, as I'm just having this conversation, it came up with most AI training fails not because AI is difficult, but we teach the tools instead of how to use them. So it's all about an AI skills gap and context gap. So from that perspective, if I were looking at this, the objective is to explain

Daniel (08:40.617)
Yeah.

Scott Schuette (08:51.926)
is so that people have a better understanding of how to use AI to help them get their work done. And then also explain some of the common pitfalls that organizations and departments use when rolling out AI training. What do you guys think?

Daniel (09:09.236)
Okay, so basically we're looking at, I love it. So basically today, once we start the timer, we want to build a course on why, what was the topic again? Why AI adoption fails.

Scott Schuette (09:22.092)
why most AI training fails.

Daniel (09:24.658)
Why most AI training fails and we want to our objectives, at least the things we want to cover, the topics we want to cover are basically how organizations aren't basically like doing a good job rolling them out. They're dropping tools over skills. And what was the other one?

Scott Schuette (09:45.103)
tools over skills and then taking a look at Common pitfalls, although if to to be fair while we were talking I actually Asked it to do two different things like come up with an outline that that would be okay and then create objectives look Which you can do so I actually have five objectives. Do you want to hear them or not?

Daniel (09:50.942)
common pitfalls. Okay.

Daniel (10:02.93)
Yeah, well.

Daniel (10:06.948)
No, not yet, because again, we're going to start fresh. We've got common pitfalls. Why most AI training fails. And I'm just writing these down.

Scott Schuette (10:08.462)
Alright.

Scott Schuette (10:16.046)
You know what would been a really great idea while we were doing this, but we didn't do it. But we should have been, well, we're kind of doing it. But anytime you're having a brainstorming meeting with folks on this kind of stuff, what should you automatically do?

Daniel (10:23.348)
What's that?

Daniel (10:34.1)
I'm not sure what do you automatically do.

Scott Schuette (10:36.394)
Well, I ask if it's okay to transcribe the meeting because you're taking notes and writing stuff down and I don't want to do that anymore because I don't need to do that anymore.

Samuel VanTassel (10:36.544)
Ha ha ha.

Daniel (10:39.501)
yeah, yeah, yeah.

Daniel (10:48.476)
Yeah, I don't think you're wrong. I've got an AI transcriber that I do use. I will say this, I find not as much anymore. We're getting sidetracked. getting sidetracked. We've got a course to build inside of 30 minutes here. So, common sense is why most AI training fails and teaching tools over skills. All right. So.

Daniel (11:20.616)
When you're ready, say go. I'll share my screen. All right. So literally, I am going to share my screen just like I said. Here we go. Again, this is not a screen that you guys will be able to see, but this is so that way Sam and Scott can throw questions at me. I've got a blank page open up to my Google homepage. I'm going to dive in. Scott asks, which agent do I often use? So just so you guys know, we've got our topic and some things that Scott asked to cover.

Scott Schuette (11:24.514)
Go.

Daniel (11:48.392)
And so I'm going to dive in by first opening up my tools. And I've got a couple of tools I like to use. Obviously the present leader-ish is ChatGPT. That's going to open up. I'm also probably going to pop open Claude. For those of you guys who haven't used Claude, Claude is really good at creative writing, at writing things that sound and feel human. And next up, I'm going to open up Kimmy. Kimmy is kind of a newer tool. I find that Kimmy does a lot of creative stuff.

really, really well. It's got a great kind of like agentic kind of use in its OK computer mode. And we're actually going to use that to do some heavy lifting. I'm probably also going to open up, just so we've got to open up, I'm going to use Google's AI Studio.

so I can open up. That's not what I wanted.

There we go. Just go back to the dashboard. There we go. So that way I can open up Nano Banana Pro, because I'm probably gonna need that to make some stuff. All right. What's that?

Scott Schuette (12:53.56)
Why are you gonna need Nano Banana Pro? Why are you gonna need Nano Banana Pro, sir?

Daniel (12:59.22)
I really like Nano Banana Pro. I will say that recently ChatGPT's update has made it a lot better, but still Nano Banana Pro, just for ease of use, is still kind of king. There's some tools you can run kind of like offline at home, like the new Flux model is really good if you've got enough horsepower to run that, but I'm trying to stay focused on stuff that anybody can run as long as they've got a PC. So I'm going to start here.

Scott Schuette (13:15.854)
Mm-hmm.

Scott Schuette (13:26.243)
Got it.

Daniel (13:29.276)
And I'm basically just going to get my prompt and I'm going to drop it into kind of my two hard workers, chatcpt and Claude. So the prompt I'm going to drop in basically is, want to build a brief training on why most AI training fails. Easy enough, nothing fancy. There's all kinds of tools and tricks you can do for like

really in-depth prompts and stuff, but we're gonna start kinda simple. Let's see. The two biggest things I'd like to cover are...

Daniel (14:14.74)
Let's see. Businesses teaching tools over skills and common pitfalls in AI training usage. Scott, you feel like that kind of covers what you wanted to see well enough?

Scott Schuette (14:35.566)
Well enough, um Well enough a quick question. Oh, I know keep going I guess never mind forget it forget my question

Daniel (14:37.426)
Well, well enough. Okay. Yeah. Yeah, so I'm dropping this into ChatGPT and Claude. All right. For ChatGPT, I'm gonna put this on thinking mode. I want it to think through that might, what's up?

Scott Schuette (14:55.561)
Why... I was gonna say why are you putting it on thinking?

Daniel (14:58.042)
I'm about talk through that. So because we're wanting to build a training and I just don't want this to just kind of like pick up like the auto mode and just run through some quick stuff. I want this to think at least for a couple of minutes. Claude, I don't need it too much, but I do have extended thinking already turned on in Claude. Again, just so you guys know, these tools do have like free usage tiers that don't always involve thinking or pro or extended thinking. So...

If you're looking for a model that has a generous free usage, I would also suggest we've got Google's AI Studio also open that we're going to use for images. But you can use it as well to build. And with the new launch of their Gemini 3 Pro, it does a pretty good job. I don't have any complaints about it. But I'm going to hit go on these. Now, if I really wanted to get into some more complicated prompts, I could tell it to ask me some questions. We could get into like...

response rankings, things like that. But we're looking at just a place to get started. And also I'm trying to be very mindful of the fact that we've already into our 30 minute timer on what we can build inside of 30 minutes. so Claude is still thinking. looks like Claude gave me an error. That's not great for it. We're try that again.

Scott Schuette (16:11.064)
Gotcha.

Samuel VanTassel (16:18.689)
you

You can do it, Claude.

Daniel (16:24.02)
I believe in you. All right. There we go. It looks like it's still thinking. There we go. ChatGPT, already has some stuff for us. Let's take a look at what it suggests. So in this case, ChatGPT said, here's a tight, brief training. You can run in 20 to 30 minutes or stretch to 45. That nails your two points. Tools over skills and common usage pitfalls.

It suggests a title for it, why most AI training fails and how to fix it. It's already thinking about what our audience might look like. Audience plus outcome. Audience, any business team adopting AI. By the end, participants can explain why tool-first AI training breaks down, use a simple skills-based framework to pick the right AI approach, avoid the most common day-to-day pitfalls. Its agenda, it's suggesting for 25 minutes is spend three minutes on our hook while your AI training doesn't stick. Its core idea for eight minutes is stop teaching tools, start teaching skills.

And then it wants us to spend 10 minutes on pitfalls, the eight ways AI usage goes sideways and fixes, and then micro practice for four minutes, rewrite a tool-based prompt into a skill-based workflow. And then from there, it kind of lays out, oh, go ahead. From there, it kind of does a little bit more in depth over what it's going to cover and what it's going to look like. We're going to hold on to that. Claude has come through. It says it took it a minute and 10 seconds to think. We had an error the first time, so it took up two attempts, but close enough.

Scott Schuette (17:30.86)
Hmm, interesting. That looked good. No, that's good. Yeah, yeah, yeah.

Daniel (17:50.004)
So it says great topic and one that's generally underserved right now. Let me help you shape this. It says tools over skills is drawn anchor point. The core argument writes itself. And it's kind of got like, just kind of like a quick brief, much briefer than what we saw with chat GPT. But already here, it's giving us a high level so it can help guide what we want to build and what that's going to look like. So this is where there is one thing.

that we would diverge from that we're not going to do in today's 30 minute go go go scenario.

AI, I'm getting to it, AI is only as good as your research. AI is a great tool and you'll see this slapped on every chat bot right now. It can make mistakes, please double check responses. It can make mistakes, please double check responses. And it's so easy to fall into looks good enough.

Scott Schuette (18:26.988)
which is

Daniel (18:54.45)
And I think that is probably one of, we're talking about why your AI training fails and we're building this quick 30 minute thing. This is one of those places where I do believe if we weren't trying to do this kind of under the gun, so to speak, and we were just trying to build something and I wanted to use AI as a partner, I would think about it like a partner. Somebody that, you know, I need to go back and check behind all the work. It's really easy to skip this step because it's going to look good enough.

when you do a first pass. But if you don't do that research, time and time again, you will find yourself put in tight spots that you'll have to kind of go, well, I used AI. And that can maybe get you in hot water, depending upon your job, what you're doing and where you work.

Scott Schuette (19:40.451)
You know, I think it's really, really important. Our value today is really in inspecting our work and the impact. And we have to do that. In case in point, I used, I had to create a VO script for eight online learnings. All I had was the PowerPoints. I didn't have the VO script. Normally that would take for eight online learnings. How long would that take you, Dan?

Daniel (20:08.164)
for what? For eight voiceovers? scripts?

Scott Schuette (20:09.516)
writing a script, eight voiceovers. Yeah, each one of them about 20 slides minimum.

Daniel (20:14.992)
Yeah, that would take me a while. Yeah.

Scott Schuette (20:17.166)
I have a couple of weeks probably, at least a week, right? If I was heads down and didn't have a lot of meetings, yeah, yeah. So I asked my agent to go ahead and do that. And for the most part, it did a really good job with that minus ... And it was cool because I asked it to write the scripts and then I asked, could you put it in the notes page of the PowerPoint? One at a time, by the way. Don't give your agent too much to do, it'll erp. And then one of them, it kind of just repeated a script.

for like five slides straight and I didn't double check the work. So I uploaded it, generated the voiceover, blah, blah. But I was smart enough to have somebody on my team that had no visibility to what the work was at all. Could you, do you have some time to kind of listen through this, just listen to it? And I got some feedback right away like, yeah, this one just repeated itself five times. And I missed it, I totally missed it. And so don't allow.

Daniel (20:47.912)
Yeah.

Daniel (21:10.366)
Yeah.

Scott Schuette (21:12.554)
our ability to say, I got something done in a day to get in the way of inspecting our work, I guess is where I was at. I did fix it and I used my agent to fix it, so that was great.

Daniel (21:23.454)
You're eating my clock. I got to get back to it, Scott. My boss is going to yell at me at 30 minutes. All right. So the last thing that kind of in our Claude post, it asks is like, what format are you envisioning? A facilitated session, an e-learning module, a one pager, and how long do you want it to run? That'll help me figure out how much we can pack into what structure makes sense. This is something to consider as you're working on building something. ChatGPT made a bunch of assumptions. I don't think they're entirely wrong. And for the most part, like it's kind of going to be like,

Scott Schuette (21:25.966)
sorry. OK, all right.

Samuel VanTassel (21:30.517)
You

Daniel (21:52.552)
what we're gonna do for today. But Claude at least had the wherewithal to be like, okay, what does this look like? So I am gonna just in one prompt, this is just for Claude say, hey, I'm envisioning a 30 minute instructor led.

training session.

There we go. And again, that's not a prompt I popped into ChatGPT because ChatGPT kind of already said, you're probably thinking about 25 minutes. And again, I kind of took that as that's what we're going to do. And so it's already coming in. It's like, 30 minutes is tight, absolutely doable for two focus points. Words are tough. It's absolutely doable for two focus points. Here's a structure that should work. And listen if this sounds familiar. Spend three minutes in an opening hook.

Spend 10 minutes on skills versus skills. Spend 10 minutes on common pitfalls. Close an application. And then two minutes of buffer built in for questions and transitions. Right here, Claude says, hey, what do you want me to build out for you? I can put together a facilitator guide, a slide deck, discussion questions, or all of the above. Also, who's the audience that helped me pitch the examples right? Corporate generalists, a specific department, leadership, et cetera. So now I'm at a place where I've got

already kind of like a high level overview from ChatGPT and haven't quite gotten there yet in a couple prompts from Claude. Again, if I had had more specifics to plug in, I think it might have gotten there as well. But let's go ahead and let's tell Claude that like, let's build this out as a presentation. Say presentation deck.

Daniel (23:42.324)
I'm gonna say, let's go slide by slide. Can each slide contain.

title on screen text a suggested image

Daniel (24:04.722)
and we'll say speaker notes.

Anything else you think I'm missing there, Scott or Sam?

Samuel VanTassel (24:15.489)
did you ask it if it was or did you tell it was a computer-based learning or a instructor-based?

Daniel (24:27.482)
I did, did. Up at the top, was like, hey, this is a 30 minute instructor led training session. Though to your point, Sam, I could say this is a virtual training session, but I think for this example, just as we're trying to build, we're going to stay focused on just like it's a presentation, it's instructor led. There's one other thing I did add that I really find helps. And I'm saying, hey, I also want you to include the purpose of the slide. Not every slide needs the purpose. Me as a stickler, somebody who makes slides, I love to put...

Samuel VanTassel (24:32.566)
Fair enough.

Samuel VanTassel (24:42.517)
Got it.

Daniel (24:55.356)
slide purpose or purpose of slide of my speaker notes that way as an instructor or facilitator. I ever get lost, I know like, okay, the purpose of this slide is to teach this. I found as I've used ChatGPT and Claude and other AI, you know, methods that telling it to think of the purpose of each slide helps keep it on task. So there we go. I'm going to copy that prompt. Yes, Sam, what you got?

Samuel VanTassel (25:18.515)
I, what if we told a little bit more about its audience? Like who is who are the learners and how many learners there are?

Daniel (25:29.684)
So I will say, Claude did ask if you want to know about little bit about our owners and our, not owners, excuse me, Claude did ask if want to know a little bit more about our audience. It suggested generalist, corporate generalist. And I'm going to say, let's assume this is for corporate generalists. All right. So, and then I'm saying, let's build a presentation deck. Let's go slide by slide. Can each slide contain the title?

onscreen text, a suggested image, speaker notes, and purpose of the slide. So I'm going to, I'm copying this, almost exactly the same. I'm leaving the, let's assume this is for corporate generalists because ChatGPT has already assumed that. And I'm hitting go on both of these. So I've hit go on ChatGPT. I'm hitting go on Claude. It's sitting there while it's thinking. We can already see, one thing that you'll see in Simi's models is that you can actually click to see their thought process.

I find that if it's taking a long time, being able to pop that down and see kind of gives me like an idea of like, is everything going right? Are we thinking the right way? Have I forgotten anything? Already we can see that Claude has already started dropping. right. Like, Hey, slide one, a lot of speaker notes, a lot of purpose of the slide. And then it stops. It's like, Hey, are you ready for slide two? Let's go over and see what chat GPT is doing. Chat GPT is still thinking. We can pop out it and kind of see what its thought process is, which is still loading.

So we'll go back to Claude and I'll say, yeah, I'm gonna tell it to let's generate all of the slides we need for this presentation. That's probably gonna take a minute. And I will say this, if I weren't doing this underneath like a 30 minute deadline and we weren't already like halfway through, I'd probably let it do it slide by slide. And Scott and Sam, the reason why I do that is because

It really helps it to stay focused. It helps me catch errors sooner. It's much easier as regular humans to read through slide by slide. Yep, that's right. Yep, that's right. Yep. Nope, that's not right. Hey, we on slide three, we'd said this. I don't like that. Let's rebuild it. And this is why, but again, the timeline is the tyrant. So I'm saying, let's generate all the slides we need for this presentation in the cloud.

Daniel (27:53.716)
All right, we're going to bounce back over to chat GPT. And already it's got, it looks like.

kind of our already kind of got a 10 slide presentation.

Scott Schuette (28:09.047)
Nice.

Daniel (28:10.334)
So in its presentation, slide one is the title and opener. The purpose is frame the session as skills-based and set the tone. Slide two is hook looks good, does not equal works. Purpose of slides create immediate recognition and urgency. Slide three, the core diagnosis, purpose of slide, name the main problem early. Slide four, what tool first training creates and shows the consequences so the fix feels necessary. The replacement, train durable skills, purpose of slide.

Provide the new mental model and structure. Slide six, skill number one dash two, framing plus context. Teach the first practical upgrade people can apply immediately. Skill number three and four, iteration and verification. Replace single prompt behavior with repeatable habit. pitfalls one. Give concrete watch outs and their remedies. Common pitfall number two. Address risk and governance without getting heavy.

And slide 10 is Micro Practice and Close. Let's take a look at Claude. Claude has built out the full deck. Let's take a look and see what it's done. So Claude has built this out. Hey, ooh man. It's still working. We're on slide 19. It's doing a lot of thinking. looks like Claude has built out almost double the presentation. It's thinking about 20 slides.

Scott Schuette (29:31.512)
Wow.

Daniel (29:32.936)
So I won't go through all of those, but just kind of hit some of the highlights here. looks like slide one is our title slide, slide two is the problem, slide three is an agenda. That's something that we didn't have before. Slide four, section header, tools versus skills. then slide five gives us a whole analogy. Slide six is tools, expire, skills, transfer. Genuinely, I like that. Slide seven, skills that matter.

Slide eight, quick activity. Slide nine, section header, common pitfalls. Slide 10, we get to pitfalls. And then we cover pitfalls all the way until slide 14. And then slide 15 is basically what we do instead, how we train for thinking, how to use AI like the collaborator, takeaways, recap, questions and answers. All right, Sam, Scott, I've done a quick dive through both of those. I know we are quickly eating into our 30 minutes. We want to see how far we can get.

Which one do you guys want to run with?

I know you guys haven't had chance to take a look, but what's your gut say?

Samuel VanTassel (30:38.047)
My get says Claude because it is giving a lot more reasons behind it. Now don't think Claude will have the most cleanest. I think that reward would go to the Google AI Studio. But Claude gives a lot of reasoning behind why it's doing it. And I like that as someone who's going to go back and improve what it's saying.

Daniel (31:04.37)
I like that. Scott, what do you think?

Scott Schuette (31:07.886)
I mean, I get more with Claude. I do like how succinct and directional chatGTP is, but I'm not going to die on either hill.

Daniel (31:22.484)
Yeah, I mean, we shouldn't, it's, you know, AI stuff. I am, yeah, I'm gonna go ahead and what I'm doing here is rather than leave it to chance, I am taking the outline that Claude made and the outline that ChatGPT did and I'm dropping them into each other's chat. And I'm saying, hey, can you review this alternate outline and compare and contrast them? If we weren't on a 30 minute timer, I...

Scott Schuette (31:25.902)
Yeah, I guess it's whatever works for you, whatever you feel better about.

Daniel (31:49.78)
probably would just, this is the step I would do myself and then kind of recombine them into one outline. But again, time is essence. And so I'm going to let that thinking up to them for now. There's a scary phrase. All right. So I've kind of given it to both of them and now they're thinking. So Chachapiti is thinking, Claude is thinking.

So ChatGPD says, yep, your alternate line is stronger as a real presentation, flow, pacing, narrative voice, while the earlier 10 slide version is stronger as compact framework and checklist. Here's the compare and contrast. It tells us what the same, it tells us key differences, it tells us tone of buy-in, tools versus skill expression, skill list design. It's still going through it, it's talking about pitfalls, section depth.

Here's what Claude says, what the alternate does better, tighter slide count, explicit skill framework, pitfalls paired with fixes, stronger hook, ends with practice, not recap. What my version did differently, not necessarily better, how humble of it. Its recommendation is use the alternate as your base structure, it's cleaner, more actionable, and better suited to 30 minutes, but consider adding one mid-deck engagement moment, borrowing the collaborator, not vending machine frame, and keeping the word and writing analogy.

ChatTPT is still kind of doing it. It says what your outline does better, where you can tighten your outline. Recommended best of both deck, your goal is to serve events. I'd run your deck structure 20 slides, but trim to 14 or 16 by combining if you recap transition slides. Use the five durable skills from my presentation. Keep your four pitfalls. So I think I'm going to, at this point, I'm gonna stick with Claude and I'm gonna say, okay, let's rebuild with your...

suggestions.

Daniel (33:42.9)
Checking to look at the clock, I think we've got less than 10 minutes left. Scott, Sam, any exacts on that?

Scott Schuette (33:53.218)
No, think we're getting pretty close. He's done a decent job.

Samuel VanTassel (33:55.233)
I think we're getting there.

Daniel (33:56.532)
All right. Yeah. So.

Scott Schuette (34:00.761)
So as you're thinking about this, right, so from an audience perspective, like I don't want to dive too much into it, but it might be a good place to start before we start building content. Like we've discussed how to define an appropriate topic. We've taken a look at building objectives. We took a look at defining an outline without diving too deep into it. What are some overall...

things that people should be aware of as they're thinking about leveraging new techs, new tech like ChatGTP or Cloud or whatever, to help them really design training and do so in a way that helps them show up prepared and enhances their value at the table from a leadership perspective.

Daniel (34:46.654)
Man, I will say we've talked about AI tool usage a lot on our podcast. It's an exciting new tool. It really has a lot of potential to like change. Basically like, I mean, it has a lot of potential to change how we work and what our jobs are. That can be scary. We've talked before about maybe some of the scaredness that's coming through and what that might look like. But.

I'd say the biggest thing is like use them. I had a lot of apprehension at first. Again, as these tools, when they first hit and they're able to just like create walls of text and looking at them and seeing that early on, it was just a lot of warble garble and being like really worried about like what that might look like and like how that's gonna work. I was very unsure.

I was very unsure of like who the tool was really for. And, and I was worried about what people would do with it. As time has gone on, it's not like people have used these tools any less if anything, we're using them more, but use them. Like that's, that's the biggest piece of advice is take time, use them, even if you're just using them for dumb things. Scott, know upfront, you said you're all in that you use the same chat service because you want it to know you.

I'm not even at that depth. get nervous when I'm worried that my chat bot knows me too well. I don't, I'm going to say it'd mean. I don't want to be friends with my chat bot. It's a portal to the cloud and other things and other people. And I don't know how comfortable I am with everything getting shared. tend to not use it for not work stuff. I'm not asking it personal questions or God forbid giving it health or

mental questions. I'm always trying to be mindful of what I'm putting in there, but I am using it on a regular basis and I'm looking at new tools. I'm looking at the new things that it can do. And then I oftentimes find myself trying a tool for the first time being like, don't like that. And then telling myself I'll come back in a month and I'll see what changes happened. And usually that's enough for me to be like, wow, we've done a lot of cool stuff in a month and this is what it can do. I don't know. Does that help answer your question?

Scott Schuette (37:10.478)
Yeah, absolutely. Totally, totally get it. Although I'm all in, it knows everything. It knows what I had for breakfast and whatnot. By the way, you can go into settings and you can change all of that. Like if you don't want it being shared out, just go to your settings. And you should do that, by the way. You should totally, totally do that. I mean, it can learn on its own in the environment that you allow it to versus sharing all that information out with everybody, which is a great, great thing.

Daniel (37:21.672)
Yes.

Scott Schuette (37:41.294)
Cool, so this is fantastic. We'll have to pick this up because I think the content creation and the content generation and how we create content would be a really nice place to pick up when we have more time. What do you guys think?

Daniel (37:56.008)
Well, I think I'm actually, we've got about two more minutes left on our clock, I think. And I am right now, just while you were talking, I've already gone over to Kimmy and I've basically asked that, Hey, I want you to build this presentation. And then I put in the presentation that we've got.

Scott Schuette (38:01.113)
Okay.

Daniel (38:20.98)
And because I goofed, I think we didn't end up turning on OK Computer, which I'm going to do right now and then hit go.

Obviously, this is going to take us to our 30 minute break and it's not going to spat out the content. So we didn't quite get any content done. Again, I think we talked a lot as we were just talking through stuff like that. But honestly, goodness, by the time this prompt is running, as we start the 30 minute countdown, we're going to have a rough draft of a presentation. It will definitely not be the end all be all. It's not something that I would turn around in hand and be like, all right, cool, this is ready to go.

Samuel VanTassel (38:38.709)
Ha ha ha ha.

Daniel (39:00.53)
And it's something that Scott, Sam, think if you're fine, we can share on the website so everybody can see it. But it would be a place for you to like start like polishing and finishing. And instead of spending all day producing this piece of content, you might only spend a couple hours, especially if you've got a site that you can grab the elements for, your visuals and things like that.

Scott Schuette (39:28.344)
That is fantastic.

Scott Schuette (39:37.432)
Take a look at Kimmy. That's great. Hey, Dan, do me a favor. Could you go ahead and let our audience know how they could get a hold of us?

Daniel (39:46.964)
Absolutely. Guys, you know this stuff. You know this by now. Email us at nerds at TheLearningNerds.com. Email us any questions you may have and join in on the conversation. If you're on Facebook, you can find us at Learning Nerds. For all of our Instagram peeps, Fab Learning Nerds. And lastly, for more information about us, what we do and updates, www.TheLearningNerds.com. Scott, back at you.

Scott Schuette (40:09.966)
Hey, Dan, thanks so much. everybody, do me a favor. Could you go ahead and hit that Like button, hit that Subscribe button, share this episode with your friends? I know that they need it. I know they're sitting there maybe doing things the old fashioned way and could do it a whole heck of a lot better. And I think this is a nice framework that allows people to go ahead and explore and find out what works for them. So go ahead and do that for us if you could. Also, go ahead and leave us a review.

either on Spotify or iTunes or wherever you get this podcast, because you know what? It helps the algorithm get this information out to more folks. And with that, I'm Scott.

Daniel (40:45.78)
I'm Dan.

Samuel VanTassel (40:49.012)
I'm Sam.

Scott Schuette (40:49.871)
We're your Fabulous Learning Nerds, and we are out.