If you have been wondering how to get started with AI at work, you are not behind. You are not in a technical deficit. What you are dealing with is an organizational communication failure: your company bought the technology and skipped the change management. And that gap has landed squarely on you.

In this episode, I break down exactly what is happening and why so many smart, capable people are sitting at their desks feeling stuck. Senior leaders are pushing for AI adoption. Employees are expected to figure it out on their own. The result is a workforce split between super users building AI agents and everyone else quietly closing the tab.

I share four strategies to get you started, no tutorials required. First: pick a specific problem you already have, not a platform to master. Second: start where the stakes are low so you can experiment without consequences. Third: use AI as your thinking partner, not your ghostwriter. I walk you through a personal example from my own life (involving a difficult email and my homeowners association) that shows exactly how this works. Fourth: set your own learning boundary, because no one is going to build that on-ramp for you.

I also share how I use AI in my own work, including how I use podcast analytics and AI together to decide what episodes to record next, how I use AI to stress-test my episode outlines, and why I never cite a statistic AI gives me without tracking down the original source.

Listen and catch the full episode here or wherever you listen to podcasts. You can also watch it and replay it on my YouTube channel, JanelAndersonPhD.

LINKS RELATED TO THIS EPISODE:

Episode 215 - The Latest in AI: Autonomous Agents Working for You

Episode 179 - AI in the Workplace: Will it Replace Humans?

Episode 150 - 7 Things to Try in AI

Episode 149 - The Power of Generative AI to Transform Meetings

Episode 95 - AI Won't Steal Your Job, but Here's What it Will Do

My company is using AI, but I don't even know where to start. Every time I sit down and try something, I just feel overwhelmed, and I close the tab. How do I even get started with this? Well, if that sounds like you or one of your employees, I've got you covered.

People are feeling overwhelmed about using AI at work these days, and so much anxiety is hanging out around using AI. You see, here's a little bit of the context, senior leaders want organizations to adopt AI, but for the most part they're not giving any training or any practical advice about how to get started. So middle managers and frontline employees just don't quite know what to do. A lot of folks are feeling behind on AI at work, and what I see with the people that I talk to is that is very much of a bimodal bell curve.

We've got people who are really super users and are creating AI agents. On the one hand, I'm in that camp myself, but then you've also got folks who are just like brand new and don't know where to start, and so what I aim to do in this episode is to really help those of you who don't know where to start, especially if you're feeling pressure from your organization that you're supposed to be using it and you just don't know what to do or where to start. Now, if you are already using AI, I still want you to listen to this episode because I can almost guarantee that you are a manager who's got someone in that latter category who is supposed to be using it, or maybe just wants to use it. It's not even a mandate, and they just don't know where to get started, or maybe it's a colleague, or a friend, or somebody else that you know who could really benefit from this, and who knows, maybe even if you are a super user, you pick up a couple of tips yourself, either for what you might do differently, or for how you can share how to use AI a little bit more easily and strategically with folks around you who aren't quite as savvy as you are just yet.

So this is a little bit about getting started with AI at you at work, but I really also use AI myself a lot as a thinking partner, so I'll be, and I'll be talking about that, and how to strategically use AI as your thinking partner. Now, I also want to just preface, before I get started into any of the nuts and bolts, that I am speaking platform agnostic here, so I don't care if you use Claude or Gemini, or Copilot, I don't, or any of the other half a dozen ones out there that are common household names these days. So, this is completely agnostic to any tool you can use this in every tool. So, if you don't know where to start on AI, or you know somebody who doesn't know where to start on AI, or your whole company wants everybody using AI, and you want to give folks some guidance.

This episode is for you. Okay, so every time that you might have said something like this, like, how do I even get started with AI, or thought it, I just want you to know you're not behind, you are not in a technical deficit, and you're definitely not alone. This is one of the most common things that I've been hearing from technical leaders that they've either got staff who don't know where to get started or maybe they don't even know where to get started themselves. So, again, I am going to give you a way through it today, and I'm going to make it really easy and break it down, and I'm going to give you an example from my life that is a place where AI really helped me recently, and I think it's a very easy and relatable example.

It's not something super high tech, okay? It's an email that I needed to write that I was completely putting off, and I was having a lot of emotion around this email, and so I'm going to tell you how AI helped me. So that's where we're headed. I'll use that example throughout.

Now, again, here's what I want you to understand before we get into the how and the nuts and bolts overwhelmingly articles and research reports are coming out saying senior leaders want their staff to use AI, but when we peel back the curtain and when we look under the hood, we find that those same organizations are not training their staff at all. I have a training that I'm rolling out, so if you're, if your organization is interested in it, let me know. But again, it is very agnostic. It doesn't matter which tool you're using, but it is a how to get started with AI as a thinking partner, how to be more strategic about AI.

I'm not talking about like how large language models are constructed or any of that. This is just very much a hands-on kind of thing, and we can easily customize it to whatever work, whatever tool your organization is using. Okay, but again, senior leaders are really wanting their folks to be using it more, and folks who are boots on the ground just don't know where to get started. So, I don't want you to feel overwhelmed by that anymore, and I certainly don't want you to feel like it is any sort of a personal failure on your part.

If anything, it is an organizational communication failure, and that's where my expertise lies, is organizational communication. So, again, I'm going to walk you through it, and if it's something that your whole organization needs, let's talk. All right, so senior leaders, and you know, organizations are heavily investing in AI right now, and just simply expecting adoption, and talking about adoption as if it is something that is just easy to do and easy to pick up, but again, the training is not being provided, the context of how to use AI in your day-to-day operation is not being provided, and there is no on-ramp for most employees, in terms of how to integrate AI into their workflows, so your company bought into the technology, but skipped the change management piece. That's what I'm finding with a lot of organizations, and that gap, unfortunately, is landing on you, and that is not fair.

Or, if you are one of those managers who has a grasp on it yourself, but your team just isn't quite getting on board again. It's still landing on you because your staff isn't up to speed. Okay, but this is the reality that we're in, even if it's not fair, and even if it's not your fault. Okay, so the question underneath the question here is not really, how do I use AI?

It's, how do I build some confidence with something that I know nothing about that somebody just handed to me, and that again is a communication and change management challenge, and that is something that I can squarely help you with, that is in my wheelhouse. So I am going to walk you through four ways to get started, and kind of a four-step process, or four different ways to think about it. So, the mistake most people make when trying to learn AI is to treat it like a category, like I need to learn this piece of software, just like maybe if you were learning. I mean, imagine all the way back to the first time you used a spreadsheet, and you didn't necessarily, and I know, for some of you, that's so long ago you don't even remember how you learned to use a piece of software like that.

Maybe it's something a little bit more like a project management tool, or something that you actually had to, like, watch a couple tutorials on, and learn step by step how to do something. Okay. Well, I want you to think about AI completely differently. I don't want you to think about it as a software product.

I want you to think about it as a helper. Okay, so it is a helper. It is not a software product that you need to learn. Okay, now there are lots of features in the different platforms, and you, the more confidence you get, the more willing you will feel to explore some of those things on your own, but this is just really how to move from being stuck to getting started.

Okay, and as I said, I've got four different ways for you to cut into this. So, the first one is I want you to pick a problem, not a plat, not a platform. I mean, if you have multiple platforms like Chachi BT and Gemini and Claude at your disposal, then just randomly pick one. Okay, it doesn't matter which one you choose.

Now, a lot of times, though, if you're in a large organization, a particular license has been rolled out in your organization, in which case I want you to use that one that is sanctioned by your organization and licensed for your organization. All right, but if you have multiple to choose from, just pick one. It doesn't really matter, that's not what we're teaching you here, the ins and outs of any specific platform. This is, as I mentioned before, platform agnostic.

So, I don't want you to think about learning AI broadly. I want you to think about applying AI to a specific business problem or task that you have. It doesn't even have to be a major business problem, just a task. Okay, so pick a task that already frustrates you or slows you down, or maybe like the example I'm going to share, one that I had been putting off.

So it's a problem, a task, something you already have on your plate. So we're going to apply AI there first. Okay, so the people that I see who are gaining real traction with AI at work are not the ones who sign up for a course or watched a dozen training tutorials on YouTube. They are the ones who said, I hate writing status updates, and then they started using AI to help them write their status updates, or they are the ones who said, I always procrastinate on writing performance review self evaluations, and then up against the wall, up against the deadline for writing their self evaluation, they get started with AI.

Okay, and then they've used it for that one thing that put in a repetition, then they try it for something else, they get another rep, another rep, and the more reps they get the more confident they get, and so that's what I want for you. If you're hesitating, or if you've got staff members who are hesitating, and you're sharing this with them, whether you're sharing the podcast episode with them specifically, which I highly recommend, or taking the strategies that I'm talking about here today and just relaying the content yourself, either way is good, but I prefer they hear it from me. Okay, so you're going to pick a problem, something that's been frustrating you, something that you've been putting off, and now you've narrowed it down to something that is absolutely manageable. Okay, so that category feels smaller once you have something to start working on.

Okay, so that's just your first strategy, it's just pick something to apply it to. Okay, you haven't even had, you haven't even opened the AI tool yet. You've just done the first thing, which is pick a problem. Okay, so I want you to also think about choosing a problem that is relatively low stakes.

Okay, so your step number two here is start where the stakes are low. Do not start with your most visible deliverable, or something that is going to get the attention of senior leaders right away, that's due into them. Okay, you're going to start totally low stakes, something easy peasy, something where you can get a quick win for yourself, maybe for your team, for your manager, but you're not going like all the way up the org chart on this. So start where you can try and experiment and adjust and learn with very low consequences.

Okay, again, I promised an example, and here is my example, and I'm not going to share a work example, although I use AI a lot in my work, and I'll share a couple of ideas about how I use AI as I continue to talk through this in my work, how I, yeah, how I use AI in my work, as I talk through this other personal example that really doesn't actually have anything to do with work, but it's very similar to a problem you might have at work. Okay, so I've got my low stakes problem. Here's my low stakes problem. I am on the board in my homeowners association, and recently I had this difficult email that I was dealing with a resident in our community on, and I needed to reply back to this homeowner who was upset about something, and I'm not going to give away any more details in this podcast at all, because some of my neighbors might be listening to the podcast.

Okay, so I don't want to out anyone or get anyone in trouble, myself included. Okay, so I'm not going to say what the issue was, or anything like that, but this homeowner was upset about something, and it was my job in my specific role on the board to get back to this person, and I was irritated just by the nature of their email, because I thought, like, come on, like, you live in an association, you should expect things like this to happen, this is not a big deal, so I was irritated first that the I was irritated that the other homeowner was irritated. I was irritated that I needed to deal with it. I was irritated, like all the things, and so I had put off this email, and I knew that putting off this email was only going to make the situation worse, so I needed to get on it all right, so again I had been putting it off for days, not weeks, thankfully, because that would have gotten me probably in more hot water with this homeowner, but I had been putting it off for a few days, and again it was a low, very low stake situation, nobody at work was watching, even the other board members didn't necessarily know that I was dealing with this particular homeowner, it wasn't anything I needed to escalate to anybody, it was just like in my job on the HOA board, I just needed to deal with this, so it was relatively, you know, low stakes, and so the perf, and something I had been procrastinating, so the perfect starting point, okay, it meets these first two criteria: something that you, that needs to get done that you don't really want to do, and something that's relatively low stakes.

Okay. And so I'm going to tell you exactly how it went, because it leads perfectly into my next point. Okay, and that is point number three, and I've already alluded to this, but that is use AI as your thinking partner. Now, most people think of AI as a tool that is going to do something for them, like write them an email or draft, you know, write some code or something like that, but I don't want you to think about it like that.

I want you to think about AI as your writing partner, maybe your coach, if you will. So I should say, I shouldn't even say your writing partner. I want you to think about it as your thinking partner first, and then it can be your writing partner after you've thought through what it is you're going to write about, or whatever the task at hand is. It could be designing a slide, it could be, you know, again, in my case, drafting an email.

Okay, so the most valuable thing that AI did for me in this particular situation had nothing to do with speed, nothing to do with even really not that much to do with drafting the email, because I drafted the original email. So, here's what happened. I had been avoiding this email because I was genuinely frustrated, the homeowner was upset, and honestly, I was a little upset right back, and I knew that I couldn't match their tone in the email, even though I wanted to, because their tone was, was their tone was upset, I'll just call it upset, their tone was upset, and I knew I could not be upset back to them, that would not help the situation, and it would not resolve the situation, so I knew I couldn't match their tone, that was part of me, like putting a pause on it and waiting for a day or two before getting back to them, but again I had waited probably a day or two too long, and so by the time I finally sat down to start drafting something, I kind of had overcorrected, because what I did is I was like, okay, let me think more critically about why they're upset about this, and the more I thought about why they might be upset about this particular thing, the more empathetic I became, and I was like, "Wow, okay, I probably hadn't.. I, well, I definitely hadn't been thinking about it from that way before.

And so, you know, I puzzled through it, and I got really empathetic to what they were upset about. Now, then, I drafted my email, and I, quite frankly, was a little worried about sounding defensive myself, or, you know, because initially I had swung pretty far towards, like, putting them in their place and telling them what for, but then again, in those cooling off days, I got a little empathetic, I think. I swung too far to the other end of the spectrum and over corrected, because I was really, really intentional about not being defensive and trying to be empathetic, and so on. And so then I read my email draft before I sent it to them, and I thought, "Oh, I think I might be over accommodating here, I think I might just be a little too much, a little too nice, a little too much in their balls, in their court, kind of thing, or too much understanding, because there was, there is something in our homeowners covenant that they actually had to do, and so I feel like my email was like maybe I was gonna let them off the hook or something, although I don't necessarily think I thought that, but here's what I did, is I took my draft, and I went to my AI tool with it.

I brought the draft to AI, and I said, "Here's the situation: the homeowner is upset about this, and I just explained the situation that the homeowner was upset about in just a couple of sentences. I didn't copy and paste the homeowner's email in. There was no.. and you know, and again, when we get into stuff like that, we start to think about, like, what is the tool you're using, and what privacy protections are around it, and is that information then available as part of the AI's training, and so forth, and so I thought, like, I don't want the text of the actual email going into AI, they don't need that, they just need enough context to know what my response is about, so again, two, three sentences of context, and then I copied and pasted the draft of the email that I was about to send right out of my email and into AI, so that AI could assess it. So I, so what I said to AI is, here's what I wrote, here's the situation, and I explained the situation.

I want to be empathetic, but I don't want to concede anything that I shouldn't concede. Help me find that balance. So that was my prompt to AI, and I wrote, or I not wrote, I did a couple of episodes about AI a long time ago, and we'll link those up in the show notes, but writing a really good prompt is also one of the most critical keys to getting the best out of AI, so give AI a lot of context, whatever task you're having it do, give it a lot of context, because the more it really understands the task at hand, the better job it's going to do in getting it right when it produces something for you, so that was my prompt, essentially. And I mean, I don't have the exact prompt in front of me, but that was more or less what I wrote.

And then I waited for AI's response, which takes, you know, like four seconds, and it, and what it did is it didn't rewrite the email for me. Now I did not prompt it to rewrite the email for me. What I asked it to do was assess the email for empathy and if I was over correcting, and so it pushed back on some places that I was being vague, and so it operated as a thinking partner, and it said when you use this phrase, you're giving the homeowner a potential avenue to push harder and get what they want, and if what they want is counter to what is in your hoa covenants, then they can get that. So, don't you know, don't open the door an inch, basically, is what AI told me, and so that's what I mean about using AI as a thinking partner, because it helped me see clearly that I was opening the door an inch or so, that they could then push further and further on, and so I did ask it then.

Okay, can you help me tighten up my language in that sentence, so that I'm still being empathetic and understanding to their situation, but then I'm not opening the door an inch for them to potentially take a mile, and so then it came back with some exact language, and I didn't use the exact language exactly as it was written, but it came back with some more, you know, precise language about how I might say that, and I was very careful to make sure it still sounded like me, but not to soften it, because that would be my temptation, is to, like, you know, take the rough edges off of it, so it didn't, you know, didn't really have rough edges, per se, but it was firm and polite, and I just made it sound a little bit more like me, without softening the language or giving them an inch to push through and take a mile, so that's what I mean, again, by using AI as a thinking partner, not as your ghost writer, so you bring a draft and your context and your intention, and you ask AI to evaluate what you wrote, and you know, make any potential suggestions or give you its assessment. So this works for emails, yes, absolutely, but it also works for meeting agendas, slide decks, that self-evaluation on your performance review, that you might have to do project proposals, anything like that, any situation where you generally know what you want to say, but you're stuck on how to say it, or you've been procrastinating, or sometimes, like this situation was for me, you're just a little too close to the situation to see it clearly. Okay, so that is trick number three, or idea number three, and that is, use AI as your thinking partner, not your ghostwriter. Okay.

And then number four, tip number four, set your own learning boundary, so nobody's going to give you the dedicated time to learn AI. You don't get to take half a day off and go take a course on it. It's not going to show up on your calendar if you wait for your organization to build you the on-ramp, I guarantee most of you will still be waiting a year from now, and they might be frustrated with you for not using AI, and yes, I know they're not giving you an on-ramp, they're not giving you strategies, or like these, or even training, so you have to claim the time yourself, and it doesn't have to be a lot of time, you know, 20 minutes once a week protected, just like you would a small meeting, no multitasking, no, I'll get to it later, just a small consistent practice where you're actually getting an important piece of work, and again, a very low stakes piece of work, but an important piece of work that you need to get done anyway, because I don't want this to be yet again another thing, and I think that's where a lot of the anxiety that people are feeling about when they're like being pressed to add AI to their workflows is they're like, oh gosh, that's one more thing I got to figure out, I don't have time for the one more thing, so again, this is where I want you to take AI and apply it to something that is already on your plate, so that you can make some good headway and progress with that. So, again, it is a great thinking partner, and I use it as I promised I was going to say a little bit more about how I use AI in my work.

So, before we wrap up, I will share just a couple of ideas about that, and then it probably merits its own episode, since I haven't done a lot of episodes on AI recently, I did a few right when Chat GPT first came out, and then when Agentic AI first came out, I did some episodes then, and again, we'll link those up in the show notes, because those are still relevant and valuable. But the four things that I talked about today, the four tips and strategies: pick one problem, one issue that you need to work with, don't pick a platform, pick a specific issue, something you've been procrastinating, or that you just don't like doing that much. Number two, pick something where the stakes are low. Okay, this should not be a big, important project.

Stakes are low. Number three, you're going to then actually use AI as a thinking partner by writing a prompt that gives it some context and how much you've done so far, or what you're thinking about it is so far, and then asking for its collaboration on thinking about it with you, not necessarily writing anything for you. You can then later ask it to assess your writing, or to evaluate your writing, if you really do want that word by word, or sentence by sentence perspective, like I did. And then number four, set your own learning boundary, pick an amount of time that feels reasonable, and then apply that time again to something that you're already working on anyway.

Okay, so those were your four strategies. Now, I mentioned I would use AI like that, I use AI like a thinking partner in my business. Here's how I use it related to the podcast: I may have like three or four or five different ideas about what I might want to do a podcast episode on, and here's what I do. I go into the analytics of my podcast platform, and I look at the download analytics, and I share the download analytics, say from like the last 10 episodes with AI, and I say, here is what people have been downloading, here are the here are the downloads, and then if you're on my email list, you know that I send an email every Tuesday that says, "Hey, a new episode of the podcast came out yesterday, because the new episodes come out on Monday.

Here's what it's about. So I might give AI the analytics data of the downloads, and then the analytics data from my emails - how many people opened, and then of those who opened, how many clicked through. So now AI has some good context about what's going on in my audience, in terms of who is, how many episodes are getting which number of downloads, and what is my closest audience. That would be the people who are on my email list.

And if you're not on my email list, we'll put that in the show notes too, so you can click through and get on that list. So, what is my closest audience thinking about the episodes based on how many emails they open and how much click through there is, and then what is my more broad general audience in terms of like the downloads writ large? What does that data look like? And then, so I'll give that to AI, and I'll say, here's what's here's this is the latest of what's been going on in on my podcast, and I'm considering the following five or six topics as I figure out what to record next, so I'm giving it the data, and then I'm giving it my ideas, and I, and I'll say something like, my specific prompt will be after your analysis of the data, both the downloads from the podcast platforms, as well as the open rates and click through rates from the emails that I send, what would you suggest that my audience wants to hear me talk about next from these six ideas, and then I list out those six ideas, and then I will also say, and if there are other ideas that come to mind that are in my wheelhouse, because I have trained my, my particular AI tool that I use to know exactly what my wheelhouse is, what I am qualified to speak on, what I am not qualified to speak on, what's fun for me to speak on, but requires a little extra research, and so on.

So AI, my AI agent knows that about me. So then I'll say, you know, what do you recommend for my next, say, three episodes, and it will spit out its, you know, recommendations again based on the data that I gave it, and, and, and the ideas that I had already. Then, once we get an idea in motion, then I will write an outline, and I'll give the outline to AI, and I'll say, what am I missing, or where would somebody push back on this idea, and then again I'm using AI as my thinking partner to make the episode more robust. Sometimes, if I do need research, because a lot of my episodes do have research in them, I will ask AI to research a topic for me to find the latest statistics on something, and then I always ask AI to give me a list of sources, along with links to those articles or reports, or wherever it found the research that it gave me, so that I can go find the original source and make sure that AI was not hallucinating, because I'm sure you've heard of hallucinating, that is when AI makes stuff up, and it's happening less and less, but still, if it's something important, you need to check and make sure you've verified your data.

So I'm not going to cite a statistic on this podcast or in a keynote that I give that AI gave for me without me following up, tracking down, and actually looking at the research study or the industry report or wherever that statistic came from, so that's just a little bit about how I'm using AI these days as a thinking partner in my business, as it relates to my podcast. Now, you could probably translate that to when I'm working on keynotes, when I'm working on training curriculum as well. I'm always coming to AI with the context and with my draft, and basically asking it to poke holes or challenge my thinking, that sort of thing, and I have a, as somebody who primarily works alone, unless I'm collaborating with a client on something, I have just really enjoyed the company of AI, not like, you know, not like it's a person sitting next to me, but just like the banter of ideas and the exchange of intellect, and I don't know, intellect is maybe pushing it with AI, I mean it's pretty smart, but I'm not going to call it intellect because that feels like it's a little too human, but but just the back and forth of ideas has been a really welcome change in my professional life, in my professional world. All right, so the four steps that I talked about in the podcast, they will work for any place you want to experiment with AI at work, whether that is drafting a meeting agenda, tightening up an email, like I was doing, building out a slide deck.

Now, and as I, as I say, for slide decks, most AI, unless you're using something like Gamma dot app, most AI is terrible at building slide decks, and even Gamma, which I have a subscription to, and I use for maybe I have to talk about how I use Gamma in a whole separate episode, but when you're using AI for slide presentations, here's what I do: I give AI a prompt, I don't ask it to write my slide, I explain what my slide needs to be about, and what the context is around it, and then I ask it for, you know, suggestion on visual treatment, or, you know, I need a three step - this is a three step process I want to illustrate on this slide, something like that. So, I know I'm going kind of down a rabbit hole here when I'm trying to wrap things up, but I did want to just add that on about building a slide deck. Don't just ask AI to build a slide deck for you, but you could ask it to, you could also give it a screenshot of a slide and ask it to evaluate that, to see if it's too many words on the screen, or if it's clear, if it makes sense, if you've got like a model or a drawing or a diagram or something like that. Okay, but so again, whether you're drafting a meeting agenda, building out a slide deck, drafting an email, writing yourself an evaluation for your performance review, or again, finally doing that task you've been putting off, like I was.

I want you to take on AI as your thinking partner in it. So, pick one problem, start where the stakes are low, use AI as your thinking partner, and not just a ghost writer. And then find some time every single week, or if you want to really challenge yourself every single day to be using AI to help you with some small task like that. So, your action this week: identify one work task that you have been putting off because it requires either writing something or doing something you don't want to do, or whatever that is your starting point, and if none of that is a problem for you, and you're using AI all the time, here is your action for this week.

Find somebody who has been putting off using AI and share this episode with them again, whether you summarize it for them in a conversation, or whether you just share the link on Spotify or podcasts, Apple Podcasts, or right from my website. This is episode 270 so that would be Janelle Anderson forward slash 270 for the show notes that has the embedded player in it. You could also find it on YouTube on my YouTube channel. So, if this episode was useful, share it with someone on your team who's also feeling stuck.

If you're, if you excel at this, share it with someone who is not as far along as you are, and if you are leading a team or an organization that's trying to get AI adoption, you know, let me know, because we can talk about that more one on one and see if a training program is right for you. All right, my friends, if you're watching on YouTube, make sure you hit the subscribe button, and if you're listening on a podcast player, make sure you hit the subscribe button there too. Like and share the episode, it is a great way to support the work that I'm doing in the world, and it's zero cost to you. All right, my friends, be well.

I'll catch you next week. Bye.

Download Full Episode Transcript