Ria Dispo

Pete Mockaitis
Donna, welcome back to How to be Awesome at Your Job.

Donna McGeorge
Thanks for having me, Pete. Great to be here.

Pete Mockaitis
Well, I’m so excited to be chatting with you about ChatGPT. It’s overdue, frankly, that we have an episode dedicated to this. And, Donna, you just happen to be an expert, and we already love you, so I’m stoked to be chatting again.

Donna McGeorge
Oh, look, it’s not as overdue as you think. I’m quite surprised at how I thought ChatGPT would be taken up by millions, well, it has, but it’d be over the place by now, but it’s been a little slower than I thought.

Pete Mockaitis
Well, yeah, I just saw some research from Pew, and this was maybe in July, so it’s probably a little higher by now, that in the US, about 60-ish percent of people had heard of ChatGPT, and of those who had heard of it, who have college degrees, which is most of our listeners in the US, about 32% of folks had used it. Does that sound about right from what you’ve seen in your research as well?

Donna McGeorge
Yeah. And so, my research is mostly standing in a conference and asking people to raise their hand. And so, when I’ve got a room full of people, and I say, “How many of you have heard of it?” It’s the same, it’s around 60-70% of hands up. And then, “Keep your hand up if you’ve actually used it or you are using it?” and the hands dropped considerably.

And then I’d say, “Now, who’s loving it and using it pretty much for their everyday world?” and then the hands dropped again in terms of using it consistently. But, again, that’s pretty anecdotal just from watching crowds.

Pete Mockaitis
Okay. Well, now, it’s just good to get a fresher test on that because it’s sort of, in my world, it comes up a lot. And so, it’s just good for context.

Donna McGeorge
I don’t know if I’m just a bubble but, like, everyone I know is using it but it’s like I go out into the world and I find all these people that some have never even heard of it, it’s like, “Wow, you’ve been living under a rock,” because that’s how it feels to me anyway.

Pete Mockaitis
And I think it’s funny, when it comes to many tools, I remember it had been years until I learned that I could enable shortcut keys in Gmail, and I was like, “How come nobody told me about this all these years? How many hours have been burned without me knowing I can enable shortcut keys in Gmail?” But then I did and I never went back, and, actually, I went with Superhuman to kick it up another notch. So, yes, tools can take a while to permeate.

And that’s what I want to talk about because there’s a lot of hype, and maybe why don’t we start with it since, hey, not everybody’s familiar. Let’s take maybe three minutes. What is ChatGPT? How do we get it and use it? And why bother?

Donna McGeorge
All right. So, I’m not going to go down the massive technical path that people are hugely interested in, the technical backend of it all. I’ll let them go Google that. But in terms of what you need to know to use it every day, it’s a large language model it’s been trained on. What that means is it has access to all published written communication up until about September 2021, so that’s all books, all articles, all websites, all research papers, pretty much everything.

So, the way I like to think of it is it’s like a librarian that has read every book ever written, read any paper, looked at every website, remembers anything, and can quote from it, ad nauseam, really. You ask it a question; it can pull from all of that knowledge to give you a reasonable answer. And so, look, it has its pitfalls. If you know a little bit about it, you probably read some of the negative stuff because it’s pulling data from all over the place, sometimes it puts stuff together that’s not true. The technical term of it is hallucination.

And, certainly, sometimes when I’ve asked it to give me references for various bits of stuff that I’m looking for, it makes up whole references, puts whole names of scientists together, and says they’ve written a paper, and they just never did.

Pete Mockaitis
Right, indeed. Well said. Thank you. All right. So, we’ve got an artificial intelligence, a large language model, it’s called, and it’s read the whole internet, or a fraction of it, and a bunch of books and stuff, and, thusly, you can interact with it. And so, if someone is like, “Whoa, that is cool. How do I do it? How do I get there? How do I play with it?”

Donna McGeorge
Okay, the easiest way is to go and register a free account with OpenAI, and you can get started straight away. What I’ll do is I’ll send you, and you can put it in the show notes. I’ve got a list of prompts across a range of different aspects of both our professional and personal lives, but it’s straight out of the book that I’ve written. But it can give people something to play with rather than just sitting there, looking at it, going, “I don’t even know how to start.”

Because it’s so big and it’s got so many potential uses that I hear someone, even today, say, “Hey, I used it for X, Y, Z,” and I’m like, “Well, I would never have even thought of that.” So, it’s not exactly something you can just go and sit in front of and start playing with. You got to have a reason. So, you go register for a free account. It’ll talk to you probably about the paid version. I would say 80% of my use it with the free version.

I’ve got the paid version but I mostly use the free version. And go play around with it with something that’s relatively harmless, like, I don’t know, meal planning and holiday planning just to see what it’s capable of.

Pete Mockaitis
That’s right. So, OpenAI.com, and you click Try ChatGPT, and so you can start taking a look. And you’re right, you just have a box, you can type anything into it, and then it will go for it. You can say, “Hey, tell me a poem about sand timers,” and it will tell me any number of things. I have a buddy who likes to say, “Hey, rewrite this email I just wrote but make it more polite,” and then it does the thing he forgets to say, like, “Hello, hope you had a great weekend.” He’s like, “Oh, yeah, I probably should’ve said that. Okay, yeah, thanks, ChatGPT.”

So, that’s really where I think you can be super useful here, is to help us understand, like, what is overly hyped and just sort of silly? Like, are the robots taking over mankind, they’re going to enslave us? Is this the answer to everything? We’re all going to get fired. Like, what’s too much hype? And what is really possible for us right here, right now that’s useful that can save us time and increase our results at work?

Donna McGeorge
Okay. So, the overhyped stuff is, now, look, it could be because I come in from a more Star Trek optimistic future of the world where humans are awesome and everyone is getting along nicely, and we’ve got the occasional attack from Klingons. Like, that’s kind of it. So, I have a more utopian view of the future so I don’t believe robots are going to take over the world. I think it could be a thousand years from now, an evolutionary marker where humans just get, again, a whole level of smarter than the technology that we’re creating, it could be.

I certainly don’t think we’re going to see a wholesale loss of jobs no more or less than any other technological breakthrough that has created some loss of jobs but created some new ones. So, I’m old enough to remember when the internet first came out, and all the kind of palaver we’re hearing now about it’s going to take jobs, it’s going to destroy the world, all of that started to happen when the internet became a thing that we carried around in our pockets, it was on our desktops, it was readily available.

So, it’s the same kind of technophobia that’s been around since, frankly, the printing press was a thing. And so, what we’ll see is new jobs or new ways of doing work. So, my advice around this is let’s not worry, “Is it going to take my job?” I can’t remember who said this but it’s been floating around for a while, “Don’t worry about it taking your job. Worry about someone who adopts AI and use the tools of this. They’ll probably take your job.” So, if you don’t keep up with it, then you’re at more risk of that than losing it straight out to AI.

Having said that, gosh, there’s some mundane administrative things that we do on a day-to-day basis that AI could be really helpful for, like any kind of repetitive processing or data entry that has a human looking at some kind of written file, and then typing it into some kind of system. I mean, there’s already systems that do that. So, I suspect that’s the kind of work that would go away.

Pete Mockaitis
And for right now, today, something that professionals can use ChatGPT to assist them with, I’d love to get your perspective. I know that some folks, if English or whatever language, is not their first language, and is maybe a little bit rough, they’ll say what they’re roughly trying to say in the language that they mostly know, and then ChatGPT just give it an automatic polish that has a little bit more smarts to it than, say, a spell check or grammar checks, so that’s sure handy. Tell us, what else are you seeing is super useful that folks are doing right now?

Donna McGeorge
Right. So, look, so I don’t even know where to start but I actually think this is a massive literacy gamechanger. So, you talked about English as a second language. I’m going to go back to the kids that struggled at school because written comms wasn’t that easy for them or they were dyslexic so they left early and have now considered themselves not terribly well educated, can’t spell, don’t know grammar terribly well, and that holds them back.

And so, this is an absolute literacy gamechanger so we don’t ever have to worry about that again. In fact, there are already stories about people who are saying, “Here’s what I wanted to write,” misspelled, no grammar, really poorly strung together sentences, whack it into ChatGPT, and it comes with a “Dear Madam,” like perfect kind of phrased email.

So, what I know, as someone who writes a lot about productivity, is one of the biggest bugbears of many people is email. And so, first of all, it’s volume, so volume of email, then other ChatGPT can’t help with that just yet. But, certainly, sometimes the time we spend responding to something that’s a bit tricky, so, as you said before, your friend that had to, “Here’s what I want to say. Now, make it sound slightly empathetic and friendly,” well, people are using it, heaps are writing emails.

In fact, I predict, once Microsoft get their act together with this, there’ll probably be a button in your Outlook email in the future, maybe Google as well, that you click on, that says, “Compose a response,’ and it will automatically generate a response for you. So, that’s going to save us a pile of time. But right now, you can already do that.

I even did it myself recently, a delicate no letter. Someone wanted to work with me, not really my thing, they were pretty insistent, so I said, “Hey, help me write a really delicate letter that’s assertive in my no, I don’t want to roll over, but maintains the relationship.” So, I’m going to say any written comms. So, as a business owner, I struggle with writing about myself, bios, website copy, email newsletters, social media copy, that’s the sort of stuff I struggle with, so I use it for that.

In corporate, so I’m hearing people using it for similar things, putting proposals together, writing about products, and getting ChatGPT to edit and get feedback on product descriptions. So, if you say, “Here’s a product description. It’s for this market. Can you please make it sound more attractive or irresistible to this market?” and it will then put the words in it. Because of our humility frames, we’re not terribly good at talking up stuff, whereas, ChatGPT is shameless. It’ll talk to your stuff up no end.

Pete Mockaitis
Yes. Oh, Donna, I think you’ve just really nailed something there. Our humility frames that we have as humans, ChatGPT does not, is not a human, and it can be shameless. And I think that’s great, you said, “Hey, I have to say no but in a way that’s polite.” I think, whenever you have some written communication with some emotional resistance, like, “Aargh, this is kind of complicated and tricky. I sort of feel like I owe them but it doesn’t work for me,” it’s like you could just say, “Hey, write an email response to this letter, copy/paste, or email copy/paste, that is very polite and says no.”

Or, emotionally, I remember I had a landlord who, I think, just had some unrealistic expectations for what a tenant was supposed to do. I won’t go into details. But at one point, I thought, I’m sure this landlord is going to drop in and find fault with all sorts of things, and demand that we do all kinds of things, like, “How about you repaint some walls?” I’m like, “I’m pretty that’s not my job as a tenant.” But whatever.

So, I said, “Hey, write an email from a landlord that’s utterly disgusted with the condition of a unit in the nastiest language possible, taking the role in the nastiest thing, whatever.” And so, what’s funny was so I read it and it was sort of like an inoculation or a vaccine or a preventative measure, because it’s like, “Okay, this is not real. This is not a real human but this is just the AI writing it up.”

And sure enough, once I read this harsh language that was AI-generated, later on I did get a harsh email from the landlord that was like, “Oh, I was expecting this, I prepared for this, and this isn’t so bad.” And it genuinely helped my emotional coping with that situation because I don’t like being judged and told I’m doing a terrible job at something in any context. That’s me.

Donna McGeorge
Right. So, it took the sting out of that because you had somehow prepared for it. And chances are, the actual email you got was nowhere near as harsh as the one that ChatGPT generated.

Pete Mockaitis
No, it was about half as harsh.

Donna McGeorge
Right. And lots of people are doing around things like feedback. And so, if I’m writing something, and I say, “Give me feedback on my style,” I don’t get offended by…I call it Charlie, by the way. I don’t get offended by Charlie because it’s just a robot, and it gives me really good structural, editorial, to-the-point, very distinct feedback.

Now, I don’t know why I would take it better from Charlie than I would from a human, but two things I reckon. One is the emotional aspect of working with a human. And, secondly, the human probably wouldn’t be that harsh with me. They’d probably couch it with a little bit of cottonwool around it or something like that. And so, that’s one side of it.

The other side of it is Charlie doesn’t get offended. When you tell it, it writes you something, and you say, “Actually, that’s a big pile of rubbish. I need it done this way, this way, this way.” And it apologizes, “Sorry about that. I’ll have another go.”

Pete Mockaitis
“I’m sorry, Donna.”

Donna McGeorge
Yes. Not quite in the dulcet tones of how, which is probably showing my knack for anything but, yes, no it doesn’t quite speak spit to me yet.

Pete Mockaitis
Okay. Well, so another’s thing. I guess now it can speak to you in the iPhone app and/or with some plugins. I haven’t played with that yet but I think that just might be handy like on the car. Like, you’re thinking about something, you could have a little bit of a conversation, I assume you could use the Siri activation button, I don’t know yet. Be safe, everybody, in your vehicles. Don’t look at your screen.

But I assume you can have some voice activation, get back-and-forth conversations, which can be handy so I can then read that later, and maybe actually get some good thought work done while driving, which is often hard to do because you’re not looking at a screen or a notebook to write it, so that’s really cool.

Donna McGeorge
Well, the thing that I find when I’m driving is accessing. Usually, if I’m driving, particularly long distances, which is just about everywhere in Australia that you’re driving long distances between one place and another. And so, I go into that really awesome alpha brainwave state, which is often when my creativity kicks in, so you’re absolutely right.

So, whether it’s, “Hey, Siri, make a note of this,” or, “Hey, Charlie,” or ChatGPT, “look, let’s have a conversation about this,” I think it’s a useful tool to think about. But you have done something else in here around, again, the aspect of literacy or creativity where some people say, “All right, I’m terrible at written communication but I can talk about my ideas.”

And so, maybe you’re struggling to explain in written form so you talk it through, record it into ChatGPT, ask it then to construct whatever output you’re looking for, like an awesome proposal letter or something like that, or lots of people are using it to help them get their resumes in good shape, their cover letters in good shape. But just be very careful that you edit it because it’s starting to get pretty obvious when people…there’s things you can do to not make it obvious.

If you’re not very well-versed with it, a straight copy and paste out of ChatGPT is a little obvious, it can be a little obvious.

Pete Mockaitis
That’s right. I found, because we’ve been experimenting with it in all kinds of contexts, in terms of writing episode titles and teasers and descriptions, it doesn’t do as good a job as we humans do, in our humble opinions of ourselves, but it can spark a few ideas, like, “Oh, that is a good turn of a phrase. Yeah, I’ll grab that. Okay. Oh, okay. I like that sentence there. Okay, we’ll take that.”

So, it can be a nice little starting point, and sometimes your copy doesn’t have to be smashingly captivating, it’s like, “Yeah, good enough. Good enough for this email response, copy, paste, done.” Other times, it can be a launching point. But one of my favorite little tricks is I’ll just ask for sheer quantity. I’ll say, “Give me 20 potential titles for this summary,” and then it’s like, “Ooh, I like this word from number two, and that word from number seven, and, thank you for your inspiration. Your work is done here,” even though I took none of the titles that it actually gave me.

Donna McGeorge
And that’s the shift that a lot of people are struggling to make because many think of it as something like Google, where you go in and you put a command in, and say, “Hey, give me a recipe for a banana bread,” or something like that. Whereas, you can go into ChatGPT, and say, “Give me 10 recipes for gluten-free sweetish snacks, and generate a shopping list that goes with that,” and it’ll give you the whole thing. So, the volume aspect is really powerful.

I did the same thing. I played around with blog titles. So, one of the tricks that I do is if I know my target audience, which is often women in leadership positions who want to level up, so I say, “This is my target audience. Give me a list of their hopes, fears, dreams, and aspirations. And then give me three suggested blog titles for each one. And then give me an absolutely irresistible captivating headline that will draw people’s attention.”

And, boom, before I know, I’ve got a quarter’s worth of social media, not copy because I’ll still go in and create much of it. I’ll get it to help me but at least I’ve got a plan in place and all my topics sorted out. And that sort of thing, this is where it really, for me, this is where I became interested in it. I became interested in the time savings that we can garner from it.

So, whether it’s spending time and energy agonizing over an email response, or the time energy to generate a social media plan for your target audience, or whatever it might be, research out of MIT says you’re saving at least 37%. I reckon it’s more for me. I reckon it’s taking me maybe about 30% of the time to do some of this stuff.

Pete Mockaitis
Okay, that’s cool. So, lay it on us, what are some of the hugest timesavers, and if you’ve got them, special prompts, that you’ve generated that had helped you realize these time savings using ChatGPT?

Donna McGeorge
Probably the first one is teaching. So, I’m a writer, so that may be a bit different to some of you, so I’m regularly writing articles and things. So, I’ve taught it how to write in my style. So, one of the very first prompts I did was I told it its personality, “You are an expert, a writing style analyzer. Analyze the following passages and play it back to me in a way that I can then,” this is a bit convoluted, “…that I can then feed it back to you so future writing will be in this style,” and so it made sense of that.

I then copied and pasted a couple of chapters from one of my books, and it spits out, and says, “You have a very engaging conversational style using anecdotes and rhetorical questions.” I go, “Awesome!” Now, in the free version, I then I copied it and say, “I need an article, 700 words, for this publication on this topic using this style. Get a start for me.”

Now, to make a start on any kind, like a blank page, whether it’s an email, or a proposal, or anything, that’s often the hardest bit to overcome. So, you never have to wait again. So, I go in there, it gets me a start, and then I’d say, “Rubbish first drafts are around 50% useful.” That’s one thing, teaching it to write like me.

Just FYI, the paid version now has an option in it where you can permanently put information like that, “Anytime I ask you to write an article, use this information.” And so, you can now train it with your stuff. There’s also plugins where I’ve been able to put PDF versions of my books, and I say, “Write this article using the following content from the following PDFs.” So, I don’t know if we’re about plagiarism now because it’s using my stuff.

So, there are a couple of timesavers for me, straight away, that means that I can generate good quality content, still human edited, in a matter of an hour. Whereas, it could’ve taken me half a day, to a full day, sometimes to write something of reasonable quality. So, that’s the first thing, any writing task. I would say anytime you’re stuck, as someone who’s done a bit of research into what happens in our head when we start to get overwhelmed, we end up in cycles that uses a lot of energy, and two hours of agonizing and we’re still having got more than a sentence on a page. So, whatever you’re agonizing over, ask it, and it will give you at least some response.

I think the volume thing is a good one because we can cut straight to the chase, “Give me 10,” you don’t have to ask for one, “Give me 10, 20, 30” however many you need. Things of a personal nature, “I’ve got my 55-year-old sister-in law, likes 1980s country and western music. Can you suggest 10 gift ideas for her that aren’t records, as in CDs or music or whatever, under 50 bucks?” because I’m a cheapskate. So, boom, all my Christmas shopping now will be a list generated there.

Meal planning. Holiday planning, “I’m about to jump in the car with a couple of pre-teens, what’s some great podcasts we could listen to?” So, it’s an entertainment curator. They’re all the things I’m using it for that just mean I can put my time and attention on the things that only a human can do in my life.

Pete Mockaitis
That’s good. I like these sorts of themes we’re collecting here. So, emotional writing, we’ve got some hangups, we’re humble, we’re awkward. Just starting the writing. It’s a blank page. It’s intimidating. “Give me a volume of ideas. Give me 20 options.” A curator of things. And with that, I like it how you could say, “Hey, I’m looking for the music,” for example, like, ‘80s music, or, “I want music kind of like artist A, B, C and D. Now give me some more,” which I think is pretty cool.

And, likewise, even with podcast guests, it’s like, “Hey, ‘How to be Awesome at Your Job’ is about this. We’ve had some guests such name, name, name. Who might be some others?” And it’s funny, it’s sort of like, “We had them, and them, and them, and them.” It’s like, “Well, we’ve already had them but thanks for trying, ChatGPT. You’re in the right zone.”

Donna McGeorge
That’s why sometimes I think of it as an eager intern. It’s eager to give you more of the stuff that you want but doesn’t always give you the right thing. But, hey, just quickly on the playlist thing, there’s a Spotify plugin now, so you can now give it access to your Spotify account. So, when you say, “I like artists like this. Give me playlist for this. Oh, by the way, then whack it into Spotify for me.” So, that’s pretty cool.

Pete Mockaitis
That’s good. Also, with curation, it’s cool. Sometimes I’ll say, “Hey, I’m looking to do this. Give me a great book that will help me do just that.”

Donna McGeorge
Or, even better, “Give me 10. Give me 10 great books.”

Pete Mockaitis
Exactly, yes. Like, “What’s a good book to help me reprogram my brain to enjoy effort, struggle, and mistakes?” ChatGPT recommends Mindset. And I say, “Can you give me 10 more?” And it says, “Oh, yeah. All right, here we go.” So, we got Grit, Antifragile, Ode to Happiness, etc. So, I just think that’s pretty darn handy.

And it works differently because sometimes it’s different than keywords because it’s, like, sometimes I don’t yet know the keywords, and sometimes I’ll specifically say, “Hey, what’s a word that means like Washington counts for like the seat of US governmental power that’s a figure of speech?” It’s like, “Oh, you’re talking about metonymy.” It’s like, “Oh, yes, thank you, if I’m even saying that right.” It’s like I didn’t know the keyword but it can generate that for me, so I dig that.

Well, maybe let’s flip it on its head now, Donna. We’ve been speaking breathlessly about how great this tool is. What are some of the limitations? What are some requests that it’s probably just going to fail us on, disappoint us on, and we would take its advice at our own peril?

Donna McGeorge
Well, first of all, it does make stuff up because it’s pulling information from all over the world, so it sometimes puts stuff together that’s not quite right. So, if you’re writing, for me, if I’m writing a book, and, by the way, to write my recent book, I did get Charlie to help me do that. So, if I’m writing a book, I still go back to Google to check my references and stuff like that because that’s important.

Look, I had to say it but it’s often about the quality of the prompts that we do that means that you get a bit of rubbish from ChatGPT. So, when people say, “Oh, I tried it once and I got a terrible response,” I’m like, “Well, did you go back and have a conversation with it? Did you tell it, it was terrible? Did you give it some more parameters?”

So, I got a bit frustrated this week because I was trying to get it to write me a story in the first person about some famous people, and it kept giving me almost obituary-style responses. And I asked it three or four times, and it still wasn’t getting it right, so I kind of pause, went off, had a bit of a break, came back, and re-crafted my prompt, and put the words in, “And I don’t want an obituary,” and I started to get the right thing.

So, occasionally, it kind of is smarter than its own good. It thinks that that’s what, in that case, it’s trying to outthink or be that eager intern that says what it thinks you’re looking for and add a bit of extra value, when I just didn’t need it to do that. So, for me, it comes back to the prompts. Certainly, if you’re looking for current data, just my point, asking it what the weather is like in San Francisco today because it will say, “No, I can’t help you with that because that’s not what I’m for.” I have had once, I violated their terms, the way in which they operated, there’s this little message that comes up.

Pete Mockaitis
Oh, boy, what did you do, Donna?

Donna McGeorge
I know.

Pete Mockaitis
You naughty.

Donna McGeorge
I was researching for a fictional book, and I was using a real-life person as the model for what I was doing. And so, I asked a pretty tricky question about this real-life person, and I think it was implying that I was either going to stalk him, or murder him, or something like that, I don’t know, so I had to go in, there’s a little, “Please explain” thing that I had to fill in.

So, on the one hand, that was a little bit frustrating, but, on the other, I was kind of encouraged by that. So, I would say that’s the efforts by the OpenAI folks to try and alleviate some of the fear that people have around it being used for evil.

Pete Mockaitis
Understood. And any other shortcomings you’ve seen? I guess I’ve asked it to write compelling headlines, it sort of fallen a little bit short but it just gave me a ton of options and I can mix and match and edit, and that’s cool. And I think I’ve also found that someone said it very well that AI can tell you, because that’s how it works, it generates the next most likely word to come, so it tends to give you the most obvious answers, as opposed to wildly creative answers, which I think I found that to be the case. But sometimes, the obvious thing is actually really handy, like, “Thank you. I should’ve thought of that.”

Donna McGeorge
Right. Yeah, if you need any obvious thing, it’s awesome, because sometimes I do that. I say, “Here’s a pattern that I’ve created. Here are two points in a pattern, because we know the world loves the rule of three. So, what could be the third? Give me 10 options for the third thing.” And the first one is the obvious one, and I go, “Of course. Why didn’t I think of that?” So, that’s useful.

But I’ve told it, if I’m really wanting something different, I’ll ask it to be a critic of whatever I’m talking about. So, one time, I was writing about burnout, and I said, “Be as if you’re a critic of burnout, and write me three paragraphs on why you think burnout, or what criticisms you have of burnout.” Anyway, so it spat out this piece that said, “Burnout is just a thing made up by people who are lazy, and dah, dah, dah.” And I’m like, “That was fun.”

Like, I would never use it in the article, but it was just fun to kind of get kind of like what you did with your “Write me the nastiest possible email from a landlord,” thing. It gave me a bit of, “Whoa, that’s interesting.” But I do ask it several times, “Give me nonconventional, give me something out of the box.” It’s not quite capable of doing that. It’s not able to do really, I don’t think, yet the really abstract stuff that the connections that a human mind can make as well. So, often when I say, “Give me this stuff that’s nonconventional,” I still get pretty conventional stuff from it.

Pete Mockaitis
It’s true. And there was a Wall Street Journal article, which we’ll link to, and I’ve mentioned in a recent podcast episode, about how AI was generating more and more new creative product ideas than some MBA students, and I had to read the full text of the paper, as I do, and I wasn’t that impressed. I guess it can generate a lot of ideas fast, no doubt. That’s very impressive that it can do that. But the quality of the ideas were like, “A portable printer,” I was like, “Well, yeah, that’s nice but you didn’t invent that. That exists, there are many options for that.” So, it can’t do that.

But what I do like is sometimes I’ll ask for it to engage me in metaphor, like, “Hey, let’s say that running a podcast is like captaining a ship in the ocean. What would be some of the elements of the metaphor?” And then that can help just spark my own ideas, and I find it handy in that regard. Or, I’ll say, “Give me advice on this issue as Yoda would, or as Tony Robbins would, or as Marshall Goldsmith would.” And so, that’s just sort of fun to say, “Oh, yeah, I guess Tony Robbins probably would say something kind of like that. David Allen probably would say something like that.”

So, sometimes I think that helps me a little bit in terms of it’s not groundbreakingly novel but it gets my own brain a shove in a direction to help me get to novel with a little bit of help.

Donna McGeorge
Yes. So, I’ve got it because a couple of my favorite writers, because I love the way they think, Malcolm Gladwell and Steve Levitt, so I’ll say, “This is the topic. What are some quotes, if they were writing the article, what might they say about it because I’m looking for a slightly different angle?” But you’re right, it’s usually something to give me a bit of a poke or a bit of a shove in a direction when I’m stuck.

And so, I think this kind of comes back to this idea of, “In what aspects of your world do you just get stuck and you end up wasting your time spinning your wheels because you can’t find an answer?” On the more emotional level, I had a woman recently say to me that she’d been using it to help craft responses to her ex who she was divorcing.

Because she couldn’t afford a lawyer, and a lawyer had said, “If you give us the basic information we need, we’ll then spend a small amount of time crafting the legal documents that are needed. But all the research-y stuff and all the kind of the backend stuff, if you can do the bulk of that, it’ll save you a truckload of money.”

And so, she was using, she told ChatGPT, “Act as if you are my divorce lawyer, ask me a series of questions to be able to fill in all the paperwork,” and she was able to get all the documentation that they needed collated, and saved herself a whole pile of money as part of her divorce, which I thought was quite…

Just quick disclaimer now. Please do not use ChatGPT for legal documents. You’d still need a lawyer to submit all that stuff, but, yeah, it was a real gamechanger for her.

Pete Mockaitis
I also recommended ChatGPT to a friend going through a divorce in terms of like, “I’ve got all these questions I’m supposed to answer,” I was like, “Well, for a first draft, let me show this.” He’s like, “Wow, that’s pretty impressive.” So, that is cool. You say when you’re stuck, I’ve also found it helpful when you’re stuck, when you’re researching and your search engines aren’t getting it done.

And it’s because, well, hey, the sad state of affairs of the internet in terms of searching is that many of the top search results are there very intentionally by companies with a budget who have hired search engine optimization professionals to accomplish that very goal, and they have succeeded. And so, you might not actually be getting the most useful information. It’s just like the most “relevant and authoritative in the eyes of Google” information, and that’s, in many ways, gamed intentionally. And not everywhere, and often it works just the way it should, and so we’re delighted with the result.

But sometimes I found, when I can’t find a product, ChatGPT can find it. Like, “I need to find a car seat that’s super narrow so I can get three across,” and it’s like I’m having a hard time finding that in Google and in Amazon, and then this thing is recommending, ChatGPT is recommending stuff that was not popping up in those searches, like, “Well, that’s very helpful.”

Donna McGeorge
Look, I’m going to catch this with at this point in time, there doesn’t seem to be that kind of product bias. Like, if you pay someone a chunk of money, your products end up being at the top of the list no matter what search criteria is put in there. I would agree. But I’d also say the risk is that we treat it like a search engine because it’s not a search engine. It’s someone you’re having a conversation with.

It’s more like you’ve got someone sitting next to you that you turn around, and go, “Hey, I’m really struggling finding this product, and, clearly, my search string in Google is not working. Can you help me maybe redefine what might be the parameters I need to get Google to work better for me that also bypasses all the paid ads so I can actually get to the product that I need?” and get it to help you craft your Google search right.

Pete Mockaitis
Yeah, it is. I think about it, I read somewhere, it is an intelligence but it’s an alien intelligence, and I thought that was well said. And I like to think of it as my alien intern who has read the whole internet or a good chunk of it, and so it’s like, “You’ve read a lot of stuff, alien intern. What do you think about this?” Alien because it’s got to be different than human, so watch out. And intern because, “Hey, I’m in charge. I am never going to blindly copy/paste what you say. I’m going to, at the very least, read it, and most likely edit, pick and choose, edit heavily.”

Donna McGeorge
Right. What you’ve actually got is an alien intern that has a hangover.

Pete Mockaitis
Oh, yeah.

Donna McGeorge
And so, they’re keen and eager to do the work for you. It’s going to be a bit nonhuman and alien in its form, but you better check it because sometimes it’s just a bit dim on certain days if it’s had a big night the night before.

Pete Mockaitis
Well said. An alien intern with a hangover. We can get some AI to generate art to that effect for us as well. That’s a whole another episode, I guess.

Donna McGeorge
That’s a whole another episode.

Pete Mockaitis
Well, Donna, any final thoughts before we shift gears? Well, first of all, your book is called The ChatGPT Revolution. Got to make sure that title gets in there, right? It sounds like it talks about any sorts of things. Anything, specifically, you want to mention about the book proper?

Donna McGeorge
Only that you can grab it from any reputable online bookselling place, on my website. But, look, I’d say when my publisher approached me to write it, we were like, “Well, what’s the angle we want on it?” So, it’s very much “Get me started, I’m interested.” It’s probably already, in fact, I know this passage in it, it’s already a little bit out of date because the technology is moving very quickly, but it’ll get you started and get you going, and get you interested.

And in terms of, I think you started to say what would be a last message, I think, around this or any kind of last comment I’d say, I would say get interested, get curious, and a saying we have in Australia is just have a crack, have a go at it, go in and try it out, and play with it. Be playful at first, which is why the list of prompts that’ll be in the show notes, are not terribly earth shattering but they’re something to get you going with.

And you can’t break it. It’s not like you can get in there, and go, “Oops, I broke ChatGPT.” The worst that can happen is you just get a somewhat rubbish response, in which case you tell it, “Gee, that was a bit rubbish. Try again from this angle.” So, have a go, have a crack is what I would say.

Pete Mockaitis
All right. Well, now could you share a favorite quote, something you find inspiring?

Donna McGeorge
From David Allen, you mentioned him earlier, “The human mind is for having ideas, not storing them.”

Pete Mockaitis
And a favorite study or experiment or bit of research?

Donna McGeorge
So, Taylor’s study, it’s called the pig iron studies from the late 19th century. I love him. He discovered that you can actually achieve way more if you take plenty of breaks throughout your day.

Pete Mockaitis
And a favorite book?

Donna McGeorge
Apart from my own, I can’t stop thinking about Four Thousand Weeks by Oliver Burkeman. Awesome.

Pete Mockaitis
And what do you love about the book?

Donna McGeorge
Look, it just got me thinking differently about the finite nature of time. And he has a really interesting angle around settling. So, we’re told nearly all our lives, “Don’t settle. You could always go for more.” And his position is, “Well, why wouldn’t you settle and make good with what you’ve got rather than constantly seeking this better job, better relationships, better something?” And I’ve not stopped thinking about that, actually. That’s really got me going.

Pete Mockaitis
All right. And a favorite tool?

Donna McGeorge
I’m afraid I’m going to talk about my own, which comes from one of my books, called The First 2 Hours, which is a way in which I think about how I do my work and how I do my to-do list. There are some things that are better to do in the morning and some that are better to do in the afternoons. So, I’m happy to share a PDF of that tool as well in the show notes if you like.

Pete Mockaitis
Oh, yes, thank you. And a favorite habit?

Donna McGeorge
Early to bed, early to rise. That’s me. I like to get plenty of sleep.

Pete Mockaitis
Absolutely. And is there a favorite nugget you share that really seems to connect and resonate with folks; they quote it back to you often?

Donna McGeorge
Yeah, pay attention to the clock in your body, not the one on the wall.

Pete Mockaitis
All right. And if folks want to learn more or get in touch, where would you point them?

Donna McGeorge
Just my website, DonnaMcGeorge.com. And I’m a shameless self-promoter, you’ll find me on my social media platforms, and my name is a bit unusual, Donna McGeorge.

Pete Mockaitis
Okay. And do you have a final challenge or call to action for folks looking to be awesome at their jobs?

Donna McGeorge
Absolutely. It’s a version of something that Sean Patrick Flanery said, which is, “Do something today that your future self will thank you for.”

Pete Mockaitis
All right. Donna, thank you. This has been a treat. I wish you much luck and fun with ChatGPT and all you’re up to.

Donna McGeorge
Thanks, Pete. Great to be on the show. Thanks for having me.