What Does It Mean to Be an Author Now? with Jonathan Stark

How this transcript was made

Transcribed by machine from the published audio, then corrected by hand. It is not a verbatim record: the changes below were made, and are counted.

This episode

62 speaker turns, 303 changes.

• closing CTA identified as CTA 1 - Share (match 100%), replaced with the authored script and attributed to Max
• 2 mid-sentence speaker change(s) left as transcribed: the rule could not settle them safely
• ~2:49 - speaker corrected by ear (owner, 2026-09-25): Jonathan Stark speaks from "it's like calling yourself cool".

What was changed in these transcripts, and what was not

Generated: 2026-09-25 from 2026-09-25-named/, itself transcribed from the audio that is live on the feed. Every change below is mechanical and counted. No sentence was reworded, no claim altered, nothing removed for being awkward.

The editorial decision

Owner's choice, 2026-09-20: light clean. "Um", "uh" and stammered word repeats ("and, and") are removed, because nobody intends to say them. "You know", "I mean", "kind of" and "sort of" are KEPT - they run to roughly 2,300 instances across the catalogue and stripping them would make the page read as somebody else.

Counts across the whole set

Change	Count
stammered repeat removed	667
um / uh removed	654
space before punctuation	285
split contraction	136
organization -> organisation	71
ai -> AI	57
sentence tail returned to its speaker	26
synthetic outro restored from script	21
organizations -> organisations	18
hyphen followed by space	15
realize -> realise	14
A.I. -> AI	10
recognized -> recognised	9
gray -> grey	9
summarize -> summarise	9
color -> colour	9
behavior -> behaviour	8
synthetic intro restored from script	8
analyze -> analyse	7
diarisation fragment returned to the host	7
behavioral -> behavioural	7
recognize -> recognise	6
percent -> %	5
personalized -> personalised	5
skeptical -> sceptical	5
analyzing -> analysing	5
center -> centre	5
optimization -> optimisation	4
realized -> realised	4
favorite -> favourite	4
realizing -> realising	4
canceled -> cancelled	4
recognizing -> recognising	4
skipped passage restored	4
summarizing -> summarising	4
speaker corrected by hand	4
labor -> labour	4
behaviors -> behaviours	4
colors -> colours	4
organizational -> organisational	4
announcer turn relabelled from Guest	3
customized -> customised	3
analog -> analogue	3
optimize -> optimise	3
honor -> honour	2
optimizing -> optimising	2
summarizes -> summarises	2
personalize -> personalise	1
prioritize -> prioritise	1
prioritized -> prioritised	1
organize -> organise	1
rumor -> rumour	1
organized -> organised	1
capitalize -> capitalise	1
centered -> centred	1
organizes -> organises	1
emphasize -> emphasise	1
split name repaired (Schaefer)	1
humor -> humour	1
traveling -> travelling	1
defense -> defence	1
maximize -> maximise	1
Organizations -> Organisations	1
apologize -> apologise	1
digitized -> digitised	1
monetize -> monetise	1
analyzed -> analysed	1
narrator opener re-attributed to Max, words left as transcribed	1
theater -> theatre	1
specialized -> specialised	1
split name repaired (Wadsworth)	1
aluminum -> aluminium	1
specializes -> specialises	1
skeptic -> sceptic	1
customize -> customise	1
personalization -> personalisation	1
optimized -> optimised	1
visualize -> visualise	1
spoken domain -> humanspark.ai	0

What was deliberately left alone

• "Program". A computer program keeps that spelling in UK English; only a broadcast or a scheme is a "programme", and only a reader can tell which was meant.
• "License" and "practice". Both have a noun/verb spelling split in UK English that a rule cannot decide.
• Capitalised words mid-sentence that would otherwise have been respelled: 4 of them, most often part of a name. Left as spoken.
• Retired show and brand names. Spoken by the host and guests. A transcript that removes what was said is not a transcript.
• "Co-pilot". Sometimes Microsoft's product, sometimes the metaphor. A rule would corrupt the honest ones.

The synthetic segments

The opening voiceover on the archive episodes and the closing call to action are read by Max, the show's AI narrator, from a script written in advance. Transcribing them means guessing at words already on file, and the guesses were wrong in places - one opener came back as "an AI that boned him up" where the script says "phoned him up". Those passages carry the authored text instead, and are attributed to Max rather than to the guest: the diariser groups Max with the non-host speaker, which would otherwise put a real person's name on words they never said.

Alastair McDermottA lot of people are suffering from AI fatigue, but there are a lot of us, business leaders, thought leaders, embracing AI and using it in a day-to-day way.

In today's episodes, I'm gonna speak with a business and thought leader who is heavily using AI, Stark, Jonathan, you're very welcome to the show. Hey, thanks for having me.

Well, Jonathan, I know you, quite a long time, I've been following you, I've been subscribed to your email probably for, I don't know, nearly 10 years now, I guess. It's getting close.

I have, I have worked with you as a business coach and as a podcast coach as well, actually helping me with the Recogn ize Authority podcast.

And I knew that you are somebody who I'd love to get your opinion on AI. So let me ask you this first.

If somebody was, if a friend of yours was to describe your thought leadership credentials, how would they describe them?

How would they, how would they make the case for saying that you are a thought leader? How would they make the case?

Jonathan StarkProbably the email list. That would probably be the thing been writing daily since 2016. I think it was June.

And yeah, every day I put out a new email where I explore mostly topics around pricing for experts, some kind of independent professionals like freelancers or consultants,

software developers, designers, copywriters.

And yeah, I just try and I just try and write every, I do write every day and I try and come up with insights that will help people understand how to make more money in less time without

hiring people. I love it.

Alastair McDermottHow would you think about thought leadership?

Jonathan StarkI mean, if you, I think of it like the term, like literally, so you're leading people.

So it's a type of leadership where you're going off on a mission into perhaps into the unknown, probably into the unknown. And perhaps people will follow you.

I don't, I don't, I don't see it as a, I don't see authority in the, in the, in like the term of like, you will respect my authority and,

you know, like a, like a, you know, policeman or something. It's, it's more like someone who you can trust to have a good answer or a good take on a thing that you care about.

And I just see it as like leading people's thoughts, you know, like I have these thoughts I'm going to explore, I'm going to go down the rabbit hole on these things.

And, you know, Hey, everybody come on with me if you want to, I see it like that. Yeah.

Alastair McDermottThe way you say it, like, it sounds very practical because I know a lot of people kind of maybe have mental hangups about the term.

It sounds a bit, I don't even know how to describe it, but it sounds a bit, you know, like a $10 word or something, you know,

Jonathan StarkIt's like calling yourself cool. It's like, Hey everybody, did I mention I'm cool? Yeah. You know, it's, it's, you can't, I don't really call myself that.

I think, I think I do the things that would make people call me that, but I don't tell people I'm an expert. I don't tell people I'm a thought leader. I don't tell people I'm an authority.

If they say that back to me, that's up to them, but it's not something I, it's not a goal.

Alastair McDermottYeah. Yeah. And I think, and that's one of the reasons why I asked that first question, like how would a friend describe you in that way?

Because I think that it's, I think that it's something that, you know, you can't really self-anoint, as an authority, as a thought leader. Yeah.

Jonathan StarkIf you turn around and no one's following your thoughts, I guess you're not a thought leader.

Alastair McDermottWell, that's true. so let's talk about AI. And, like I have, I started this podcast recently, this will be episode four of the AI power thought leader.

And the reason why I started this was because I had some conversations with folks who said, you know, if you're a thought leader or if you're an expert,

you shouldn't be using AI to help to assist with your writing and with your thinking and things like that. And I think that that's crazy personally.

so I'm like, I want to find out, like, I think there's a line here somewhere and I'm just, I'm trying to explore that.

So that's why I'm really interested in having you on, cause I know you're a huge fan of using AI.

So can you, maybe tell me a little bit about how you think about that conflict and, and how you're, you're approaching the problem.

Jonathan StarkYeah. So I agree with you that, I mean, crazy is strong, but it's like, do you use a word processor? Do you use a computer?

I mean, it's just another tool and it's not having thoughts, you know, you still have to, you still have to get in there.

And, you know, it can, it's like, I described this to one of my brothers recently. I was like, he was like, what? Cause he hasn't even used ChatGPT. Right. And he's like, what is it?

I'm like, you know, like your phone autocorrect stuff. He's like, yeah. And I'm like, it does that, but it'll like, it'll suggest like an entire article.

He's like, I want an article about this. And it suggests the whole thing. My experience has been that with rare exceptions, it never puts out something that I would publish.

You know, I treat it more like an editor or a writing, not a partner's the wrong word, but like sounding board.

It's a place, it's someone, it's someone air quotes that I can kind of, it's almost like sparring, like explore an idea.

It's, it's the same thing I would have done before where I Google something and research it. It's just faster and fewer ads, like things flying in my face.

So I think of it as like an assistant more than, you know, like if you're, if you're a thought leader or whatever, if you're an author or a writer, it's okay to have an assistant, right?

It's okay to have researchers, right? Like it's, to me, I see it like that. It's not coming up with original ideas.

Alastair McDermottSo the issue I think is that there are ways to use it for brainstorming and ideation in a way that it will help you to come up with ideas.

Or at least it'll help you to create like almost exhaustive lists when you're trying to think of, you know, what are the different formats that I could use for something?

And it can, it can kind of shortcut the thinking or brainstorming process, I think.

So, I mean, that's, that's a way that I use it personally is to come up with lists of, you know, if I do this, how else could I do that?

And what I find is that it will come up with stuff that I never would have thought about. And so that's the way I would use it for brainstorming.

Now, I have had, you know, several conversations with folks in LinkedIn comments in different places.

And one person said, I will happily use it for spelling and grammar, but I would never use it for brainstorming or ideation because that is, you know, a purely human thing.

And you can't use it for that. And now I know, I know I use the word crazy, but like, I do think it's crazy. I think it's like a, an accountant refusing to use Excel, you know? Right. Yeah.

Jonathan StarkThat's just, I flatly disagree with that. That person's comment. Yeah. It's not.

Alastair McDermottSo I'm like, where is this line, you know, where is the line where we say, you know, if we're, if we're calling like, okay.

If other people are calling us a thought leader or considering us to be a thought leader, and we use, you know, ChatGPT or Claude or one of these tools to come up with ideas or to sanity

check our ideas for, let's say a thought leadership, white paper, that is clearly in the, in the thought leader realm, you know, is that okay?

Jonathan StarkSo the quote that keeps coming to my mind is like from the music space.

So like Eddie Van Halen was like a famous, amazing groundbreaking guitarist and very well known for being a fast player.

He played very fast and he was being interviewed about, you know, like what's good music.

And, you know, cause a lot of guitar players, especially at the time would sort of do this thing that they call the guitar Olympics and they want to do the fastest shredder.

And he was kind of the king of that and they were asking him about it and he's, and his quote was like, look, if it sounds good, it is good. It doesn't have to be hard.

It doesn't have to be, it doesn't have to be anything. As long as it sounds good, that's good.

So if the ideas are good and they're novel and they're opening people's eyes, I kind of don't care who did it or got credit for it.

So if some back and forth with an AI or a partner or a group of people or a Google search or your researcher or from some movie you watched,

whatever gives you the idea, wherever it comes from, makes you think of the insight.

You're the one that detects the insight and is like, whoa, this is like ChatGPT can't tell you what's amazing.

It'll tell you, it might give you new ways to see things or reorganize things or different words to use or whatever,

at least the way I've used it so far, but only I am the one when I'm doing it is going to have the detector.

That's like, wow, that is, that turned on a light bulb for me and I'm going to share it. I don't, I don't care who gets credit for it. So the idea is either good or it's not.

So if you're using ChatGPT or something to generate a garbage white paper, well, it's still a garbage white paper, or if it's something that's been,

everyone knows, or it's just put in a new way, you're just like putting a fresh coat of paint on some old idea.

Well, it's just not going to resonate, but if it does resonate and people are blown away, you've led them into a place that they've never been before.

Like who cares if you used a headlamp or pickaxe or whatever tools you used, I really, I just don't care. I don't think about that whatsoever.

It's like, I want, I'll use any tool that'll get me down that cave faster. So, you know, is there a, I don't think about the ethics of it. And maybe I should, maybe I should, I don't know.

I don't care. It's like, it doesn't, it's not relevant to me. I'm not hurting. I don't believe that I'm hurting anyone.

Alastair McDermottYeah. I guess like one of the things that you mentioned there was, you know, giving credit or claiming credit. And I think that maybe that's one of the issues here.

And so I think, and I think it's quite nuanced as well, because like, I know that long before computers ever existed, we had ghostwriters and we've had ghostwriters.

We've had editors, we've had developmental editors, you know, helping authors to completely restructure books to make them much better to read.

And, you know, many of these people are credited and many of them are not credited in books.

And, you know, we have many sports people who supposedly write a column for the newspaper every weekend. But actually it's a journalist entity, you know, interviewing them and taking that.

So I think that the whole thing about the credit is nuanced and not as, not as straightforward.

The other issue is, and this is something that my friend Al istair McBride was talking to me about last week, was that Amazon have now put in place a limit where authors can only upload

three books per day.

Jonathan StarkThere's a, in fact, I was going to talk about this on this, on this podcast, there's a checkbox that's, that says something to the effect of,

you know, AI was used or not used in the creation of this book. It's like, well, what do you mean? Yeah. Right. Like I use Grammarly on everything. That's definitely AI.

Alastair McDermottWell, you know, it's, it's funny because you mentioned that to me before the start of the show.

And I know like I use Grammarly or an equivalent, I can't remember what it's called, but, and it runs my browser and has done forever. I didn't even think about that.

It didn't even occur to me. And the tool that I use actually does offer, if you go into its website interface, it does offer you an option to completely rewrite a sentence to restructure it.

Yeah. But I didn't think of that as AI because it's just, well, it was around since before the kind of the ChatGPT AI hype. And so, you know, it's just a software tool.

Jonathan StarkSo I just don't see how, I mean, the checkbox pretty soon is going to say, did you use a computer to write this book? It's like, it's going to be in everything.

And maybe, you know, I don't want to be defeatist about it and be like, well, this is all moot because it's going to be in everything.

But it is going to be in everything unless there's some like government regulation or something, which I don't, I don't see that coming. So it's like, you know, the issue there though, is if

Alastair McDermottWe say, yes, we did use AI, what is, what implicates, what implications does that have?

If you're publishing, you know, a really high quality book up onto Amazon where you happen to use Grammarly, and I say, well, yes, I did use AI.

And does that get you lumped in then with people who are generating, you know, 30 books a day using the free version of ChatGPT and uploading this total rubbish.

You know, does that mean that you kind of end up in this pit of doom where your book never sees the light of day?

you know, and so there's, there are issues because like, we know that the fact that Amazon brought in that limit means people were absolutely abusing this.

So, so I think it goes back then, you know, do we, it comes back to the, to the flip side of the credit thing of giving credit and claiming credit.

Do we admit to using these tools and if we do admit, then what happens, you know?

Jonathan StarkYeah, it doesn't say what the implication of checking that box is. And it's like, it feels like if it were going to be more, honest even now, which have to be the early days, even now,

if we're going to be more honest, it would have a little bit more nuanced version of the checkbox.

Like, what did you use it for, you know, if you used, so like I've used ChatGPT to like read my, like, here's, here's a PDF of a book.

I wrote, give me a first draft of the marketing copy for the back cover. And it does a great job of that. Did the, did I use AI to write the book? No.

Did I use it to summarise the book and come up with marketing copy for the back? Well, yeah, but am I going to check that box?

Alastair McDermottNo, that's, that's the issue.

And by the way, it's, it's absolutely super at doing those kinds of, cause those, that type of blurb for covers and things, that is where I think it really comes into its own.

Cause that's, I think that's difficult for us to write as humans, whereas for an AI to write it's very easy. but I think it's like that the, where we give it a lot of input.

And I think this is for me where the line is when I feel like when I've given the AI a lot of input and I tend to use the dictation feature quite a lot.

when I give it a lot of input, I feel like, yes, I have written that output because I've given it all of the ideas that it's using.

It's just reformatting, tidying up the spelling and grammar and, you know, maybe, putting some sentences together to string the thoughts, but, and connect the thoughts together.

you know, if I, if I ask it to, you know, write show notes for a podcast episode, I'm giving it the transcript.

And so I feel, well, yeah, I can claim that I wrote that with AI assistance, but it was me that wrote that. And I think this is where my friend, Joe Casabona, he had a great example.

He said, he said, great analogy. He said, using ChatGPT to write a book is saying, I used a car to run a marathon, which is, which is very funny.

Well, I think that it really depends on what level of input you actually have, because if you give a single line prompt. Yeah.

Then I think that's very different than if you like, for example, I've got a custom GPT where I've uploaded the text of five of my books and I've uploaded every blog post I've ever written.

And so when I asked that custom GPT to write something, I feel like I've created a lot of the input that it's using.

Jonathan StarkSo a lot of things going on here. So I think, I think I see where the, where people are getting tweaked because the car analogy is great because what you're doing,

what it's doing using a tool that's that powerful, regardless of how much you prompt it, you know, regardless of how much is based on your own ideas.

The, it's redefining the notion of what an author is because for forever, like an author was someone who wrote a book. And so it's like redefining writing. It's redefining what an author is.

It's, it's, it's, it's tweaking these norms that we're used to.

And if people are uploading, you know, 50 books a day to Amazon, then that threatens the notion of what an author is, what a book is, what you can expect from a book.

But if I come back to the A.D. Van Halen quote, like if someone uploads 50 crap books to Amazon every day, they're not going to get any reviews. They're not going to get any stars.

No one's going to read them. Like they're not good. If they are good and they get 3,000 five-star reviews, then who cares who wrote it? It's, I just don't care.

You know, it's the, but I do see what the marathon example is a good one.

Cause it's like, well, that's breaking the, that's like the whole thing of saying I ran a marathon is like, you're, you're, it's a status thing.

It's, it's saying a lot about the person and that is not true. It's implied. Okay, here it is. It's implying a bunch of stuff that isn't true.

So I suppose it's fair to think, okay, if I say I'm an author, my name is on the cover of this book. I'm saying that I wrote it or I'm implying a bunch of things.

Like I was up late, you know, in the middle of the night typing and you know, I've struggled through this whole marathon of writing a book. And that is how I describe writing a book.

And if you actually didn't do that and it was actually easy, then yeah, it just redefines the narrative or the mystique, I think of being an author.

But at the end of the day, I don't care though. It's like, is the book helpful or is the book not helpful?

And if the book is helpful, you know, and you did it very easily and made a billion dollars from it.

Well, that's kind of my whole mission, which is like create more value with less work so that you can capture enough, you know, enough that you can support your lifestyle goals.

So I'm all for finding easier ways to create value, like ways to create value that are less work.

And if AI is one of the ways I have yet to see a deal breaker problem that would make me be like, oh, wow, this is actually really dangerous or bad.

Alastair McDermottYeah. There's, there's, what I find interesting about this is there's so much nuance in all of this and the lines are really blurred.

You know, and that's not even getting into all of the ethics and, you know, privacy and all of these kinds of issues as well. Sure.

and the fact that these things have been trained on people's intellectual property without permission in many cases.

so the idea of having these books, like we know that there are a lot of people out there who talk about, you know, a book is the best business card you can ever have.

And, you know, that writing a book is a kind of a route to credibility and authority. And so now I can imagine that people like that.

A lot of people are going to see that with less credibility and let, you know, a lot more skepticism in the future.

Like I can see that the entire concept of being an author and I hadn't thought about this until he said it. but the entire concept of being an author may mean something different.

like I know that, you know, like if we set our minds to it, either of us could probably produce a book in a day that would look, it would look plausible, you know?

and it may be like, it may be like a B minus or a B plus, it might actually not be, not be that bad. So, but then the other question is, does it matter who wrote this?

Well, it may get to the stage. Well, if it doesn't matter who wrote it, I might as well just go to ChatGPT and not bother to work with you, Mr. Expert in your field, you know? Right.

Jonathan StarkI mean, yes, it could redefine what a book is. It could make books irrelevant. I don't know.

But right now, but I, you know, it's like if something gets, if books get replaced by something better than books were replaced by something better.

So like, I don't have a nostalgia about books. I mean, I was, I should, I mean, my stepfather ran a used bookstore.

I worked there for years when my mom would take us to the beach when we were little kids. I'd bring, like, I wouldn't go in the water.

I would stay under the trees and like read Star Trek paper backs. Like I'm totally a book person. I've written five books, like the old way.

So I should be like all about books and defending books. And I just have no, I'm just not precious about books. Like if there's some better way to help people improve, then so be it.

I just, you know, but it could, it could happen that, you know, Amazon gets flooded with 10 X more books.

So never, I just, but I feel like the, I feel like they're, if they're bad, they won't gain traction. And if they're good, they will.

So like atomic habits, I don't care if James Clear wrote that. It's a good book that first Hormozy book, it's a good book. I don't care.

He may be, and that was AI era that could have been generated by some large language model. It's a good book. So I just don't, if the book is good, it's good. If the book is bad, it's bad.

If somebody, you know, has their name on the cover, it's just not going to spread. I don't see, I don't see a bad book spreading. I mean, I've written good books that didn't spread.

Alastair McDermottThat's the other issue then, because there are plenty of authors who have written great books that haven't got many reviews.

And I think, I think it would be fair to say that all of the books that are 99% of the books that have got many good reviews and many sales are very good.

But there are also many that didn't, you know, didn't go viral enough or didn't have an author who knew how to play the game properly that didn't make it.

And so their books are sitting side by side with all these, this AI generated content. And they're saying, you know, this isn't fair.

Jonathan StarkGet in line. I mean, progress. I don't know. I just don't have sympathy for the argument. It's like, stay ahead. Like, do the stuff that computers can't do yet.

Like, keep climbing up Maslow's hierarchy of needs. Just like, if you're like, well, I'm going to be an author and I'm going to succeed being an author.

That's a very self-centred, you know, it's like, I want to spend my days engaging in these kinds of activities. And the world owes me a salary for doing it. Yeah.

It's like, I don't see it like that. It's like, you go out and create value and you will be compensated for it.

But if you don't, you don't automatically, you're not entitled to a living engaging in your hobby. Anybody that's tried to make a living off their hobby knows this.

So I'm kind of, you know, I, you know, it feels mean to say it, but it's like, look, create value.

If the packaging of your expertise, if you're, I mean, if people who are aspiring or are considered thought leaders are listening to this and books are no longer a viable way to,

project or demonstrate or spread your authority, then do something else.

You know, it's like, do the book and like, wow, that only took me a day and it, and maybe it'll do okay, but maybe a podcast, maybe a podcast or a video is the new way. I don't know.

I don't know what the, or speaking back in in-person actual speaking until we have Androids, like that's not gonna, that'll be the new scarcity, I suppose.

Alastair McDermottYeah. This is, continuation of what I was talking to my friend, Al McBride about, and we were talking about, you know, like what is hard to replicate.

And so for right now, it would be difficult for AI to generate this conversation between the two of us. Right.

for example, you just took a drink and you took pause, you moved all of that is difficult for AI to replicate right now and fake, and it would look uncanny would look unreal.

but I think, you know, it would be, it would be dangerous to think that I can't do that in three or four or five years time. I assume we'll be able to. Yeah.

And then, so I think, I think if you, if you consider, okay, well, let's keep climbing the ladder.

I think that, so for right now, us being on video like this is very difficult for it to replicate right now. And I think that the live in-person event is going to be very difficult.

We're not going to have humanoid, like, you know, commander data, type robots. We're not going to have that, for a long time.

So I think that live in-person events may in fact become, you know, that may become the new, the new kind of pinnacle of thought leadership because,

you know, that's cause that's the way it goes.

Jonathan StarkLet's, that has been a tried and true way. I mean, speaking at conferences is like, you know, author, speaker, consultant, that was the, that was in the, in the 2010s.

Well, I guess 2005 through 2015, that was my jam. I wrote books and flew around and spoke at conferences.

And I think the conference one is probably going to probably, I mean, if I had to bet which one was going to produce authority,

like benefits for the next 10 years, I bet on speaking live speaking events and not writing books.

Alastair McDermottSo if you were talking to a coaching client about this now, who was in the position where they wanted to build authority and build thought leadership,

kind of profile, where do you think you'd be advising them?

Jonathan StarkI still have people that are putting out, I'm in the middle of a book launch with one of my students right now.

And, but it's a good book, you know, like in this particular case, AI was not involved other than perhaps the back cover.

But, but again, it doesn't, it's, but an editor was involved, you know, and in the, and had a big effect on the quality of the output. And, you know, I, and it's, it's a really good book.

It's a really good book. So it could have a huge effect. So I think this thing that we're worried about, it hasn't happened yet. I mean, books aren't dead.

That's, I mean, maybe if I had to bet, maybe I'd be like, or I might short books in the longterm, but you know, not now it's still,

I think it's still a very valid way to, if you write a very good book.

Alastair McDermottThe other thing that might happen is there might be a swing back towards, commercial publishing away from self-pub lishing.

if, you know, because what the commercial publishers were bringing to their table was their stamp of approval and almost nothing else. whereas now that may become important again.

Jonathan StarkIt's possible. Yeah. So third party, some sort of social credibility, yeah. Or third party, you know, validation that, I mean, that's, that's true.

Now, I mean, it, but to your point, it could become more important, but I would almost traditional publishing is so messed up at the moment that you could be right.

That could be one of the things I think one of the other things that has the same effect, but doesn't mean you have to go through a publisher is to have someone famous blurb the book.

So like right on the cover, it says, you know, Barack Obama says, this is the best productivity book in, on planet earth. And then it doesn't matter what the author's name is. Yeah.

Alastair McDermottYeah, absolutely. okay. So let's get away. Well, have we wrapped it up on books or is there anything else we need to talk about about that?

Jonathan StarkMy experience is maybe my standards are high, but it is just not been my experience that you can directly take any kind of output without a lot of correcting for voice.

Like it can't write like me. And I've trained, I've done the same thing with you. I've trained it.

I have someone who put together a, you know, a J stark bot and you can ask you questions and like 80% of the time it says the right answer, but it doesn't sound like me saying the answer.

And then like maybe 20% of the time it gives an answer that's like, that's not wrong, but that's not what I would have said.

And then 5% of the time it says something that I absolutely oppose. And I, you know, so it's, it's not ready for that stuff like that is not ready for prime time.

My experiences has been the same with books where, you know, I'll write a book about productivity that has these three things and ingest the slides and transcript from this hour or two hour

long webinar that I did. And like turn that into a book, it's, it's not good. It's okay. It's an okay first draft, but you still need to go through that thing.

Like word by word and be like, oh, no. It can. Yeah. So it probably is good at writing garbage books right now, but yeah, we can move off of books.

Alastair McDermottYeah. The only thing I'll say to all of that is it's improving exponentially and what it can't do now. It may be able to do in a year's time.

Jonathan StarkI hope so. I wish there were more good books.

And by good, I mean, you know, we're talking nonfiction, but we could talk fiction too, but in terms of nonfiction, I would love it if there were more good books.

I would love it if there were books that I could point students to that were, that were, that were useful by good.

I mean, useful in terms of producing insights that turn into improvements in their lives. Again, I don't care who gets credit for it. Like just their lives are better.

Alastair McDermottGood. Yeah. I mean, the way I think about, particularly in the case of fiction, and I'm thinking it's probably going to happen with movies as well,

where, you know, in 10, 20, 30 years time, we'll be able to say, Hey, Netflix, generate a movie for me, which is a murder mystery set on Mars, which is a crossover between Isaac Asimov.

And, and, and Sherlock Holmes.

Jonathan StarkIf Netflix isn't working on exactly that, I don't know what they're doing.

I mean, that has to be what they're doing because it's, they do that now where they take all of your, you know, in order to recommend movies to you.

And I've, I've heard, I guess, production studios or directors or whoever pitches movies to Netflix say like they'll greenlight something right away.

If it ticks these three boxes, because they know that some percentage of people will watch this.

So right now they've got a big bottleneck, which is that someone has to make the movie, but they can pretty much tell production companies, here's the movie or TV show that we want made.

We know it will be a decent hit. People will watch it.

And then they, so they're already telling these, create the creative people, like, here's the, here are the components that people want to watch.

So like you take the production company out of that. I mean, you could see it making movie trailers.

Like you just said, I could imagine right now it is capable to say, make me a movie trailer.

That's like you said, you know, a crazy mashup of like a buddy cop, sci-fi procedural, and just the trailer. And it could probably make a 30 second trailer that you'd want to watch.

Alastair McDermottAnd then you could test that trailer. The trailer does well in testing. Then you go get your director and say, Hey, here's the trailer. Go make the movie. Yeah. And we can do that now. Yeah.

Yeah. Yeah. okay. Let's talk about how you actually use AI on a day-to-day basis. Cause I'm interested in knowing more about what you're doing.

Jonathan StarkYeah. So Grammarly is probably the biggest one. And I really had to sit down and think about it when I was getting ready for this, because I feel like I use it a lot.

And then I'm like, do I use it a lot? And in some sense I do.

I use it in little ways a lot with like dictation and, the, and Grammarly correcting all of my, you know, I write every single day and I don't have an editor.

There's no process for review and Grammarly like upped my game massively when I started using that, you know, no more stupid, you know, there typos or commas in the wrong place.

None of that stuff. It's very, it's a huge improvement. I'm, I do two podcasts a week, sometimes three, and I always have it.

In fact, it automatically transcribes the thing and, create, and if you click one button, it'll create show notes, suggest a title. And if you want it to make shorts, it'll do that.

It's not that great at that, but, the transcripts and the summaries are very good, much better than I would have. I mean, it's so hard to do that.

I wouldn't, I would, no one would manually make transcripts. That'd be bonkers. It'd be, to be so expensive. So, you know, I didn't use to have transcripts.

I didn't have, used to have show notes.

I never did show notes, with business of authority, Ro chelle used to do show notes, but I was like, I'm not doing show notes, but now they're automatic. So cool. Captions are automatic.

That's cool too. So yeah. So that is a common use case.

other ones that I don't know if you mentioned this before the show or after the show, but, or during the show, but summarising long things, it's fantastic. It's summarising one things.

And if, and if you were there, like where we're here, we're in this conversation, if we, if we have a summary of it after you can be like, oh yes, this is correct.

I get a little bit more nervous if it's summarising something else. Cause I'm like, well, I don't know if it's correct.

Like maybe it's wrong, but you know, so I use it for mostly for stuff that I wrote or I was involved with the creation of, and then I can be like, oh yeah, this is right.

This is really good. This is like, this is what I remember happening here.

Alastair McDermottThe timestamps, asking for time.

Timestamps and direct quotes along with the summary can be really useful for fact checking when you want to bypass the hallucinations and be really useful to say,

Hey, give me the quote with the timestamp for that. Yep.

Jonathan StarkI use it. I have a bi-weekly of a coaching community where every other week we do a Q and a session, like a live Q and a session like this.

And after that's over, I have a checklist of things that I go through that is, you know, it takes like an hour to go through all of these things.

There's uploading and downloading and all sorts of stuff. And I have, I do, I use ChatGPT to transform things.

So like some of the apps that I use will produce, like HTML that I need in Markdown format.

So I use it to switch stuff from like HTML to Markdown or the other way, or like convert this to Jason, you know, it's, it's really good at that. And you can tell it's right.

You know, it's not one of those things where you're like, I'm not sure if it's right. It's like, that's definitely right. I can see that it is.

So that the idea of doing data transformations is really helpful. It's so tedious doing that stuff.

and in fact, now that we're talking about it, there are other big chunks of this checklist.

It's probably like 150 things that I have to go through to make sure that, you know, all of the, all of the production post-production things are done. There's probably big chunks of it.

I should probably just take the checklist and put it in ChatGPT and say, which one of these things can you handle? And it would be like, oh, I can do this, and this like, great, do that.

So it's pretty cool. I should try that.

And another thing is I'll just like, I've been really, so a friend of mine, Kelly Shaver, and I doing like a writing short story writing challenge this month.

And I've been using it to sort of research physics of what would be potentially happening on Mars in the future if there was a base. Right. So like what would happen?

So I've been, I've been like, I'll spend two hours in ChatGPT being like, is the speed of light really constant? Or is that just a convenience of the mathematics?

And, you know, and just like, well, what would happen if it wasn't, if it turned out it wasn't constant? What implications would that have?

And the stuff it's giving me back, I don't know if I trust it, but it sounds, you know, it's good at sounding trustworthy.

So you get these things back and I'm like, huh, well, I don't know if this is true, but I do feel smarter about relativity, for example.

And I could go look it up to double check, but it's good enough for fiction.

You know, if it's plausible enough to convince me that this could be true or what time dilation, time dilation works, then it's good enough for fiction.

So I'll just use this in the story, but I feel like I'm getting smarter about physics.

Alastair McDermottYou should check out Werner Wings, A Fire Upon the Deep, if you haven't already. Not, have not heard. Yeah, you, well, it speaks almost directly to what you're talking about.

So, but I won't get too sci-fi nerdy on the, on the audience. Okay. So there's some interesting things you mentioned there.

And so you talk about tedious data transformation and I'm using it for tedious.

Things I think is like the best thing about AI is we can get to do all the grunt work that was just so time consuming.

But I think the interesting thing that you said is you can tell that it's right. And the fact that that was noteworthy, I think is something to just dig into a little bit.

Why is it noteworthy that you can tell if the output is right or not?

Jonathan StarkWell, I'm sort of speaking to the hallucination scare headlines, you know, it's like, but it hallucinates and it's like, well, how would you have researched this before ChatGPT?

You would have Googled and then you would have went to a bunch of websites that you don't know the source of, which is what ChatGPT was trained on.

So of course it's going to be wrong sometimes. And, you know, if you're a lawyer presenting a brief to a judge, don't do that. That's insane. Yeah. You know, referencing an actual story.

But if you're just writing a short story that you're going to submit to some, to Asimov magazine or whatever, big deal.

You know, it's, it's just, there's certain things you can use it for where you just know it's right. It's like, you ask it to write some code and the code runs. It's correct.

So for, you know, for people who sort of dismiss the whole thing by like, sometimes it, by saying sometimes it's wrong is like,

well, just use it for stuff that you can tell if it's right or wrong. And then if it's wrong, just be like, ah, the code, you know, there's a bug in the code. Here's the error message.

Rewrite it so that this error doesn't happen. And then just runs and compiles and it works, or it transforms the data in a way that you're like, yeah, that's correct.

I just didn't have to type all the semicolons. So, yeah.

So for things where you're pretty much everything I use it for, except for the special relativity stuff, I, pretty much everything I know that it's, it did a good job.

I can see that it did a good job and the special relativity stuff. It's no professor would have the patience for the kind of interrogation I give ChatGPT. No one would have patience.

Like it's dumb questions. It's huge questions. It's, you know, how do we know it just, it's just, I will literally badger it for hours and try and catch it in inconsistencies.

And it'll say like, okay, yes, that's true that I said this way back when, and now I'm saying this, but there's a nuance and it explains it. And I'm like, oh, that's a good explanation.

Like, and we're so far down the rabbit hole. It just really feels correct, but no teacher, no college professor would ever put up with me texting them for two hours on a Friday night.

Cause that's the kind of life I have. I'm chatting with, I'm talking to ChatGPT about special relativity on a Friday for two hours. But that's actually true. That is how I spend my Fridays.

Alastair McDermottSo, okay, let's wrap this up.

I think that one of the most interesting things that, so there's a few highlights from what you said, chat, ChatGPT can't tell you what's amazing that it doesn't have that ability yet.

I mean, maybe we can train it, trying to figure, to learn, to identify what's really great. How would you do that? Yeah.

Jonathan StarkYou got my gears turning. I think there's probably a way to do that, but it would involve human input. Yeah. I don't know. I feel like that, I feel like that's a moving target.

I think that is always a moving target. I could be wrong, but.

Alastair McDermottI think the other thing that you mentioned about AI redef ined the notion of what an author is or what that means and what the implication is to say that you've written a book,

because like, I think that that requires some thought. and then, you know, the idea that you can, you can tell, you can tell that it's right.

The output, one of the things that I've, I've been outs ourcing to subcontractors for years since I started my business in 2007.

One of the things that I always did was I almost never outs ource something that I didn't know how to do myself.

And I think that for me, it's a little bit like this with the, with the AI tools is I'm getting it to do things that I can do already. So I can sanity check it.

I think it's when you push past the envelope of what you can sanity check yourself personally, that things become a bit more fraught with potential for error.

Jonathan StarkYeah, I get that. Like if I was asking it to write a book about something I don't know anything about, then that's, you know, that's putting garbage out.

My assumption is that that book would get zero traction, but it could be wrong.

Alastair McDermottYeah. And then when it's doing those transformative, tasks, like when you're inputting things like transcripts and, getting output as summaries and show notes things,

that's where it really shines for us right now, at least. Oh yeah.

Jonathan StarkYeah. I mean, for me, that's not putting anybody out of work either. Cause I just didn't do it before and now I can do it because the cost is basically free. So, so yeah.

Alastair McDermottYeah. I probably wouldn't do this show because it would be another podcast for me to produce and create and do all of that. If I didn't have AI tools, I just wouldn't bother. Right.

that's a bit meta because it wouldn't exist, but, you know, you know, you know what I mean? I just want to talk about one last thing.

I know there's a lot of people saying I'm getting sick of the tire of this AI stuff in my feed. And, it was, it was one of the reasons why I almost didn't create this show actually. Yeah.

Well, I'm kind of concerned by the idea that people could dismiss AI or kind of, you know, get this fatigue where I think that it's going to be potentially either,

you know, it's, it's potentially humanity changing, either positive or negative or both.

like, what do you think about the concept that people have are feeling AI fatigue and just don't want to hear about it anymore?

Jonathan StarkI mean that it's, it is hype cycle and a lot of people are just riding the hype cycle like they always do with hype cycles,

just adding no value, all noise, especially using AI to create more noise. There's definitely gonna be a lot more noise. So, okay. More noise. Your filters will just have to get better.

I mean, we've already gone through that, but you know, if you look at the, if you look at the most recent previous hype cycle, it was crypto or blockchain and all of that stuff.

And that never, like, I found that interesting from a technology standpoint and NFTs and my sort of music background. And I was like, okay, there's, this is kind of interesting.

I want to understand how this works, but that burned out very quickly. Like I got to the, I got to the edges of that area very quickly.

And I was like, eh, this is, this is either nothing or it's not ready for prime time or Bitcoin is the only use case.

It just, you know, people were always struggling to come up with use cases beyond Bitcoin. Like Ethereum's reason for being is essentially like to create use cases for blockchain.

And I was like, eh, and that was huge hype cycle. Tons of people constantly talking about every headline, even mainstream news, like crypto.

And then chat TPT came out and it was like, my kids were using it before I even heard of it. And I'm like, oh my, there's use cases everywhere.

And then the adoption curve of the open AI specifically, but just more broadly Midjourney was another huge one early. And I was like, this feels like mobile.

Cause I was, I was like eyes wide open when mobile happened. And I was like, I'm jumping on this. And it ended up being the most widely distributed technology of all time of any kind.

There's more people have mobile phones and smartphones than can write like literacy.

It doesn't have the penetration that smartphones has toothbrushes, bicycles, nothing mobile by far is the most successful technology to date.

And it enabled all of those people to have access to AI and AI feels like, I mean, specifically chat TPT because of the interface, it feels like mobile did.

It feels like it's like crypto felt like it was like a, it was like a, it was like a, it's like a little wave, nothing really just like a, but this is like a tie.

And it just, the idea of ignoring it and just writing it off because you read some headline in the New Yorker about how it's unethical to,

or it's burning so much energy, whatever the thing, whatever the typical complaints are, it's like, yeah, it's a disruption and people aren't going to like it.

Some people are not going to like being disrupted, but ignoring it seems kind of like, okay, if that's how you feel. Yeah.

Alastair McDermottYeah. Yeah. It's that's a good point. It is a disruptor and, you know, the status quo has always had a disruption.

So yeah, that's, that's, that's the way that it's been throughout the evolution of humanity. So, Jonathan, if people want to find you, where can they go?

Jonathan StarkJust put my name into ChadGBT and, you can Google me, I'll come up, I'll be the first hit.

Alastair McDermottOkay. And Jonathan has an amazing daily email letter. I suggest that you, you, sign up for that as well.

he also has a podcast, with Rochelle Moulton called the business of authority and another podcast called ditching hourly. Do you have any other podcasts that I don't know about?

Jonathan StarkI have one with my brother, Matt called stark reading, where we talk about philosophy and politics and all the, all the stuff you're not supposed to talk about and mixed company.

Alastair McDermottI didn't know about that one. So that's kind of cool. Yeah.

Jonathan StarkI don't talk about it much. I'm sure I'm going to get cancelled.

Alastair McDermottI did ask, ChatGPT, who is Jonathan Stark. And, so there's an actor, a tennis player.

Jonathan StarkYeah. I've got, so they're, they're still, they still have me beat in ChatGPT apparently, but I've got them being Google.

Alastair McDermottAll right. Yeah. And you're a software consultant and a former software developer, according to ChatGPT. Okay. So there you go.

Jonathan, thank you so much for taking the time to come on and chat with me about

MaxThanks for listening to Something About AI. If you got something out of this one, share it with someone who would get something out of it too. That is genuinely how the show finds people.

There is more like this coming, so do subscribe wherever you are listening. And there is a pile of free resources over at humanspark.ai. Go and have a look.