AI across the publishing stages

WEBVTT
Kind: captions
Language: en
00:00:00.000 --> 00:00:02.389
I founded Page in Podium Press on the
00:00:02.399 --> 00:00:05.349
very human premise that our lived
00:00:05.359 --> 00:00:08.310
experience teaches us so much about the
00:00:08.320 --> 00:00:11.030
world on helping people to claim that
00:00:11.040 --> 00:00:13.990
expertise and step forward to both tell
00:00:14.000 --> 00:00:16.470
their stories and share what they have
00:00:16.480 --> 00:00:19.189
learned. And because I see this as such
00:00:19.199 --> 00:00:22.550
a deeply human proposition, y'all, it
00:00:22.560 --> 00:00:25.750
shook me how I have seen AI coming into
00:00:25.760 --> 00:00:28.550
author's projects. So today I want to
00:00:28.560 --> 00:00:31.349
continue our four-part series on AI by
00:00:31.359 --> 00:00:33.430
talking about the ways that I have seen
00:00:33.440 --> 00:00:37.510
AI creep into author's projects at every
00:00:37.520 --> 00:00:40.869
step from writing to publishing. I'm Dr.
00:00:40.879 --> 00:00:42.869
Amanda Nell Edgar. This is the page and
00:00:42.879 --> 00:00:45.430
podium podcast and today is our second
00:00:45.440 --> 00:00:48.069
episode in our four-part AI series
00:00:48.079 --> 00:00:51.510
discussing how AI can infiltrate every
00:00:51.520 --> 00:00:54.150
stage of the publishing process. Stay
00:00:54.160 --> 00:00:59.430
tuned.
00:00:59.440 --> 00:01:01.590
Most memoirs don't stall because of lack
00:01:01.600 --> 00:01:04.229
of time. Most memoirs stall because of
00:01:04.239 --> 00:01:06.950
lack of clarity. And that's why I put
00:01:06.960 --> 00:01:09.190
together the memoir method checklist.
00:01:09.200 --> 00:01:11.670
The idea here is to give you every
00:01:11.680 --> 00:01:13.990
single step in the process. Not because
00:01:14.000 --> 00:01:16.310
that makes it easy, of course not, but
00:01:16.320 --> 00:01:18.870
because no project is ever completed
00:01:18.880 --> 00:01:20.950
until you know what the steps in the
00:01:20.960 --> 00:01:23.510
process actually are. You can download
00:01:23.520 --> 00:01:31.190
it free at pagenpodium.com/checklist.
00:01:31.200 --> 00:01:34.149
As I teased in the intro to this video,
00:01:34.159 --> 00:01:38.149
I am seeing AI everywhere. Places that I
00:01:38.159 --> 00:01:40.630
never anticipated that I would see it.
00:01:40.640 --> 00:01:42.950
Particularly because I know that the
00:01:42.960 --> 00:01:44.710
folks that come to work with Paige and
00:01:44.720 --> 00:01:47.749
Podium share the same human values that
00:01:47.759 --> 00:01:50.469
I do. I know that when folks come to
00:01:50.479 --> 00:01:52.389
work with me and particularly some of
00:01:52.399 --> 00:01:54.950
these are folks that I have talked to, I
00:01:54.960 --> 00:01:57.590
know that we are on the same page in
00:01:57.600 --> 00:02:00.630
terms of our values. And what really got
00:02:00.640 --> 00:02:02.950
me thinking about AI in terms of
00:02:02.960 --> 00:02:04.709
writing, what really prompted this
00:02:04.719 --> 00:02:07.510
series actually was realizing that the
00:02:07.520 --> 00:02:09.350
reason that that's happening is not that
00:02:09.360 --> 00:02:11.830
anyone is trying to necessarily game or
00:02:11.840 --> 00:02:14.790
cheat the system, but it is the same
00:02:14.800 --> 00:02:16.790
thing that I have talked about week
00:02:16.800 --> 00:02:19.510
after week for now going on over three
00:02:19.520 --> 00:02:22.790
years on this podcast. The reason that
00:02:22.800 --> 00:02:25.430
folks are turning to LLMs in so many
00:02:25.440 --> 00:02:28.309
cases on these books they care so deeply
00:02:28.319 --> 00:02:31.350
about is the same old story. It is the
00:02:31.360 --> 00:02:34.869
sense of insecurity and fear that LLMs
00:02:34.879 --> 00:02:37.750
promise it will push us past. They
00:02:37.760 --> 00:02:40.150
promise us if we just use Claude to
00:02:40.160 --> 00:02:42.630
quote unquote clean up our writing, we
00:02:42.640 --> 00:02:45.030
won't have to face that difficulty, that
00:02:45.040 --> 00:02:47.430
struggle, that discomfort of not being
00:02:47.440 --> 00:02:51.030
as good at the process as we would like.
00:02:51.040 --> 00:02:54.790
Even though our idea is so rooted in
00:02:54.800 --> 00:02:58.710
humanity. Now this is a really human
00:02:58.720 --> 00:03:02.070
issue, right? We as humans really want
00:03:02.080 --> 00:03:04.229
security. We want predictability. We
00:03:04.239 --> 00:03:06.470
want safety. And one of the things that
00:03:06.480 --> 00:03:08.790
bookw writing is always going to do is
00:03:08.800 --> 00:03:12.309
push us out of those exact arenas. Bookw
00:03:12.319 --> 00:03:14.710
writing is one of the things that is the
00:03:14.720 --> 00:03:18.149
most vulnerable and particularly when
00:03:18.159 --> 00:03:20.070
we're talking about books that pull from
00:03:20.080 --> 00:03:22.869
your personal story. It feels like you
00:03:22.879 --> 00:03:25.990
are naked in Time Square. So I
00:03:26.000 --> 00:03:29.110
understand why AI is coming into these
00:03:29.120 --> 00:03:31.910
workflows. I get it. I am sympathetic in
00:03:31.920 --> 00:03:35.030
fact. But what unfortunately I'm seeing
00:03:35.040 --> 00:03:37.509
over and over is that even though I can
00:03:37.519 --> 00:03:39.990
understand why folks would pull in an
00:03:40.000 --> 00:03:42.789
LLM or other AI to finish or quote
00:03:42.799 --> 00:03:46.149
unquote improve their manuscript, it is
00:03:46.159 --> 00:03:49.750
never a good choice. So I wanted to talk
00:03:49.760 --> 00:03:51.509
about some examples that I have seen
00:03:51.519 --> 00:03:53.830
recently that I think can just help give
00:03:53.840 --> 00:03:55.589
you some kind of groundwork, some
00:03:55.599 --> 00:03:57.910
framework to think about how this might
00:03:57.920 --> 00:04:01.190
happen to you and if you want to choose
00:04:01.200 --> 00:04:04.390
to use AI, how you can at least make
00:04:04.400 --> 00:04:06.149
sure that you are making that choice
00:04:06.159 --> 00:04:09.750
from a grounded, measured, informed
00:04:09.760 --> 00:04:12.789
place rather than from the fear and
00:04:12.799 --> 00:04:15.509
insecurity and panic that I know come up
00:04:15.519 --> 00:04:18.390
for literally everyone at some point in
00:04:18.400 --> 00:04:20.310
this process, myself included, by the
00:04:20.320 --> 00:04:23.510
way. So, as we dive in, as I said, I
00:04:23.520 --> 00:04:25.990
want to move through each stage of the
00:04:26.000 --> 00:04:28.469
publishing process. And specifically, I
00:04:28.479 --> 00:04:30.070
want to let you know that I am, of
00:04:30.080 --> 00:04:31.909
course, going to be narrating the way
00:04:31.919 --> 00:04:33.830
that Page and Podium uses this
00:04:33.840 --> 00:04:35.830
publishing process. Now, if you're
00:04:35.840 --> 00:04:37.110
interested in learning a little more
00:04:37.120 --> 00:04:39.110
about what we do in terms of publishing
00:04:39.120 --> 00:04:41.030
at Pagen Podium, you can visit
00:04:41.040 --> 00:04:43.830
pageandpodium.com/publishing
00:04:43.840 --> 00:04:46.310
and it will talk us through our process.
00:04:46.320 --> 00:04:48.390
Now, one of the things that is really,
00:04:48.400 --> 00:04:50.550
really integral to how we do our work at
00:04:50.560 --> 00:04:53.270
Page and Podium is that every step of
00:04:53.280 --> 00:04:56.950
this process is collaborative. We really
00:04:56.960 --> 00:04:58.790
pride ourselves and in fact have built
00:04:58.800 --> 00:05:01.590
out our entire structural framework
00:05:01.600 --> 00:05:04.469
around the idea that firsttime authors
00:05:04.479 --> 00:05:07.909
not only need but deserve an industry
00:05:07.919 --> 00:05:10.550
professional to really not only walk
00:05:10.560 --> 00:05:12.710
them through the process but help them
00:05:12.720 --> 00:05:15.189
understand why different decisions are
00:05:15.199 --> 00:05:17.029
being made, why they need to be made
00:05:17.039 --> 00:05:19.270
within what parameters they need to be
00:05:19.280 --> 00:05:22.310
made so that the author never feels as
00:05:22.320 --> 00:05:24.390
though their work has been ripped away
00:05:24.400 --> 00:05:27.189
from them. We always want you to feel as
00:05:27.199 --> 00:05:29.430
though we are in your corner, holding
00:05:29.440 --> 00:05:31.430
your hand if you need it, but mostly
00:05:31.440 --> 00:05:33.590
cheering you on and making sure you have
00:05:33.600 --> 00:05:36.070
all of the information you need to make
00:05:36.080 --> 00:05:38.790
the decisions that you need to make in
00:05:38.800 --> 00:05:41.670
again this highly, highly vulnerable
00:05:41.680 --> 00:05:44.950
kind of scary process. So, as I am
00:05:44.960 --> 00:05:46.469
talking through these steps, I just want
00:05:46.479 --> 00:05:48.310
to be clear. This is the page and podium
00:05:48.320 --> 00:05:50.230
process. Different presses are going to
00:05:50.240 --> 00:05:51.749
vary in how they move through this
00:05:51.759 --> 00:05:54.310
process, but I really wanted to give you
00:05:54.320 --> 00:05:56.790
some examples that really are from our
00:05:56.800 --> 00:05:59.670
workflow. And again, from authors that I
00:05:59.680 --> 00:06:02.230
don't judge or necessarily blame for
00:06:02.240 --> 00:06:07.029
using LLMs or AI, I understand the heavy
00:06:07.039 --> 00:06:09.909
heavy impulse to use it, particularly
00:06:09.919 --> 00:06:12.309
when we're being told all the time that
00:06:12.319 --> 00:06:14.150
we're going to be quote unquote left
00:06:14.160 --> 00:06:17.189
behind if we don't get on board the AI
00:06:17.199 --> 00:06:19.590
bandwagon. Well, friends, I don't
00:06:19.600 --> 00:06:21.990
believe that's true. And in fact, the
00:06:22.000 --> 00:06:23.749
evidence, the examples I'm going to give
00:06:23.759 --> 00:06:25.830
you today, I think will make it pretty
00:06:25.840 --> 00:06:28.469
clear that LLMs are not going to make
00:06:28.479 --> 00:06:31.029
your book better. So, in fact, maybe we
00:06:31.039 --> 00:06:33.110
actually should just agree to be left
00:06:33.120 --> 00:06:34.950
behind by AI when it comes to writing
00:06:34.960 --> 00:06:37.749
our non-fiction books. I certainly am
00:06:37.759 --> 00:06:39.909
very much okay with being left behind
00:06:39.919 --> 00:06:41.909
when it comes to my own writing because
00:06:41.919 --> 00:06:44.790
I'm a better writer than Claude. So all
00:06:44.800 --> 00:06:47.510
right with that and speaking of writing
00:06:47.520 --> 00:06:51.029
we are starting at the writing point. So
00:06:51.039 --> 00:06:52.629
first thing I want to say if you have
00:06:52.639 --> 00:06:54.629
not watched the first episode in this
00:06:54.639 --> 00:06:56.950
series it just aired last week. Go back
00:06:56.960 --> 00:06:59.029
and listen to that first because I'm
00:06:59.039 --> 00:07:00.070
going to give a little bit of a
00:07:00.080 --> 00:07:02.469
framework in that first episode about
00:07:02.479 --> 00:07:06.070
how LLMs work. But for our purposes here
00:07:06.080 --> 00:07:09.270
when I'm talking about AI and LLMs, I'm
00:07:09.280 --> 00:07:10.230
actually going to use those
00:07:10.240 --> 00:07:12.710
interchangeably here. They are not
00:07:12.720 --> 00:07:14.950
actually interchangeable, but I have
00:07:14.960 --> 00:07:16.309
talked to enough clients, talked to
00:07:16.319 --> 00:07:19.350
enough authors to know that typically we
00:07:19.360 --> 00:07:21.430
are thinking of those as one thing. Now,
00:07:21.440 --> 00:07:22.710
I'm not going to fight it. We're just
00:07:22.720 --> 00:07:24.629
going to go with it. So, if you have not
00:07:24.639 --> 00:07:26.629
listened to last week's episode, go pick
00:07:26.639 --> 00:07:29.670
that up for now. When I say LLM, large
00:07:29.680 --> 00:07:31.749
language model, I just simply mean
00:07:31.759 --> 00:07:34.550
something like chat GPT or Claude. any
00:07:34.560 --> 00:07:37.110
of these um generative AI programs are
00:07:37.120 --> 00:07:40.070
typically trained as an LLM large
00:07:40.080 --> 00:07:42.710
language model. Okay. So the first stage
00:07:42.720 --> 00:07:44.790
that I want to start at as I said is the
00:07:44.800 --> 00:07:47.749
writing stage. Now most of the folks
00:07:47.759 --> 00:07:49.830
that I talked to aren't planning to use
00:07:49.840 --> 00:07:52.790
AI to write their whole manuscript. It's
00:07:52.800 --> 00:07:54.230
often they're kind of wanting to use it
00:07:54.240 --> 00:07:56.710
for bits and pieces. But this first
00:07:56.720 --> 00:07:58.869
author that I worked with for a while
00:07:58.879 --> 00:08:01.350
was somebody that had kind of imagined,
00:08:01.360 --> 00:08:03.830
I think, that she would use the LLM for
00:08:03.840 --> 00:08:07.029
the entire manuscript. Now, this is a
00:08:07.039 --> 00:08:09.430
while ago that this happened, so I'm not
00:08:09.440 --> 00:08:10.950
sure if this would have gone the same
00:08:10.960 --> 00:08:13.589
way now, but effectively what happened
00:08:13.599 --> 00:08:16.230
was as we had started to work together.
00:08:16.240 --> 00:08:19.510
She was putting out chapters by asking
00:08:19.520 --> 00:08:21.189
Claude or I'm actually not sure, Claude
00:08:21.199 --> 00:08:23.510
or Chat GPT, whichever it was, to write
00:08:23.520 --> 00:08:25.510
her chapter and then she would copy and
00:08:25.520 --> 00:08:28.070
paste it into a Microsoft Word document.
00:08:28.080 --> 00:08:29.909
Now, I did not know that this was
00:08:29.919 --> 00:08:33.350
happening at first, but it became very
00:08:33.360 --> 00:08:35.350
obvious to me. And the way it became
00:08:35.360 --> 00:08:38.070
obvious to me was that I was seeing the
00:08:38.080 --> 00:08:41.750
exact same choices being made over and
00:08:41.760 --> 00:08:44.550
over again. Now, if you have not watched
00:08:44.560 --> 00:08:46.790
that episode from last week, let me tell
00:08:46.800 --> 00:08:49.829
you this is actually a signature of the
00:08:49.839 --> 00:08:52.470
LLM of a large language model generative
00:08:52.480 --> 00:08:56.230
AI type of writing. Unfortunately,
00:08:56.240 --> 00:08:59.350
because what the LLM is trying to do is
00:08:59.360 --> 00:09:01.509
make the most probable choices with the
00:09:01.519 --> 00:09:04.870
highest level of efficiency. That means
00:09:04.880 --> 00:09:07.190
that if you like a choice it makes, it
00:09:07.200 --> 00:09:08.949
will keep making that choice over and
00:09:08.959 --> 00:09:11.509
over and over and over and over again.
00:09:11.519 --> 00:09:13.750
And unfortunately, even if you don't
00:09:13.760 --> 00:09:15.670
like a choice it makes, even if you tell
00:09:15.680 --> 00:09:17.990
it you don't like a choice it makes,
00:09:18.000 --> 00:09:20.310
what we have found in research across
00:09:20.320 --> 00:09:22.870
the board is that human writers will
00:09:22.880 --> 00:09:26.550
make a much wider range of choices than
00:09:26.560 --> 00:09:29.430
an LLM, than a generative AI. So
00:09:29.440 --> 00:09:31.430
literally what was happening as I was
00:09:31.440 --> 00:09:33.030
reading through this AI written
00:09:33.040 --> 00:09:35.990
manuscript is I would not only see the
00:09:36.000 --> 00:09:37.990
same sentence sometimes appearing
00:09:38.000 --> 00:09:40.790
multiple times, I would also see the
00:09:40.800 --> 00:09:43.829
exact same formula on the page. And I
00:09:43.839 --> 00:09:45.509
bet you've seen it too. I bet I don't
00:09:45.519 --> 00:09:47.590
even need to go into such depth because
00:09:47.600 --> 00:09:51.670
it literally was the exact same pattern.
00:09:51.680 --> 00:09:53.990
It would start with a setup and then it
00:09:54.000 --> 00:09:55.750
would go to this is what this means
00:09:55.760 --> 00:09:57.269
colon and make a really bold
00:09:57.279 --> 00:09:59.590
declaration. Then it would give a list
00:09:59.600 --> 00:10:03.190
of synonyms. Effectively, every single
00:10:03.200 --> 00:10:05.910
page looked exactly the same. There was
00:10:05.920 --> 00:10:08.150
repeated information. There was also
00:10:08.160 --> 00:10:10.470
contradicting information, of course.
00:10:10.480 --> 00:10:13.190
But the repetition made it so I could
00:10:13.200 --> 00:10:15.269
not even edit the work. I actually had
00:10:15.279 --> 00:10:17.030
to tell her I could not work with her
00:10:17.040 --> 00:10:20.310
because even though this was clearly AI
00:10:20.320 --> 00:10:22.630
generated, she was pretty dug in that
00:10:22.640 --> 00:10:25.269
she wanted to keep using AI. And at some
00:10:25.279 --> 00:10:28.389
point there is really no editing that
00:10:28.399 --> 00:10:30.389
when the choices have been made at the
00:10:30.399 --> 00:10:32.470
sentence and paragraph and page level
00:10:32.480 --> 00:10:36.389
structure to be so repetitive, you can't
00:10:36.399 --> 00:10:38.069
edit it. You're going to have to rewrite
00:10:38.079 --> 00:10:41.030
it again. And unfortunately, one thing I
00:10:41.040 --> 00:10:43.190
have seen is that when folks are using
00:10:43.200 --> 00:10:45.750
AI, it is very very difficult for them
00:10:45.760 --> 00:10:49.030
to write something then as a human. So,
00:10:49.040 --> 00:10:51.509
even if this author had decided she
00:10:51.519 --> 00:10:53.829
didn't want to use an LLM anymore, I
00:10:53.839 --> 00:10:55.430
don't know whether she would have been
00:10:55.440 --> 00:10:57.829
able to write it as a human because
00:10:57.839 --> 00:11:00.230
she'd already put all of those stories
00:11:00.240 --> 00:11:02.630
into Claude or Chad GBT or whatever it
00:11:02.640 --> 00:11:05.269
was. So, this is I know a little bit of
00:11:05.279 --> 00:11:07.910
an unbounded type of story just because
00:11:07.920 --> 00:11:09.750
we kind of don't know the outcome. We
00:11:09.760 --> 00:11:11.110
kind of don't know a lot that was
00:11:11.120 --> 00:11:13.670
happening behind the scenes here. All I
00:11:13.680 --> 00:11:16.230
can tell you is that this writing was so
00:11:16.240 --> 00:11:20.310
obviously AI to me and she disagreed.
00:11:20.320 --> 00:11:22.389
And you know, maybe she was not using an
00:11:22.399 --> 00:11:25.430
LLM. I really can't know for certain,
00:11:25.440 --> 00:11:27.269
but it seemed pretty clear to me that
00:11:27.279 --> 00:11:29.350
she was. And it seemed pretty clear to
00:11:29.360 --> 00:11:32.150
me that she had gotten stuck in an AI
00:11:32.160 --> 00:11:34.470
loop. Now, this is one thing that I
00:11:34.480 --> 00:11:36.150
really want to give you as kind of a
00:11:36.160 --> 00:11:37.430
takeaway here. So, if you're
00:11:37.440 --> 00:11:40.069
multitasking, come back to me. One of
00:11:40.079 --> 00:11:41.910
the things that I have seen, not
00:11:41.920 --> 00:11:44.550
directly tested in studies, but I've
00:11:44.560 --> 00:11:46.230
seen enough implications that would
00:11:46.240 --> 00:11:49.910
point toward this issue is that I really
00:11:49.920 --> 00:11:52.949
suspect that there is an amount of LLM
00:11:52.959 --> 00:11:55.590
use that will stop you from being able
00:11:55.600 --> 00:11:58.470
to produce your own writing. And I want
00:11:58.480 --> 00:11:59.990
to put this in a context. Someone shared
00:12:00.000 --> 00:12:01.910
this with me on LinkedIn and I found it
00:12:01.920 --> 00:12:05.509
so so helpful. They were saying, "Do you
00:12:05.519 --> 00:12:08.550
remember I do remember back when you
00:12:08.560 --> 00:12:11.190
didn't have any GPS in your car?" And at
00:12:11.200 --> 00:12:12.710
one point then you could kind of print a
00:12:12.720 --> 00:12:14.470
map quest or something like that, but
00:12:14.480 --> 00:12:16.389
you didn't have anything directing you
00:12:16.399 --> 00:12:19.269
on the ground. And at that point, if
00:12:19.279 --> 00:12:21.110
you're like me, you pretty much just
00:12:21.120 --> 00:12:22.629
knew where you were going. You learned
00:12:22.639 --> 00:12:23.990
the different routes. You learned how
00:12:24.000 --> 00:12:25.430
the streets intersected. You learned
00:12:25.440 --> 00:12:27.430
your city's grid. All this type of
00:12:27.440 --> 00:12:29.750
stuff. Well, now, do you know your whole
00:12:29.760 --> 00:12:32.389
city's grid? Now, do you know easily how
00:12:32.399 --> 00:12:34.790
to get to any place you want to go? I'll
00:12:34.800 --> 00:12:37.350
tell you, I don't. I use GPS for almost
00:12:37.360 --> 00:12:39.910
everything. Although, after that fellow
00:12:39.920 --> 00:12:41.430
shared that with me on LinkedIn, I have
00:12:41.440 --> 00:12:43.190
tried to be a little bit more cognizant
00:12:43.200 --> 00:12:46.949
of using my own brain. My point is, I
00:12:46.959 --> 00:12:49.190
don't know why we would assume that a
00:12:49.200 --> 00:12:51.750
GPS can stop us from knowing how to get
00:12:51.760 --> 00:12:54.310
places in our town, but an LLM can't
00:12:54.320 --> 00:12:56.629
stop us from knowing how to write. I
00:12:56.639 --> 00:12:58.069
think they're pretty much the same
00:12:58.079 --> 00:13:00.790
mechanism and based on the studies that
00:13:00.800 --> 00:13:03.030
I have read about AI, which is a lot at
00:13:03.040 --> 00:13:05.750
this point, there is evidence pointing
00:13:05.760 --> 00:13:09.190
toward the AI making us less effective
00:13:09.200 --> 00:13:11.990
as decision makers. That's all writing
00:13:12.000 --> 00:13:15.509
is a series of choices. So, this is the
00:13:15.519 --> 00:13:17.829
biggest caution that I have for y'all is
00:13:17.839 --> 00:13:21.030
pulling an LLM or AI into the writing
00:13:21.040 --> 00:13:22.710
process.
00:13:22.720 --> 00:13:24.389
Now I want to give a couple of examples
00:13:24.399 --> 00:13:26.310
for my second point which is the editing
00:13:26.320 --> 00:13:28.470
process. And the first of these is kind
00:13:28.480 --> 00:13:31.030
of stuck between writing and editing
00:13:31.040 --> 00:13:33.509
because this is editing that happened
00:13:33.519 --> 00:13:35.750
before this author brought their
00:13:35.760 --> 00:13:38.710
manuscript to us. Now with this
00:13:38.720 --> 00:13:40.310
particular author, this is such a good
00:13:40.320 --> 00:13:42.710
book by the way and we are working now
00:13:42.720 --> 00:13:47.030
to deiify it. So that is something that
00:13:47.040 --> 00:13:49.350
can happen. We can do that. we can pull
00:13:49.360 --> 00:13:51.350
you out of that loop if you have gotten
00:13:51.360 --> 00:13:54.870
stuck in AI land. But I'm so excited
00:13:54.880 --> 00:13:57.269
that this author is willing to work on
00:13:57.279 --> 00:13:59.030
deiifying
00:13:59.040 --> 00:14:02.069
because this manuscript was fantastic.
00:14:02.079 --> 00:14:04.470
Now, I could tell that she had used AI
00:14:04.480 --> 00:14:07.269
in some places, but I also felt pretty
00:14:07.279 --> 00:14:08.629
confident that there were places that
00:14:08.639 --> 00:14:11.990
she had not used AI. This is just a gut
00:14:12.000 --> 00:14:13.990
level reaction from me. As someone
00:14:14.000 --> 00:14:15.829
though who has been working with authors
00:14:15.839 --> 00:14:18.470
for a really long time, who has read a
00:14:18.480 --> 00:14:20.470
lot of firstdraft manuscripts, both
00:14:20.480 --> 00:14:23.189
very, very clean and very, very messy.
00:14:23.199 --> 00:14:25.189
There were parts in there that I felt
00:14:25.199 --> 00:14:27.110
really certain were human, but there
00:14:27.120 --> 00:14:28.870
were also parts that she had shared with
00:14:28.880 --> 00:14:32.310
me that they were AI written. Now, this
00:14:32.320 --> 00:14:34.150
author had submitted her manuscript for
00:14:34.160 --> 00:14:36.389
a manuscript review, and that is the
00:14:36.399 --> 00:14:38.310
first stage really in working with us in
00:14:38.320 --> 00:14:40.389
most capacities. if you have a
00:14:40.399 --> 00:14:42.150
manuscript. You can read more about
00:14:42.160 --> 00:14:43.030
that, by the way, at
00:14:43.040 --> 00:14:46.230
pageandpodium.com/manuscript.
00:14:46.240 --> 00:14:48.230
So, she'd submitted her manuscript for a
00:14:48.240 --> 00:14:50.790
review. And when we met to discuss the
00:14:50.800 --> 00:14:52.310
review, to make a game plan moving
00:14:52.320 --> 00:14:54.389
forward, decide what the book needed,
00:14:54.399 --> 00:14:57.350
all of that, she said to me, "You know
00:14:57.360 --> 00:14:59.509
what? I went back and looked at those
00:14:59.519 --> 00:15:02.230
passages that you pointed out as AI, and
00:15:02.240 --> 00:15:03.910
you're right. They don't sound at all
00:15:03.920 --> 00:15:06.790
like me." And they didn't. And this is
00:15:06.800 --> 00:15:08.470
the thing that I want to pull to your
00:15:08.480 --> 00:15:12.230
attention because I suspect that most of
00:15:12.240 --> 00:15:14.790
the time when people are using AI in
00:15:14.800 --> 00:15:16.870
spots. So maybe it's a paragraph you
00:15:16.880 --> 00:15:18.629
just you can't quite get your head
00:15:18.639 --> 00:15:20.389
around. Maybe it's something you're
00:15:20.399 --> 00:15:21.990
trying to summarize that you don't know
00:15:22.000 --> 00:15:23.670
very well and it kind of seems like a
00:15:23.680 --> 00:15:25.590
shortcut. Maybe you're wanting your
00:15:25.600 --> 00:15:27.750
scene to be very very exciting, but
00:15:27.760 --> 00:15:29.430
you're not sure how to write an exciting
00:15:29.440 --> 00:15:31.670
scene. All three examples I've seen
00:15:31.680 --> 00:15:34.230
before, by the way. What I see very
00:15:34.240 --> 00:15:36.470
often is that that will seem like such a
00:15:36.480 --> 00:15:38.949
good idea. LLM will give you the output.
00:15:38.959 --> 00:15:40.710
They'll put it into their manuscript.
00:15:40.720 --> 00:15:43.509
Sometimes edited a bit, sometimes not.
00:15:43.519 --> 00:15:46.069
But then when they come back to it, it
00:15:46.079 --> 00:15:49.350
is not nearly as good of writing as they
00:15:49.360 --> 00:15:52.230
thought. So what is happening here?
00:15:52.240 --> 00:15:54.389
Well, I suspect that there is something
00:15:54.399 --> 00:15:57.269
to be said about the speed of generating
00:15:57.279 --> 00:15:59.350
the words on the page. And actually a
00:15:59.360 --> 00:16:01.189
client said this to me the other day. He
00:16:01.199 --> 00:16:02.389
said, "Well, this gets rid of the
00:16:02.399 --> 00:16:05.350
blinking cursor problem." That is true.
00:16:05.360 --> 00:16:07.269
And the blinking cursor problem, that
00:16:07.279 --> 00:16:08.790
thing where you're just staring at the
00:16:08.800 --> 00:16:11.030
blinking cursor on a blank page and
00:16:11.040 --> 00:16:13.110
really talking yourself out of whether
00:16:13.120 --> 00:16:15.590
you can do this or not, that's a real
00:16:15.600 --> 00:16:18.069
phenomenon. And what I suspect is
00:16:18.079 --> 00:16:21.189
happening is that people are trading off
00:16:21.199 --> 00:16:23.110
getting rid of that blinking cursor
00:16:23.120 --> 00:16:25.269
syndrome in exchange for something that
00:16:25.279 --> 00:16:27.829
they know is probably not going to be as
00:16:27.839 --> 00:16:31.350
good of an output. But then the AI is
00:16:31.360 --> 00:16:33.670
almost magic, right? All of those words
00:16:33.680 --> 00:16:35.269
show up on the page and they're not as
00:16:35.279 --> 00:16:36.870
bad as you thought they might be. In
00:16:36.880 --> 00:16:39.350
fact, maybe they could work. And then
00:16:39.360 --> 00:16:41.350
what happens? We drop them in our
00:16:41.360 --> 00:16:43.430
manuscript. We forget about it. And then
00:16:43.440 --> 00:16:46.310
when we come back to it later, oh, maybe
00:16:46.320 --> 00:16:48.790
that looks a little different. Now we're
00:16:48.800 --> 00:16:50.949
suddenly not so impressed with this
00:16:50.959 --> 00:16:53.350
lightning fast output. we're not so
00:16:53.360 --> 00:16:56.230
impressed with the particular voice
00:16:56.240 --> 00:16:58.710
because again there are pretty major
00:16:58.720 --> 00:17:00.310
issues with the writing that comes out
00:17:00.320 --> 00:17:04.150
of an AI out of an LLM. So I suspect
00:17:04.160 --> 00:17:06.069
this is something that will be relatable
00:17:06.079 --> 00:17:08.549
if you were somebody who had dabbled at
00:17:08.559 --> 00:17:12.069
all in using an LLM using AI for your
00:17:12.079 --> 00:17:15.590
creative writing. So when you are moving
00:17:15.600 --> 00:17:17.350
forward as you are thinking of editing
00:17:17.360 --> 00:17:19.270
your manuscript I really just want to
00:17:19.280 --> 00:17:21.990
discourage you from bailing out at that
00:17:22.000 --> 00:17:24.390
point that it seems like it's too hard.
00:17:24.400 --> 00:17:27.990
That is in fact the exact point of
00:17:28.000 --> 00:17:30.390
learning. I've said often on this
00:17:30.400 --> 00:17:32.390
channel before when you are writing a
00:17:32.400 --> 00:17:34.950
book especially a memoir but really any
00:17:34.960 --> 00:17:37.830
book that is using your lived experience
00:17:37.840 --> 00:17:40.070
a lot of what you are doing is building
00:17:40.080 --> 00:17:43.590
your capacity for discomfort. Well, when
00:17:43.600 --> 00:17:46.150
we use AI in this way where we can't
00:17:46.160 --> 00:17:47.909
figure out just this one paragraph or
00:17:47.919 --> 00:17:50.310
just this one sentence, that is
00:17:50.320 --> 00:17:52.630
effectively shortcircuiting your ability
00:17:52.640 --> 00:17:55.350
to expand that capacity for discomfort
00:17:55.360 --> 00:17:57.590
because now you are not sitting in the
00:17:57.600 --> 00:17:59.350
discomfort figuring out what the answer
00:17:59.360 --> 00:18:01.990
is. You're skipping the discomfort to
00:18:02.000 --> 00:18:04.150
get the answer, but you still don't
00:18:04.160 --> 00:18:06.230
know. You still haven't practiced how to
00:18:06.240 --> 00:18:07.830
get that answer, right? you still
00:18:07.840 --> 00:18:10.390
haven't had to work your way through to
00:18:10.400 --> 00:18:12.630
build those synapses in your brain so
00:18:12.640 --> 00:18:14.870
that the next time you hit that problem,
00:18:14.880 --> 00:18:17.990
you now know how to solve it. Instead,
00:18:18.000 --> 00:18:20.230
you offloaded that decision-making. You
00:18:20.240 --> 00:18:22.710
offloaded that skill of making a
00:18:22.720 --> 00:18:26.549
difficult choice. So pull that back in
00:18:26.559 --> 00:18:29.430
so that in your first draft you can push
00:18:29.440 --> 00:18:31.350
through that discomfort and continue
00:18:31.360 --> 00:18:34.789
building this very very important muscle
00:18:34.799 --> 00:18:38.230
of decision-m in your writing. Okay, I
00:18:38.240 --> 00:18:40.310
have a second example from this editing
00:18:40.320 --> 00:18:41.830
bucket that I want to share with you
00:18:41.840 --> 00:18:43.590
that is at a different level. So we've
00:18:43.600 --> 00:18:45.909
got the voice level, but I'm also seeing
00:18:45.919 --> 00:18:48.390
some AI troubles with the structure
00:18:48.400 --> 00:18:50.870
level. Now, one of these really
00:18:50.880 --> 00:18:52.870
surprised me actually, and this is from
00:18:52.880 --> 00:18:54.950
a book, this is a little while ago that
00:18:54.960 --> 00:18:57.190
we had ghostritten. Now, the woman that
00:18:57.200 --> 00:18:59.909
had hired us to ghostwrite this memoir,
00:18:59.919 --> 00:19:02.950
it is a highly, highly traumatic memoir.
00:19:02.960 --> 00:19:05.270
Um, it has a lot of sexual trauma, a lot
00:19:05.280 --> 00:19:08.310
of domestic abuse, a lot of um intimate
00:19:08.320 --> 00:19:11.029
partner violence, that kind of thing all
00:19:11.039 --> 00:19:13.430
throughout. But it's quite a beautiful
00:19:13.440 --> 00:19:15.669
book. It's quite powerful. Now, of
00:19:15.679 --> 00:19:18.549
course, our ghostriters do not use LLMs
00:19:18.559 --> 00:19:21.270
or AI generated text in ghostwritten
00:19:21.280 --> 00:19:23.750
work. But what had happened in this case
00:19:23.760 --> 00:19:25.830
was our ghostwriter had written the
00:19:25.840 --> 00:19:29.029
chapter for this client. And as we had
00:19:29.039 --> 00:19:30.710
moved through the book, she had really
00:19:30.720 --> 00:19:32.950
been so so happy with the output. She
00:19:32.960 --> 00:19:34.549
had really been loving it. But you know
00:19:34.559 --> 00:19:36.310
what I have talked about before just
00:19:36.320 --> 00:19:39.029
last month? Often times when we start to
00:19:39.039 --> 00:19:41.350
get to the end of the book, things get
00:19:41.360 --> 00:19:43.990
kind of scary. And I think that was part
00:19:44.000 --> 00:19:46.549
of what happened here because we had
00:19:46.559 --> 00:19:49.510
delivered our final chapter to her. And
00:19:49.520 --> 00:19:51.270
I had read this chapter as I do with all
00:19:51.280 --> 00:19:53.270
of the ghostritten materials in page and
00:19:53.280 --> 00:19:54.870
podium. I had reviewed it before it went
00:19:54.880 --> 00:19:57.510
out and I really liked it. There was a
00:19:57.520 --> 00:19:58.950
really interesting device that the
00:19:58.960 --> 00:20:01.430
ghostwriter had used here and that was
00:20:01.440 --> 00:20:04.070
that even though this was a memoir. So
00:20:04.080 --> 00:20:05.669
throughout the memoir, it had been in
00:20:05.679 --> 00:20:08.310
first person. As she moved into the
00:20:08.320 --> 00:20:10.070
final chapter, she was kind of
00:20:10.080 --> 00:20:13.350
transitioning into a second person, a U
00:20:13.360 --> 00:20:16.070
so that she could on behalf of this
00:20:16.080 --> 00:20:18.630
author directly address the reader and
00:20:18.640 --> 00:20:20.150
just really make sure that the
00:20:20.160 --> 00:20:22.150
connection was very clear between what
00:20:22.160 --> 00:20:24.870
she was urging the reader to do and what
00:20:24.880 --> 00:20:27.830
she had experienced herself. Now this
00:20:27.840 --> 00:20:29.510
second person device is not something
00:20:29.520 --> 00:20:32.070
that we had created at page and podium
00:20:32.080 --> 00:20:34.870
and in fact in our um formerly our book
00:20:34.880 --> 00:20:37.029
club. We had even discussed a book that
00:20:37.039 --> 00:20:39.990
had used this device and it works so so
00:20:40.000 --> 00:20:43.430
well in particular types of books. I
00:20:43.440 --> 00:20:44.950
really liked the choice. I thought it
00:20:44.960 --> 00:20:46.710
was really interesting and I thought it
00:20:46.720 --> 00:20:48.789
pulled the reader in in a more direct
00:20:48.799 --> 00:20:50.710
type of way that really worked well in
00:20:50.720 --> 00:20:53.510
my opinion for this book. Well, the
00:20:53.520 --> 00:20:55.270
author didn't like that second person
00:20:55.280 --> 00:20:57.510
use. And you know what, friends? That's
00:20:57.520 --> 00:20:59.590
okay. We're not always going to agree.
00:20:59.600 --> 00:21:01.669
When folks work with us on ghostriting,
00:21:01.679 --> 00:21:03.990
we are never forcing you to accept a
00:21:04.000 --> 00:21:06.310
choice you don't like. But the reason
00:21:06.320 --> 00:21:08.710
that I bring this up is that typically
00:21:08.720 --> 00:21:11.029
how we will do this process is if there
00:21:11.039 --> 00:21:12.870
is something in the manuscript that the
00:21:12.880 --> 00:21:14.710
author doesn't like, they will bring
00:21:14.720 --> 00:21:16.630
that to the ghost writer and just try to
00:21:16.640 --> 00:21:18.710
articulate what they didn't like about
00:21:18.720 --> 00:21:22.549
it. Now, can people who are nonriters
00:21:22.559 --> 00:21:24.950
always successfully communicate what it
00:21:24.960 --> 00:21:27.350
was they didn't like? Not always,
00:21:27.360 --> 00:21:29.350
because there's a lot of terminology, a
00:21:29.360 --> 00:21:31.350
lot of different structural things. If
00:21:31.360 --> 00:21:33.029
you're hiring a ghostriter, you're not
00:21:33.039 --> 00:21:35.110
really in that type of world. That makes
00:21:35.120 --> 00:21:37.669
it really hard. But the fact that an
00:21:37.679 --> 00:21:39.190
author struggles to communicate with
00:21:39.200 --> 00:21:40.710
their ghostriter is not a deal
00:21:40.720 --> 00:21:42.950
breakaker. In fact, we are all used to
00:21:42.960 --> 00:21:45.830
that as ghostriters. And in fact, one of
00:21:45.840 --> 00:21:49.510
the things we crucially need is to hear
00:21:49.520 --> 00:21:52.310
you struggle through why you don't like
00:21:52.320 --> 00:21:54.390
something or why you do like something.
00:21:54.400 --> 00:21:57.110
The reason is often times it's just the
00:21:57.120 --> 00:21:59.350
like and don't like at the surface, but
00:21:59.360 --> 00:22:01.350
there's something deeper that we need to
00:22:01.360 --> 00:22:03.510
address to really make the manuscript
00:22:03.520 --> 00:22:05.029
work. This is collaborative
00:22:05.039 --> 00:22:06.870
ghostwriting, right? We are building
00:22:06.880 --> 00:22:09.190
this out together.
00:22:09.200 --> 00:22:12.149
But because in this case AI is so
00:22:12.159 --> 00:22:14.149
accessible, what this author had
00:22:14.159 --> 00:22:16.549
actually done was put the chapter into
00:22:16.559 --> 00:22:19.350
an LLM and coached it. Said, "I don't
00:22:19.360 --> 00:22:21.350
like this use of you. Can you tell me
00:22:21.360 --> 00:22:24.310
why I don't like it?" Well, of course,
00:22:24.320 --> 00:22:27.590
this LLM just spit out to her a totally
00:22:27.600 --> 00:22:29.990
clean response that never ever should
00:22:30.000 --> 00:22:32.070
you use second person in memoir. No one
00:22:32.080 --> 00:22:33.830
has ever done it. It is not consistent
00:22:33.840 --> 00:22:37.270
with the genre. It's bad writing. Y'all,
00:22:37.280 --> 00:22:40.789
please don't do this. The problem here
00:22:40.799 --> 00:22:43.510
is that not only do we now not know
00:22:43.520 --> 00:22:45.750
what's actually happening beneath the
00:22:45.760 --> 00:22:48.390
surface of this author's reaction to her
00:22:48.400 --> 00:22:51.909
manuscript, we now also have to reassure
00:22:51.919 --> 00:22:54.149
this author that no, in fact, this was
00:22:54.159 --> 00:22:56.070
not a careless decision. This was not a
00:22:56.080 --> 00:22:57.830
decision that was made lightly. This was
00:22:57.840 --> 00:22:59.830
not a decision that was based on a lack
00:22:59.840 --> 00:23:01.830
of education.
00:23:01.840 --> 00:23:04.630
But because AI had told her that it was,
00:23:04.640 --> 00:23:07.430
it became a much much more difficult
00:23:07.440 --> 00:23:10.149
problem to work through than if she had
00:23:10.159 --> 00:23:13.350
just come to us and allowed us to help
00:23:13.360 --> 00:23:15.350
talk through how we might make this
00:23:15.360 --> 00:23:17.750
problem a solution for her, how we might
00:23:17.760 --> 00:23:20.390
make this feel better as her book, which
00:23:20.400 --> 00:23:24.070
of course is our goal. So in this case,
00:23:24.080 --> 00:23:26.630
and in many others, by the way, what's
00:23:26.640 --> 00:23:28.710
happening is the outsourcing of the
00:23:28.720 --> 00:23:30.950
review process. Now, I put this in
00:23:30.960 --> 00:23:32.149
editing even though this is a
00:23:32.159 --> 00:23:34.630
ghostwriting problem because I have seen
00:23:34.640 --> 00:23:36.549
this with other authors who also wrote
00:23:36.559 --> 00:23:39.110
their full manuscript where when they
00:23:39.120 --> 00:23:41.270
got feedback back from an editor, they
00:23:41.280 --> 00:23:43.669
didn't like it, but instead of talking
00:23:43.679 --> 00:23:46.149
through it to see if maybe there was a
00:23:46.159 --> 00:23:47.990
different edit that could solve the same
00:23:48.000 --> 00:23:50.470
reader problem, the whole thing goes
00:23:50.480 --> 00:23:52.710
into Claude and the response comes
00:23:52.720 --> 00:23:54.230
directly back to the editor or
00:23:54.240 --> 00:23:56.870
ghostwriter. And y'all, not only is this
00:23:56.880 --> 00:23:58.310
going to hurt your relationship with
00:23:58.320 --> 00:24:00.789
your editor or ghostwriter, it also is
00:24:00.799 --> 00:24:03.110
very likely limiting choices that you
00:24:03.120 --> 00:24:05.029
actually might have liked if you had
00:24:05.039 --> 00:24:07.590
come to the table to talk about things.
00:24:07.600 --> 00:24:10.149
So when you are thinking about this
00:24:10.159 --> 00:24:13.430
editing process, this is a real area
00:24:13.440 --> 00:24:15.990
where the kind of yes man syndrome of a
00:24:16.000 --> 00:24:18.470
lot of these LLMs can really come into
00:24:18.480 --> 00:24:22.070
play. If the LLM is getting input that
00:24:22.080 --> 00:24:24.870
you don't like something, it will almost
00:24:24.880 --> 00:24:27.510
always give you feedback that you are
00:24:27.520 --> 00:24:29.750
right not to like it. And in fact, here
00:24:29.760 --> 00:24:33.669
are all the reasons. So again, as with
00:24:33.679 --> 00:24:36.149
the last editing example, this is really
00:24:36.159 --> 00:24:37.750
something that's going to prevent you
00:24:37.760 --> 00:24:39.590
from getting the practice in this
00:24:39.600 --> 00:24:42.149
decision-making that you really need if
00:24:42.159 --> 00:24:43.909
you're going to move forward with a
00:24:43.919 --> 00:24:46.149
career in writing.
00:24:46.159 --> 00:24:47.909
Okay. The third step that I want to move
00:24:47.919 --> 00:24:51.430
to is of course design. Now this one is
00:24:51.440 --> 00:24:54.310
one that probably is the first encounter
00:24:54.320 --> 00:24:57.750
I ever had with AI and now I have had
00:24:57.760 --> 00:25:00.230
many many more and the point that often
00:25:00.240 --> 00:25:03.430
times AI will come up in design is with
00:25:03.440 --> 00:25:06.070
cover design. Now I've talked many many
00:25:06.080 --> 00:25:08.630
times on this channel. I know covers are
00:25:08.640 --> 00:25:11.510
so special. They're so personal. They're
00:25:11.520 --> 00:25:14.470
so important. They are quite literally
00:25:14.480 --> 00:25:17.350
the face of your book. And for that
00:25:17.360 --> 00:25:19.830
reason, people do get very, very
00:25:19.840 --> 00:25:22.310
emotionally invested in their covers.
00:25:22.320 --> 00:25:23.909
Remember what I said at the beginning of
00:25:23.919 --> 00:25:26.310
this video and in last week's video.
00:25:26.320 --> 00:25:28.390
Often times, the place that AI will get
00:25:28.400 --> 00:25:31.190
you in trouble is when its certainty
00:25:31.200 --> 00:25:33.830
meets your insecurity. Well,
00:25:33.840 --> 00:25:36.070
unfortunately, all of us have
00:25:36.080 --> 00:25:38.789
insecurities around this whole process.
00:25:38.799 --> 00:25:41.510
And often it does feel very reassuring
00:25:41.520 --> 00:25:44.630
to be able to feel so certain. Well, I
00:25:44.640 --> 00:25:46.470
want to give you an example of a book
00:25:46.480 --> 00:25:48.230
cover that ended with a very happy
00:25:48.240 --> 00:25:50.390
ending. Now, this is um not a book that
00:25:50.400 --> 00:25:52.470
is a page and podium book. This is a
00:25:52.480 --> 00:25:53.909
book that's published through it's being
00:25:53.919 --> 00:25:55.430
published through our partnership
00:25:55.440 --> 00:25:57.990
publishing process, which is where you
00:25:58.000 --> 00:26:00.070
can retain all of the rights, all of the
00:26:00.080 --> 00:26:01.430
distribution. you are doing
00:26:01.440 --> 00:26:04.070
self-publishing effectively but with the
00:26:04.080 --> 00:26:05.990
guidance of our team with that same
00:26:06.000 --> 00:26:08.710
process that our team uses.
00:26:08.720 --> 00:26:10.950
So in this case we had gone all the way
00:26:10.960 --> 00:26:12.870
through this excellent manuscript all
00:26:12.880 --> 00:26:14.789
the way through this excellent book. We
00:26:14.799 --> 00:26:18.070
had gotten to the cover design area. Now
00:26:18.080 --> 00:26:19.590
the way that this process works at page
00:26:19.600 --> 00:26:21.350
and podium is that our cover designer
00:26:21.360 --> 00:26:24.630
will create a series of concepts. Now, a
00:26:24.640 --> 00:26:26.789
concept is not a finished book cover,
00:26:26.799 --> 00:26:28.549
but it is enough of a finished book
00:26:28.559 --> 00:26:30.950
cover that you can see the gist and then
00:26:30.960 --> 00:26:33.110
from there we've got something on paper,
00:26:33.120 --> 00:26:34.630
right? So, we can kind of manipulate
00:26:34.640 --> 00:26:36.310
some of the elements, change out a
00:26:36.320 --> 00:26:38.789
picture, whatever it might be. But the
00:26:38.799 --> 00:26:41.269
concept gives us a place to start. So,
00:26:41.279 --> 00:26:43.590
we always ask that our authors choose a
00:26:43.600 --> 00:26:45.830
concept and then we go back and forth
00:26:45.840 --> 00:26:47.909
with the designer to alter that concept
00:26:47.919 --> 00:26:49.110
to make sure that it comes into
00:26:49.120 --> 00:26:52.149
alignment. Now, we had sent the concepts
00:26:52.159 --> 00:26:54.390
to this particular author and he had
00:26:54.400 --> 00:26:57.510
written back and said, "These are great.
00:26:57.520 --> 00:26:59.510
Please thank your designer. These are
00:26:59.520 --> 00:27:01.350
really excellent." And by the way, not
00:27:01.360 --> 00:27:02.870
to brag on our designer, although I
00:27:02.880 --> 00:27:05.190
always will. Thank you, Jessica. Not to
00:27:05.200 --> 00:27:07.750
brag on our designer too much, but also
00:27:07.760 --> 00:27:09.750
this client's wife is a graphic
00:27:09.760 --> 00:27:11.510
designer. So, this is not someone that
00:27:11.520 --> 00:27:14.870
doesn't know what he's doing, but we had
00:27:14.880 --> 00:27:16.710
sent the concepts. He'd chosen a
00:27:16.720 --> 00:27:18.870
concept. And then as we went back and
00:27:18.880 --> 00:27:21.269
forth, what I noticed was that things
00:27:21.279 --> 00:27:22.789
that hadn't really even been on the
00:27:22.799 --> 00:27:24.870
table at the beginning were suddenly
00:27:24.880 --> 00:27:27.590
becoming really really important. And
00:27:27.600 --> 00:27:30.950
then the answer came through. So what he
00:27:30.960 --> 00:27:33.750
had ended up doing was putting the cover
00:27:33.760 --> 00:27:36.230
concept into AI and then coaching it to
00:27:36.240 --> 00:27:39.029
make some changes. Now when he sent this
00:27:39.039 --> 00:27:42.630
cover design back, it was a totally
00:27:42.640 --> 00:27:45.990
different cover. Totally different. In
00:27:46.000 --> 00:27:48.310
fact, this is generally a self-help type
00:27:48.320 --> 00:27:50.390
book. The book cover that he sent back
00:27:50.400 --> 00:27:52.149
would have been much more like techno
00:27:52.159 --> 00:27:54.070
dystopia. Maybe it was it would have
00:27:54.080 --> 00:27:56.950
looked much more like a fiction book. It
00:27:56.960 --> 00:27:59.430
was not the right fit for a self-help
00:27:59.440 --> 00:28:02.070
book aimed at young adults. It was very,
00:28:02.080 --> 00:28:03.909
very much something that would seem
00:28:03.919 --> 00:28:06.389
dramatic, that would seem like fiction.
00:28:06.399 --> 00:28:08.710
Now, we got this cleared up. We got this
00:28:08.720 --> 00:28:11.029
straightened out. All is well that ends
00:28:11.039 --> 00:28:12.950
well. But the reason that I want to
00:28:12.960 --> 00:28:15.510
bring this to you is this is another
00:28:15.520 --> 00:28:18.070
example of a time where not feeling
00:28:18.080 --> 00:28:20.470
totally secure that you can articulate
00:28:20.480 --> 00:28:22.230
what you want, that what you want is the
00:28:22.240 --> 00:28:25.430
right thing, often times is where AI is
00:28:25.440 --> 00:28:27.430
going to step in and tell you with
00:28:27.440 --> 00:28:31.029
certainty that your idea is a great
00:28:31.039 --> 00:28:34.149
idea. But in this case, it was very very
00:28:34.159 --> 00:28:36.070
easy to talk this author off the
00:28:36.080 --> 00:28:39.750
proverbial ledge because in our process
00:28:39.760 --> 00:28:42.230
of publishing, we always begin the
00:28:42.240 --> 00:28:44.310
design step by presenting the author
00:28:44.320 --> 00:28:46.710
with comp title covers. Now, what that
00:28:46.720 --> 00:28:48.950
means, a comp title is a comparable
00:28:48.960 --> 00:28:51.830
title. Comp title covers are just simply
00:28:51.840 --> 00:28:53.830
the covers of other books that are in
00:28:53.840 --> 00:28:56.230
the similar area on the bookshelf near
00:28:56.240 --> 00:28:58.070
your book. Now, we don't want to
00:28:58.080 --> 00:29:00.710
replicate exactly what the other covers
00:29:00.720 --> 00:29:02.630
look like, but we do need to make sure
00:29:02.640 --> 00:29:05.110
it's in the general world of what that
00:29:05.120 --> 00:29:07.269
genre is going to be if that's what your
00:29:07.279 --> 00:29:10.710
reader expects. The problem is when you
00:29:10.720 --> 00:29:13.669
pull AI in to redesign your cover, you
00:29:13.679 --> 00:29:15.669
are not designing a cover that is for
00:29:15.679 --> 00:29:17.350
your reader or your genre or the
00:29:17.360 --> 00:29:19.510
bookshelf. You are likely designing
00:29:19.520 --> 00:29:21.110
something that you have in your head or
00:29:21.120 --> 00:29:23.669
had in your heart oftentimes even before
00:29:23.679 --> 00:29:25.750
it occurred to you that you needed to be
00:29:25.760 --> 00:29:27.430
thinking about how your book cover would
00:29:27.440 --> 00:29:30.230
look in the context of the market. Now,
00:29:30.240 --> 00:29:32.389
I don't blame any authors for this. It
00:29:32.399 --> 00:29:35.190
is not actually your job to be thinking
00:29:35.200 --> 00:29:36.950
about how the cover might fit in with
00:29:36.960 --> 00:29:39.269
other covers in your genre. That's our
00:29:39.279 --> 00:29:42.549
job. But the problem is, because AI has
00:29:42.559 --> 00:29:44.710
put these quote unquote design tools at
00:29:44.720 --> 00:29:47.510
everyone's fingertips, it's just so
00:29:47.520 --> 00:29:49.750
tempting to try to show the
00:29:49.760 --> 00:29:52.870
professionals exactly what you want.
00:29:52.880 --> 00:29:55.110
Unfortunately, AI is not going to remind
00:29:55.120 --> 00:29:56.950
you that your book cover isn't just for
00:29:56.960 --> 00:29:59.110
you. AI is going to tell you you have
00:29:59.120 --> 00:30:02.310
great great ideas and put out an example
00:30:02.320 --> 00:30:03.990
that you can then send to your
00:30:04.000 --> 00:30:06.470
professional to again probably hurt the
00:30:06.480 --> 00:30:09.269
relationship a bit but also to not
00:30:09.279 --> 00:30:11.430
really help you get where you need to go
00:30:11.440 --> 00:30:14.149
anyway. So, as I mentioned, we were able
00:30:14.159 --> 00:30:16.470
to kind of pull this project back, go
00:30:16.480 --> 00:30:18.389
with that original concept that he had
00:30:18.399 --> 00:30:20.630
loved and make some tweaks to it so that
00:30:20.640 --> 00:30:23.110
he loved it even more as we move forward
00:30:23.120 --> 00:30:25.110
to the end of this partner publishing
00:30:25.120 --> 00:30:27.830
agreement. Okay? And then the last place
00:30:27.840 --> 00:30:29.909
that I want to talk about this is in
00:30:29.919 --> 00:30:32.549
your marketing. Now there are many many
00:30:32.559 --> 00:30:34.950
ways to market a book. But one of the
00:30:34.960 --> 00:30:37.029
things that we really lean on as a press
00:30:37.039 --> 00:30:39.110
that mostly works with experts, mostly
00:30:39.120 --> 00:30:40.950
works with people with a message that is
00:30:40.960 --> 00:30:43.350
very important to them. We are very
00:30:43.360 --> 00:30:45.269
often working with authors who plan to
00:30:45.279 --> 00:30:47.430
do some kind of content marketing. And
00:30:47.440 --> 00:30:48.710
if you're not familiar with the term
00:30:48.720 --> 00:30:50.710
content marketing, that simply means
00:30:50.720 --> 00:30:52.549
something like a podcast, a YouTube
00:30:52.559 --> 00:30:55.110
channel, social media, reals, that kind
00:30:55.120 --> 00:30:57.430
of thing. Effectively, we are creating
00:30:57.440 --> 00:30:59.990
content that then points our reader to
00:31:00.000 --> 00:31:02.230
our book. Our reader wants more of our
00:31:02.240 --> 00:31:04.549
content, so they buy our book. Simple.
00:31:04.559 --> 00:31:06.149
Works really, really well. And it is
00:31:06.159 --> 00:31:08.630
especially useful again if you are
00:31:08.640 --> 00:31:10.630
somebody who has a message that's close
00:31:10.640 --> 00:31:12.710
to your heart. It's going to give you a
00:31:12.720 --> 00:31:14.950
lot of opportunity and frankly a lot of
00:31:14.960 --> 00:31:17.510
push to keep building that message, keep
00:31:17.520 --> 00:31:19.350
developing that message, keep speaking
00:31:19.360 --> 00:31:21.909
to it where it is relevant.
00:31:21.919 --> 00:31:24.070
Unfortunately though, one of the things
00:31:24.080 --> 00:31:26.389
that we are seeing is that folks who are
00:31:26.399 --> 00:31:29.029
wanting to use AI in their marketing are
00:31:29.039 --> 00:31:31.269
number one not only outsourcing their
00:31:31.279 --> 00:31:34.230
whole voice to AI, which I know you have
00:31:34.240 --> 00:31:35.830
seen if you've been on social media in
00:31:35.840 --> 00:31:39.590
the last year. But bigger issue is that
00:31:39.600 --> 00:31:41.909
they're outsourcing their thinking about
00:31:41.919 --> 00:31:45.029
their topic. Now, next week's episode is
00:31:45.039 --> 00:31:47.509
going to get into why thought leadership
00:31:47.519 --> 00:31:51.269
in particular is harmed by AI use, but
00:31:51.279 --> 00:31:54.389
suffice it to say for right now, one of
00:31:54.399 --> 00:31:56.149
the things that I heard that really
00:31:56.159 --> 00:31:58.070
turned on a light bulb for me on this
00:31:58.080 --> 00:32:01.190
particular topic was my friend Jess Sto.
00:32:01.200 --> 00:32:03.190
Now, I've had Jess on the show before.
00:32:03.200 --> 00:32:05.669
She works directly with thought leaders
00:32:05.679 --> 00:32:08.149
to really help make sure that their big
00:32:08.159 --> 00:32:10.470
idea, their mission, and their vision
00:32:10.480 --> 00:32:12.549
and what they want to do in this world
00:32:12.559 --> 00:32:15.350
is central to their business. Now, just
00:32:15.360 --> 00:32:17.509
like me, has been creating content, has
00:32:17.519 --> 00:32:19.269
been doing content marketing for many,
00:32:19.279 --> 00:32:21.990
many years, finds it, like me, very
00:32:22.000 --> 00:32:24.389
fulfilling. And mostly really all we're
00:32:24.399 --> 00:32:26.149
doing is sharing what we see in the
00:32:26.159 --> 00:32:27.669
world, what we think is important, what
00:32:27.679 --> 00:32:29.830
we think you'll enjoy and learn from.
00:32:29.840 --> 00:32:31.269
So, it's a pretty pleasurable
00:32:31.279 --> 00:32:32.870
experience. It's something I really
00:32:32.880 --> 00:32:35.350
enjoy, frankly. And Jess has told me
00:32:35.360 --> 00:32:37.509
that she enjoyed it, too. Which is why I
00:32:37.519 --> 00:32:40.070
was so surprised when I started to see
00:32:40.080 --> 00:32:41.750
that she was getting less frequent with
00:32:41.760 --> 00:32:44.230
her podcast, with her newsletters. Well,
00:32:44.240 --> 00:32:47.269
then Jess put out an excellent episode
00:32:47.279 --> 00:32:49.509
on what she's doing, which is called an
00:32:49.519 --> 00:32:52.710
AI cleanse. Now, what Jess found, and
00:32:52.720 --> 00:32:54.630
you can listen to this in her podcast.
00:32:54.640 --> 00:32:56.070
It's called the Profit and Purpose
00:32:56.080 --> 00:32:58.310
Podcast, and I'll link it in the notes.
00:32:58.320 --> 00:33:00.950
What she described was that she had been
00:33:00.960 --> 00:33:03.269
using AI as I hear from many people
00:33:03.279 --> 00:33:05.590
quote just to generate ideas just to
00:33:05.600 --> 00:33:08.389
organize her thoughts. That is what she
00:33:08.399 --> 00:33:10.870
believed that she was doing. But what
00:33:10.880 --> 00:33:12.870
she found was she was struggling more
00:33:12.880 --> 00:33:15.269
and more and more to come up with even
00:33:15.279 --> 00:33:17.909
topics, let alone new ideas to talk
00:33:17.919 --> 00:33:20.230
about in her content. She just simply
00:33:20.240 --> 00:33:21.909
felt like she didn't have any ideas
00:33:21.919 --> 00:33:24.470
unless she went to AI. And of course, AI
00:33:24.480 --> 00:33:26.310
had stopped putting out any ideas that
00:33:26.320 --> 00:33:28.470
really seemed exciting or certainly to
00:33:28.480 --> 00:33:31.190
her heart. And what she found was that
00:33:31.200 --> 00:33:34.470
by taking a break from AI, she was able
00:33:34.480 --> 00:33:37.750
to resto that fire of ideas, that fire
00:33:37.760 --> 00:33:40.630
of passion that she has for her topic.
00:33:40.640 --> 00:33:43.269
So, she's working with folks now to do
00:33:43.279 --> 00:33:45.190
this type of AI cleanse. Let me
00:33:45.200 --> 00:33:49.350
recommend, check her out, jessato.com.
00:33:49.360 --> 00:33:51.830
But what I want to say here is that one
00:33:51.840 --> 00:33:53.590
of the things I think is the most
00:33:53.600 --> 00:33:56.630
sinister about this AI LLM moment that
00:33:56.640 --> 00:33:59.590
we are in is that I hear so many people
00:33:59.600 --> 00:34:02.149
saying, "Well, I'm just quote unquote
00:34:02.159 --> 00:34:06.389
just using it for X and Y and Z." But
00:34:06.399 --> 00:34:09.270
almost always the just is actually
00:34:09.280 --> 00:34:12.310
something that is much much more central
00:34:12.320 --> 00:34:14.710
to the actual project you're doing than
00:34:14.720 --> 00:34:17.430
you think. So when we say, "Oh, I'm just
00:34:17.440 --> 00:34:20.470
using it to organize my ideas." I often
00:34:20.480 --> 00:34:22.550
find that the problem is that we're
00:34:22.560 --> 00:34:24.869
undermining the actual process of
00:34:24.879 --> 00:34:27.030
organizing those ideas. And that by
00:34:27.040 --> 00:34:28.869
doing that process of organizing the
00:34:28.879 --> 00:34:31.349
ideas ourselves, we come up with new
00:34:31.359 --> 00:34:33.829
connections, new themes, new ideas, new
00:34:33.839 --> 00:34:37.030
examples. When we want that done fast
00:34:37.040 --> 00:34:39.589
instead of well, we outsource it. And
00:34:39.599 --> 00:34:42.389
what that means is we lose the ability
00:34:42.399 --> 00:34:44.869
to find all of those connections until
00:34:44.879 --> 00:34:47.109
we physically force ourselves to do it
00:34:47.119 --> 00:34:51.109
again. And that is so hard because AI is
00:34:51.119 --> 00:34:53.430
quite addictive. In the folks that I
00:34:53.440 --> 00:34:55.030
have seen, in the studies that I have
00:34:55.040 --> 00:34:58.390
read, I really do not see any way out
00:34:58.400 --> 00:35:01.030
but stepping back from the AI in all of
00:35:01.040 --> 00:35:03.829
these issues and retraining yourself,
00:35:03.839 --> 00:35:06.630
reminding yourself that there is joy and
00:35:06.640 --> 00:35:09.510
fulfillment and passion in coming up
00:35:09.520 --> 00:35:12.630
with things even if an LLM could do them
00:35:12.640 --> 00:35:15.349
faster. Especially if an LLM could do
00:35:15.359 --> 00:35:18.310
them faster. Okay, I hope this has been
00:35:18.320 --> 00:35:20.790
a helpful overview of the ways that I
00:35:20.800 --> 00:35:23.430
have seen AI come into folks projects in
00:35:23.440 --> 00:35:26.390
a way that really did not serve them.
00:35:26.400 --> 00:35:28.310
I'm certainly not trying to shame
00:35:28.320 --> 00:35:31.190
anyone. And in fact, again, if you are
00:35:31.200 --> 00:35:33.829
someone who has been, say, using AI in
00:35:33.839 --> 00:35:36.310
your business, in your work, and is now
00:35:36.320 --> 00:35:37.990
finding that your creative life is
00:35:38.000 --> 00:35:39.990
suffering, if you are somebody who has
00:35:40.000 --> 00:35:42.310
been in that LLM with your project and
00:35:42.320 --> 00:35:45.910
you regret it, come to me. I promise you
00:35:45.920 --> 00:35:49.109
we do not judge AI use. But what we will
00:35:49.119 --> 00:35:52.150
do is try to help you back out of this
00:35:52.160 --> 00:35:55.030
use so that we can recover not only your
00:35:55.040 --> 00:35:57.030
book, not only your project, although
00:35:57.040 --> 00:36:00.790
yes, but also your sense of creativity,
00:36:00.800 --> 00:36:03.109
your sense of your ability to do this
00:36:03.119 --> 00:36:05.750
and your pride in actually having
00:36:05.760 --> 00:36:08.069
written a book and not simply outsourced
00:36:08.079 --> 00:36:10.950
it to a computer database.
00:36:10.960 --> 00:36:12.870
Okay, y'all. I really want to encourage
00:36:12.880 --> 00:36:15.349
you to like, share, subscribe so that
00:36:15.359 --> 00:36:17.510
you don't miss next week's episode. You
00:36:17.520 --> 00:36:20.150
can head over to pageandpodium.com/blog
00:36:20.160 --> 00:36:21.910
to sign up for our newsletter. Make sure
00:36:21.920 --> 00:36:24.230
you never miss an episode. And
00:36:24.240 --> 00:36:26.390
otherwise, if I can help with AI or
00:36:26.400 --> 00:36:28.470
anything else, please reach out. I would
00:36:28.480 --> 00:36:30.310
love to help you bring your project to
00:36:30.320 --> 00:36:32.069
life. I hope you have a good day. I'll
00:36:32.079 --> 00:36:38.470
see you next week. And happy writing.
00:36:38.480 --> 00:36:39.910
If you've been working on your memoir
00:36:39.920 --> 00:36:41.829
for a while, but you're not making the
00:36:41.839 --> 00:36:43.990
progress you want to make, I want to
00:36:44.000 --> 00:36:46.310
offer you something really practical.
00:36:46.320 --> 00:36:48.470
The memoir method checklist is going to
00:36:48.480 --> 00:36:51.030
lay out every single step in the process
00:36:51.040 --> 00:36:53.589
from idea to finished book. And I know
00:36:53.599 --> 00:36:55.750
what you're thinking, a checklist. But
00:36:55.760 --> 00:36:57.589
truly, if you don't know what you're
00:36:57.599 --> 00:36:59.750
supposed to be doing next, there is very
00:36:59.760 --> 00:37:01.910
little chance you're going to actually
00:37:01.920 --> 00:37:04.790
do it. So, this checklist will guide
00:37:04.800 --> 00:37:07.109
your way through not only making a plan,
00:37:07.119 --> 00:37:09.349
but then executing the plan all the way
00:37:09.359 --> 00:37:12.069
through holding that book in your hands.
00:37:12.079 --> 00:37:13.349
You can download it for free at
00:37:13.359 --> 00:37:16.359
pagandpodium.com/checklist.