Episode Transcript
Transcripts are displayed as originally observed. Some content, including advertisements may have changed.
Use Ctrl + F to search
0:15
All right. Looks
0:17
like this should be good to go. Welcome... Neil,
0:20
in particular, I saw your question.
0:22
A few other people had some, but we'll
0:24
start with yours. Yeah,
0:26
thank you. I Actually I, rather than me reading
0:28
it, why don't you just explain in your own words, cause you
0:30
can do a better job.
0:31
Yeah, of course. So I've been trying to use
0:33
the basically the web search functionality
0:35
in ChatGPT to do some
0:38
kind of deeper research into different topics.
0:41
And I'll, some of it has to do with
0:43
what sort of things are trending and
0:45
kind of predictive stuff. And
0:47
when I first saw that, that was
0:49
an option I tried a couple of
0:51
different, techniques to build
0:54
the prompts. The prompts had, persona
0:56
and knowledge and traits and like steps
0:58
to the task and all that. And the first
1:01
attempt that I took at it, I
1:03
really focus in on the steps to the task.
1:05
And I would say start with these
1:08
sources that are more broad,
1:10
then dig into these sources, then dig
1:12
into these sources. Thinking like, how
1:14
would I go about researching something? And
1:17
then I put examples of URLs
1:19
to look at. And then, names of
1:21
specific companies and things like that. And
1:23
it, it didn't seem to do very
1:26
good at all with the URL
1:28
examples. So I pulled those
1:30
out and then it seemed to do better
1:33
without that information, but
1:35
I almost like by giving it maybe 11,
1:38
12 examples, yeah. It didn't
1:40
seem to do so well. I narrowed it down a little
1:42
bit more. It did better. It
1:45
was very odd. And then eventually
1:47
I just, Took the specific
1:49
examples away, made it more broad.
1:51
Only had three examples that
1:53
seemed to do the best, but it seemed like
1:55
it just didn't really go
1:57
super deep each time. It would
1:59
maybe hone in on two or three websites
2:02
and gather information from that and present
2:04
the findings. And I was really
2:06
looking for it to go
2:08
a lot deeper and I couldn't quite figure out, e
2:11
even try and continue exactly where you left
2:13
off or something like that. It would just do
2:15
the task over again and find the same thing. So
2:18
I wasn't sure what other people had done
2:20
and had success with using that capability
2:23
cuz it clearly
2:24
was a lot better than being
2:26
in barred and everything. But just,
2:29
I don't know, I was having a hard time really getting into
2:31
it and getting it to really
2:34
go to the depths
2:34
That I wanted
2:35
it to go to. Got it.
2:37
Okay. Yeah, I actually just
2:39
did a podcast episode
2:42
about this topic interesting.
2:44
Although that was a little more brainstorming
2:46
and a little less just flat out research.
2:48
But what's an example? Just
2:51
lost the video. What's an example of
2:53
the kind of stuff you were researching?
2:57
So it, it had to do with trends
2:59
in 401K plans. Okay.
3:02
Like risk trends. Okay.
3:05
Yeah. I was asking it to, take
3:09
a look at overall kind of
3:11
trends. And then I was
3:13
directing it to specific, like
3:15
law firms and things like that would typically
3:18
write about those trends
3:20
or industry organizations or things
3:22
like that. But, may I
3:25
wish I could ch I not really for compliance for,
3:27
I can't really share my screen or anything, but
3:29
Okay. Maybe something like similar along
3:31
those lines. I'm not sure if someone else has like a
3:33
particular area that they,
3:36
I don't know, maybe even AI trends or something like
3:38
that. Something that we could try to track down and
3:40
see, like we could all look at and go, oh, that's really
3:42
good. I don't know if there's another example that we could
3:44
use. Similar along those
3:46
lines. Yeah, I'm not
3:48
sure. To
3:50
me hearing, hearing like
3:53
risk for retirement funds
3:55
sounds like something that you'd be able to. Maybe
3:58
have to go into a PDF to pull out, but at least it's gonna
4:00
be there as opposed to more
4:02
of an analytical, like what directions
4:04
are AI tools trending in would be a much
4:06
more thought processy.
4:08
Yeah. Yeah. But I guess
4:11
we could start with something like
4:13
I don't know, what are the return
4:16
amounts for the top 10,
4:19
retirement plans? And then start trying
4:21
to dig from there and see, can you give
4:23
me more data? Can you explore?
4:26
Cuz one thing I have noticed is it
4:29
seems like the browsing, at
4:31
least from my experience, is much more go get
4:33
one webpage and analyze it
4:35
as opposed to go give me 30 web
4:37
pages and analyze them. Okay.
4:40
So I wonder if that might be just
4:42
part of it but let's try. So I guess.
4:44
Starting off role wise
4:47
act as a experienced
4:50
financial planner and
4:55
please go retrieve the rate
4:58
of return for the top 10
5:01
investment plans retirement
5:03
plans within
5:06
the US Cause of course that's
5:08
worth clarifying. And
5:10
then I
5:13
think we're gonna need to say yeah,
5:15
ask me questions to generate
5:17
the best output. And
5:22
actually I can already imagine it asking
5:24
what the top 10 retirement plans would
5:26
when the US are. So maybe
5:28
just based on rate of return.
5:32
It's a little recursive, but it might be okay with
5:34
that. And then No,
5:38
actually I just wanna run it from there and see what it comes
5:40
up with cuz. Okay,
5:43
good. I'll look up the most current data
5:47
specific types of retirement plans.
5:53
Oh okay. I feel like some of that's obvious
5:55
and implied, but Sure. Yeah. I
5:57
guess let's go with type as a 401k,
6:01
historical rate of return. Obviously
6:04
not future. That seems
6:06
a little silly to even ask, but I guess it. Maybe
6:08
makes sense. Would you like That would be
6:11
very
6:11
interesting to see what it comes up with cuz
6:13
that's yeah. Easy information
6:15
to, there's publicly available for some plans,
6:17
but it's not very easy to connect all
6:19
the dots to determine
6:21
something like that. I
6:23
think general plans,
6:25
regardless of company. And
6:30
then any specific industries or sectors
6:32
you're interested in? No, please.
6:35
Really just wanted to keep it broad.
6:38
Okay. It's browsing good. What
6:40
is it? Browsing to historical rate
6:42
of return for 401k plans in the us.
6:46
I don't imagine that's gonna give me a very
6:48
good result, but let's see what it does. It's
6:50
also just slow. So this happens to me a lot.
6:53
It tried clicking on it and apparently failed.
6:55
It said click failed. Why did the click fail? I
6:57
don't know. Can I even open this link?
7:00
I can, it's a search
7:03
on Bing for lots of stuff.
7:06
Okay. Okay.
7:09
So now it's going off of a pocket
7:12
sense article and
7:15
then it went back to the search.
7:17
This is a lot better than what I've most
7:20
of the, so it was, what I was
7:22
looking in was if you have, if
7:25
you're like a, an officer of a
7:27
company, you are, you
7:29
might be what is called a fiduciary
7:32
of a re you're like responsible for the retirement
7:35
plan. So that was the lens I was trying
7:37
to get a sense for is what
7:39
are the how good would it do? Looking
7:42
out there and getting a sense for what
7:44
the current trending risk
7:47
areas are, for those individuals.
7:50
And then also trying to get it to think like
7:53
of the stuff that's happening out there,
7:55
what's likely to become
7:58
a more meaningful risk in future years
8:01
that like what's trending and just developing.
8:03
And anytime I would run it like that, I
8:05
definitely would not get that many successful
8:09
clickthroughs on
8:10
anything. It would maybe just
8:12
find one or two articles and then most
8:14
of the time it would say click failed retrying.
8:16
Even in that example, you look in it's I don't know, seven,
8:18
eight websites or something. Yeah.
8:21
Yeah. Honestly,
8:23
this is the most successful clicks I think
8:25
I've gotten and it still didn't get a very
8:28
helpful result. Yeah. It's particularly
8:30
spending a bunch of time on telling what a 401k
8:33
is, which is strange, but,
8:35
okay. Even assuming nine
8:37
and a half percent is accurate, who knows? That's
8:39
it says it's coming from a link. It's
8:42
then not well, okay. Yeah.
8:44
It then even says, I ran out of time
8:47
before I could find information about
8:49
the average on
8:51
sep. Oh, okay. But that's not
8:53
what I was asking for. All right. All right, then let's
8:56
just re-ask the same-ish question.
8:59
Because what I want you to do is
9:02
retrieve the rate of return for the top,
9:05
or actually let's even just say list the
9:07
top 10 retirement plans based on rate
9:09
of return, and then
9:12
see if that does a better job
9:15
refining it. It
9:18
is a little strange to me that it was like, yes,
9:21
I will go answer that, and then it didn't stay on track.
9:25
Yeah. I didn't realize there was a time component
9:29
that it, it was on that, maybe allotted only
9:31
a certain amount of time to
9:32
perform the search. I
9:34
am honestly not surprised. You want it to be
9:38
responsive. Why is it loading
9:40
airline credit cards, cashback,
9:43
credit cards. What in the world? Okay,
9:45
now it's failing to read anything, so
9:49
I wonder what would be a better way
9:51
to, okay let's see if we could prompt
9:54
it for a second here. If I just did top
9:56
10 retirement accounts
9:59
and I guess I should be saying 401k.
10:05
All right, so here is an example
10:07
of one who knows how accurate it is, but
10:09
does this then list the,
10:13
okay, so it lists the average annual return. Great.
10:15
So let's just see if I give it
10:19
still on credit cards. Okay
10:21
we'll see what it comes up with. Based on
10:23
the analysis at here,
10:26
pick the top 10. Interesting.
10:33
Still giving me some of the same
10:35
topic content as before.
10:39
Okay. You
10:41
can stop typing at any time.
10:43
ChatGPT. Wow,
10:47
that's a really weird direction to take. All
10:51
right. How about, I'm tell
10:53
you to go here. Curious in if
10:55
anyone else is, from a
10:57
strategy standpoint, found, it,
10:59
let's say it does look at eight websites.
11:02
A prompt that it's
11:04
okay, now look at eight now. Now find new
11:07
information, but factor
11:09
the pri. Like how many times can you have it,
11:11
do a search and collect information
11:13
and organize that information and then apply
11:17
it in some way through a
11:18
prompt somehow. Yeah,
11:49
I've definitely done data
11:51
analysis where I just go get
11:53
the data and then I throw it in. And
11:55
there's a few tools that I've seen that somewhat
11:58
try and work on that. Wow.
12:00
It really doesn't like that US News
12:02
website. I wonder out
12:04
of curiosity if it's not an H T G P thing.
12:07
All right. So it'll work without that. Stop
12:10
typing.
12:12
It might be a financial thing. I've
12:14
noticed sometimes if I've
12:17
tried, I'm trying to build a stock app
12:20
and yeah, there are some
12:22
blocks. Oh,
12:24
that's interesting. I
12:28
am a little surprised,
12:30
but I guess there could
12:32
be a lot of things that, that open AI
12:34
is saying, Nope, let's not go there. Okay.
12:37
Yeah, it's not htt
12:39
p s that's weird.
12:43
I think it looks like a good search
12:45
because you're asking historical, so
12:47
it should be able to pull that information up. It
12:50
d it doesn't seem like it would be proprietary
12:52
in any way. If you were, you're asking it to do a stock
12:55
picking for future, it
12:57
won't do that. But if you were to do it based
12:59
on history of 401ks, what's
13:02
the average, the best return rate? I would think
13:04
that would be a no-brainer.
13:06
Yeah. I have yet to have any good experiences
13:08
with Bard, but let's just go try that. See
13:11
if it does anything useful. Okay.
13:15
That at least got. That,
13:18
that got something, but I'm not
13:21
sure. Hang on. Okay.
13:25
It did correctly
13:27
pull a bunch of these, it looks like. Nice.
13:30
Okay. Not so much. So
13:32
VTax is 19.1, but
13:34
it doesn't list that. It immediately then goes
13:37
on to Baron Partners,
13:39
which is 18.7, which is the next
13:41
one. But clearly it has just failed at the task
13:44
and actually skips the next one. Two,
13:47
what? That's
13:50
weird. Okay, hang on. Why is
13:52
it coming up with artisan
13:54
midcap, which is 18.4
13:58
and is not mentioned in this article? Okay.
14:01
All right. So yeah this is probably
14:03
just not using that data at all.
14:05
It's pulling it from somewhere else. I
14:09
wonder, does it say no?
14:14
Yeah, so it's not even, it's not
14:16
even loading it. Use the
14:19
article at, to
14:23
pick, like
14:26
I said, I have not had good luck with Bard,
14:28
so I'm not sure. Yeah. Okay. That's
14:30
about what I was expecting from Bard. Okay.
14:35
I wonder if we
14:37
take it the other direction. So
14:39
if we say retrieve
14:44
the best performing
14:46
funds from
14:48
companies, fidelity,
14:53
Vanguard. Baron.
14:56
I assume that's Baron Partners, but I don't actually
14:59
know. Is it gonna actually
15:01
come up with anything useful or is it just gonna
15:04
Okay, it's browsing the web. That's a good sign. Oops.
15:10
Okay, that's a good one to run. Still
15:14
coming to a Forbes
15:18
and then Best Student credit cards. It really likes
15:20
looking at credit cards. Apparently. I don't understand
15:22
this. This is weird. That's really interesting. I
15:26
almost wanna try and do something else like financial,
15:29
but not related to this, just to see if it decides
15:31
to be like, yeah, I'm gonna talk about home loans,
15:34
credit cards again or something. But we
15:36
will wait on that. Oh my God.
15:39
Bunch of bit credit card stuff and then it failed.
15:45
Huh? Let's
15:47
actually throw in a step by
15:50
step here, cuz I'm
15:52
wondering work
15:54
step by step to retrieve
15:59
ooh, this might actually be a good point to try react,
16:01
but let me try this first one,
16:04
retrieve top companies
16:09
two, retrieve five
16:12
funds from each. Three
16:14
retrieve rate of return
16:17
for each fund. Four.
16:19
Compare to select
16:22
top funds by rate
16:24
of return. And let's actually
16:26
just give it to it as a bulleted list
16:29
instead of an in line. All
16:32
right. See it.
16:36
One more try. Wow.
16:39
No, it does not like the, something
16:41
about the way I provided that list is making
16:44
it really unhappy. Let me just refresh the browser.
16:48
Wow. I think it actually just lost that entire
16:51
pro. It did. It lost an entire prompt. Okay,
16:54
let's lemme see if I can recreate that really
16:56
quick. That was this one.
16:59
And then work step
17:02
by step. Retrieve
17:05
tall companies retrieve five of their funds,
17:07
retrieve the rate
17:09
of return for each fund,
17:12
and then provide a
17:15
table sorted by rate
17:17
of return with
17:19
company, with columns company
17:23
fund, rate of
17:25
return. And I think that
17:28
that'd be good cuz that, that's
17:30
something, that's a good example where
17:32
it,
17:33
Browsing. Yeah. Okay. Run.
17:40
Okay. This time it's at least executing.
17:44
And it got the best 401K plans.
17:47
That's something.
17:50
This is good. Cause this is definitely in line with
17:52
it, you gotta look at different areas
17:55
and bring all the information together
17:56
and Yeah,
17:59
and I'm definitely thinking that
18:01
comeback screen share it
18:03
might even be helpful to use
18:06
the react. Okay.
18:10
That's at least a reasonable place to
18:12
look. Because boy, this is
18:15
not getting anywhere fast. I'm
18:18
drawing a blank on this. I've never seen
18:20
it. Just keep dying.
18:23
In lieu of going deeper on that.
18:25
David or Rhonda, do you have anything that you've
18:27
been trying to work on and wondering
18:29
about?
18:32
I've been working on something that
18:36
I put together a kind of a
18:39
slideshow showing
18:41
what I've done and
18:45
if you want, and what better
18:47
it is I thought
18:49
I was reading this book about
18:52
mysteries where
18:56
you're given clues in the mystery.
18:59
And as
19:01
a reader, you're able to figure
19:04
out the mystery on your
19:06
own. But you're
19:08
given the answer at the end. And
19:11
I thought, I wonder if t
19:14
can do something like
19:16
this.
19:17
That's an interesting idea. The reason why I like
19:19
this I'm teacher and I
19:21
would like students to be able
19:24
to try to figure out their
19:26
mystery on
19:28
their own and have this book, it's only
19:30
two pages. Each
19:32
mystery is only like two pages long.
19:35
But anyway it would be clues like
19:37
a guy comes back from the desert
19:41
and he claims that he he found gold
19:44
and that and he this is, and
19:48
he crazy found gold and he came back.
19:50
He He says he was out in the sun
19:52
and he's all canned and everything,
19:56
but he mentions that he just shaved
19:58
the day before. And then the answer
20:01
to the story as well, if
20:03
he'd been out in the desert, if
20:05
he had not shaved, then
20:08
there wouldn't be a sunburn because
20:10
once the, once you
20:12
shave off
20:15
your beard, you wouldn't, you,
20:19
I don't think it would mix, but, so
20:22
it's like there's a clue in the story. You
20:24
can figure it out on your own, kinda. So
20:27
I don't know how you are with time, but I did a quick,
20:30
if it's okay, I did a quick kind
20:32
of summary of the different attempts
20:34
I've made so far to
20:37
try to get ChatGPT to
20:39
do it. And what's interesting
20:41
is to show you. How
20:43
it's failed so far,
20:46
we might have lost you. Yeah,
20:49
we lost him. Okay. Rhonda,
20:51
it sounded like you had something in mind, so go
20:53
for it.
20:55
I just wanted to basically tell you why I'm here
20:57
and then also you might be able to help
20:59
solve some of the problems that I'm having. Building
21:02
a cup and handle financial model
21:05
to attach to, I don't know if you've heard of I b D.
21:07
It's international Business Daily. It's
21:09
a, it's an investor tool
21:11
that people use and has the 20 factors for
21:14
how do you. Check a stock and
21:16
how do you not, you still have to do a lot
21:18
of research. I want my tool
21:20
to build out that final research of a
21:22
cup and a handle, which is a pattern
21:25
that the stock takes. And when it gets
21:27
to the handle, that's when you know it's
21:29
gonna shoot up. You still have to check other
21:31
financials, but I wanna build an app using
21:34
AI that builds on
21:37
these apps that are already out there. I have shaken
21:39
analytics, I have optometry.
21:41
I bought all these apps that have lifetime, I spent
21:44
$5,000 in each and they're worthless because
21:46
I don't have the time just
21:48
in research cuz I have a million projects I wanna do.
21:51
And I'm coming here to you because I
21:53
like the perspective and I like to learn
21:55
about new
21:55
tools. Gotcha. You could
21:57
try dumping some
21:59
of that information in and then, asking
22:01
ChatGPT to analyze it. I don't know how
22:04
I, not knowing anything about the cup
22:06
and whatever the name was, model.
22:08
I have no idea how much data that
22:10
is, if it's reasonable to paste it in or
22:12
not, but Right. It sounds
22:15
like it's something where you could analyze one stock at
22:17
a time. Is that right? You don't need to analyze like 12
22:19
compared to each other? No,
22:21
that's exactly right. And what I'm thinking
22:23
is I would begin my own
22:25
investor club and everybody would join
22:28
and then everybody would take their
22:30
10 favorite stocks, if they could be mid-cap,
22:32
whatever, and then they run my program and
22:35
whoever sees a cup and handle forming would
22:37
then put it into the notes
22:39
for our group and we'd
22:42
all be able to look at it each week or each
22:44
month that we meet. We would get better and better at
22:46
identifying cup and cup and handle and
22:48
hopefully coldly tool enough
22:51
that it would be identified on its
22:53
own.
22:55
Yeah. Thinking about
22:59
the way ChatGPT works, it's not great
23:01
at graphical things, and I don't just mean like analyzing
23:03
an image, graphical reasoning, which is what it sounds
23:06
like you're saying of I'm assuming there's a
23:08
curve this way and a curve another way, and it
23:10
needs to detect that. I was
23:12
thinking though that it could actually even put it into
23:15
an analyzer. Like you could use
23:17
stable diffusion or mid journey and run
23:20
a describe on it and basically just ask,
23:22
is this describing an upward curve or
23:24
a downward curve or, I, I don't know the right
23:26
way to describe what you're talking about, but
23:29
there are some ways you maybe could make that work.
23:31
Okay. So
23:34
I love that.
23:36
You can run algorithms like that, that
23:38
are not really in ChatGPT, but
23:40
because you're talking about like technical analysis
23:43
with the formation of the cup and the handle.
23:45
That might be something to look at where,
23:47
cuz there are a lot of different, even larger
23:50
scale platforms with major
23:52
financial institutions that allow you to build out
23:54
models that detect trading
23:56
patterns and things like that based on whatever your
23:59
your criteria are.
24:00
For the, so that's an open AI seek
24:02
and run.
24:03
It's not an open ai. Is it
24:05
open? Open source?
24:06
Open? Yeah. It, I mean it should be. It should
24:08
be. Yeah. Cuz there's lots, there, there are ways
24:11
to people build, it's
24:13
not exactly charting, but they build trading
24:15
models to respond to tweets and things like that. And
24:18
you can also do visual setups
24:20
where it's double bottoming out
24:22
or hitting a peak or something like that, and
24:24
then, execute a trade or send an alert or something.
24:26
So that might be a route to look at.
24:29
Love it. Okay. Thank
24:31
you.
24:33
Yeah, I'm not familiar with the, any of those, so
24:36
I can't give a suggestion of try
24:38
X, Y, or Z, but I have heard
24:40
those exist as well. Okay.
24:45
All right. I just got the file from
24:47
David.
24:49
It's a little different because you're not saying
24:51
write a mystery story. It could do that very
24:53
easily and it writes your typical,
24:56
a bunch of rich people with British themes
24:58
in a manner all together.
25:01
And, funny, it does it really
25:03
well, it just, And it comes
25:05
up with somebody screams
25:08
and then they all end up in the drawing room
25:10
and he talks to three suspects. And, but
25:13
then what happens is it's
25:15
always here's, oh, here's
25:19
why the person's guilty, but you're not
25:21
involved in trying to figure out why.
25:23
So if you want, I can,
25:27
I have highlighted stuff, but I can explain
25:30
the process I've been through so far. It's been
25:32
interesting.
25:34
Got it. So you're trying to get it.
25:36
Oh, I see. Yeah. So
25:40
it's just not wow. Blackwood
25:42
manner. You're, it's really leaning into the Sherlock
25:45
Homes experience and clue for that matter.
25:47
Yeah. It's
25:49
funny, it'll, it can't put
25:53
clues in the story and then
25:56
Put them in there subtly and
26:00
then hold off and let the reader figure
26:02
it out and then tell you at the end.
26:05
And it, so
26:08
I tried telling it to structure the story.
26:10
First I want an introduction,
26:13
then I want there to be
26:15
a a crime. Then I want
26:17
the interviews, and then I want
26:20
the, what, the reveal at the end. And,
26:23
but make it so that, the
26:28
reader could use critical thinking and
26:30
yeah. Okay. And
26:33
it,
26:33
so it can tell you, it can tell
26:35
you this is what you should do, this is
26:37
how to do it. And it, cuz
26:39
that's what I did. I asked it to, what are the elements
26:42
of a story like that? And it writes 'em all
26:44
down. It tells you everything you're supposed to do
26:46
and then you say, okay, can you do that please?
26:49
And I can, I can't do it.
26:51
That's funny. That's so interesting. Okay.
26:54
Let's see. So
26:58
let me see if ChatGPT four is
27:00
actually working again. Hey,
27:02
it's working again. All right, let me change the screen
27:05
share here. But I get what you're saying
27:07
about Yeah. It just keeps,
27:10
that's hilarious. All right, so
27:13
share this one. Lost
27:18
the video again. There we go. Okay.
27:21
Yay. Finally. Yeah. So what I
27:23
ended
27:23
up doing was asking it just to give
27:25
clues, like just to give me ideas
27:28
for clues and for crimes that, like
27:30
a crime, a clue and a contradiction
27:32
or something, and then put
27:35
it in yourself kind of thing. But
27:39
I was hoping it could do more
27:40
than that. Yeah.
27:42
So let's see. Outline short mystery
27:44
story include the
27:47
plot twists as well
27:49
as major plot
27:51
points for each chapter
27:55
and hints towards the
27:58
plot twists subtle
28:01
hints towards the plot twists.
28:04
Let's not even include any length.
28:07
We'll just go with that. I'm
28:09
not even sure it's gonna follow my
28:11
outline quite near enough. I think I need to give
28:13
it a shot prompt or two, but let's
28:16
see.
28:18
Yeah, the only problem with this kind of problem is there's
28:20
lots of reading.
28:22
Yeah. But I am thinking
28:24
it'll probably be something like every
28:27
chapter you generate by saying, here's
28:29
the outline now generate for chapter
28:31
one and
28:34
then generate chapter two and that kind of thing. Okay.
28:40
So there's a pi, a
28:42
philanthropist, a business tycoon, and a brilliant
28:44
but unassuming scientist. Interesting
28:47
though it didn't actually
28:50
include anything about what the
28:52
mystery is. We've
28:54
got a. Little, it's
28:57
doing the plot twist hint. That's good. So
29:01
I almost think I need to just restart this
29:03
and let's just make it a murder
29:05
mystery, just to make that easier.
29:08
Include the
29:11
actual crime or the crime,
29:14
the plot twist, as well as major points. Stop
29:17
generating for each chapter and subtle hints
29:19
towards the plot twist. Oh,
29:21
okay. Then for each chapter,
29:27
major plot points, one
29:31
plus subtle hint towards the plot twist. Go.
29:33
Let's see if that works better. Suddenly
29:38
it's very fixated on boats. Got
29:40
two boat stories in a row. Okay.
29:44
So it's an elaborate suicide, but we don't actually
29:46
know that. Oh.
29:49
Okay. It's doing it. So the major
29:51
plot point is Martel's body is
29:53
discovered. Detective is called
29:55
to the scene, finds a new
29:58
Van Gogh painting. Subtle hint is
30:00
looking at the last one. Testament, align states
30:03
The art world will find my last piece most breathtaking.
30:06
How does that connect to the Oh,
30:09
cuz. Cuz the plot twist is, he
30:11
faked a murder
30:14
actually killed himself cuz he wants
30:16
to bring attention to a counterfeit van Gogh. That's
30:18
a really weird plot twist, but Okay, sure. Whatever.
30:20
We'll go with the ChatGPT like it. Okay.
30:24
So if we then
30:27
run this as a second prompt,
30:30
which it's still going. So anytime
30:32
you want to finish, that would be great. There
30:36
we go. All right, so now
30:38
actually for clarity, so this
30:41
is the outline generator,
30:43
and now we're gonna make another one that's going to be
30:45
the chapter generator, but I don't need to name
30:48
it that yet. And write
30:50
the, This
30:52
chapter for a murder
30:55
mystery. Oh,
30:58
actually I need to go back and grab
31:00
the first prompt there.
31:03
It's okay. New chat. Here's
31:08
the thing, here's the second thing.
31:10
Okay. So acts as a murder, Mr. Mess.
31:13
Best many bestsellers. Write
31:16
this chapter. Of
31:19
a murder mystery Short story using
31:22
the outline,
31:25
crime, plot twist
31:28
major points for
31:31
chapter, and then subtle
31:34
hints toward the plot twist outline.
31:40
Actually, let's do this way just
31:42
for clarity. Please
31:45
write this chapter in a fluoride entertaining
31:48
way. I don't know. I
31:52
did do, oh man. I'm running this
31:54
on three five. All right.
31:56
I'll run it on four again in a second. At
31:58
least three five's faster. Okay.
32:01
One, one turn of phrase I like. Her
32:04
gaze was immediately drawn to the unfamiliar
32:06
Van Gogh painting laying nearby. The strokes
32:08
of the master artist seem to whisper secrets
32:11
into her ear. Yeah.
32:13
Okay. It's definitely going for fluoride. All
32:15
right, let's again,
32:19
and run this on four. All
32:22
right, so the main thing that I'd
32:24
be concerned about with running it this way
32:26
is then it's going to start
32:29
drifting away from the details.
32:32
So what I would then do is say, for
32:35
chapter one, summarize this,
32:37
and for chapter two, give it the outline
32:40
and then the summary of chapter one. I
32:44
see. To keep it on
32:46
task. Yeah. So it doesn't
32:48
suddenly add a new character in chapter
32:50
three and pretend that they've been on the boat
32:52
the whole time or whatever. It
32:54
can try and keep track of who's there, what
32:57
are they doing, that kind of stuff. Yeah.
33:00
G P T four is much slower. No,
33:02
I'm not able to keep up as fast. But is
33:05
it writing in a style
33:07
that the reader can figure
33:11
out
33:13
in? Seems like it. Guilty person. Okay.
33:16
Yeah. Wow. So there's the
33:18
hint right there. Wait,
33:22
most breathtaking, that's not what it oh,
33:25
no, it was Okay. Nevermind. That
33:27
was the hint. Little
33:31
did Morro know then that the final
33:33
chapters were to be penned by the deceased himself.
33:35
Okay. Yeah. We're definitely getting the tone.
33:38
All right, so summarize
33:41
that chapter into a single
33:44
paragraph. All
33:47
right, let's, that's
33:49
closer, but let me try that again differently. Summarize
33:52
that chapter into a single paragraph to
33:55
containing all the pertinent
33:58
details to carry on
34:01
into chapter two. Let's
34:05
try that. That's
34:08
better. Okay. It's
34:10
including the important bits. It's got the
34:12
quote. Yeah. No surprise there. All
34:15
right, so now the question is, If
34:19
I go back to,
34:23
no, I think it was this one. Oh
34:27
no, that's the three five version. Okay. I
34:31
thought I named it generator.
34:35
Maybe I'm wrong. Come
34:38
on. There we go. Okay, so
34:40
now we'll take
34:42
the same title
34:44
and the chapter
34:47
two info and
34:54
include the same stuff, but
34:58
replace chapter one with chapter
35:00
two. And
35:03
obviously don't need that still, or
35:05
that. There's chapter. There
35:08
we go. Chapter two. Okay. Major
35:10
plot points, blah, blah. Okay,
35:13
outline. Now
35:15
I also need to include using
35:18
the outline and
35:20
the summary of the
35:23
previous chapters. So
35:25
now, and
35:29
where did I put that? There
35:33
we go. Not
35:36
sure this is gonna work, but we can at least try.
35:40
Well,
35:42
I'm learning a lot about how you're. Out
35:44
writing it in for
35:46
it so that the format Yeah.
35:49
Yeah. The like including
35:52
them as H T M L tags instead
35:54
of using variables, more common style.
35:57
I'm mostly just doing, I've seen a
35:59
lot of different ways of doing it. I don't think any
36:01
of them work better than any others, with
36:03
the exception of this one. You're not
36:05
gonna run into problems with quotes cuz
36:08
you know, if you do a normal like var
36:11
chapter two summary equals quote,
36:13
then like, all of the quotes from people
36:15
speaking are gonna be messed up. So
36:18
it, it seems to do a pretty good job of working
36:20
around that. Wow.
36:23
It's immediately jumping to,
36:27
huh? Oh, okay. So
36:30
suggesting its counterfeit is already in this chapter.
36:32
I missed that and
36:34
is, Yep.
36:37
So they're bringing in the new character of Claude
36:40
describing him. Oh,
36:44
I just noticed we're a couple minutes over time. But
36:47
this seems like it's actually getting a decent
36:49
story. I'm curious enough that I'll probably just
36:52
run this a little bit longer and then actually read
36:54
some of it. Cause I'm curious to see what it comes
36:56
up with. It seems very
36:58
I'm having Eric, but I'm
37:02
having
37:02
trouble. Keep keeping up. You're really fast
37:04
and good at this, but do you think
37:06
it's going to make it
37:09
so that when
37:12
you get to the end of the, at the end before
37:15
you. You're told who's
37:17
guilty, you'll be able to use
37:20
the clues in the story to figure out yourself
37:22
if if you're able who
37:24
did it.
37:27
I think it's possible. I think honestly
37:29
we'd need to generate all, yeah. What
37:32
was it? Seven. Oh,
37:34
okay. There's more than seven chapters apparently, but
37:36
at least the seven chapters and kinda get a sense
37:39
of it. Oh
37:41
seven is the conclusion. So then there's
37:43
an epilogue, so yeah.
37:45
Yeah. Won't take too much longer
37:47
to run this process.
37:50
It's a long process cuz you'd have to,
37:53
go through the story yourself and
37:55
see whether or not you can figure it out
37:57
before you read the
37:58
last chapter. And then,
38:01
and
38:01
then find out, one
38:03
of the frustrating things was that it kept
38:06
revealing new information
38:08
at the end that you had no way of knowing.
38:12
Kinda like the way they do that on tv. But anyway
38:14
and then you're yeah,
38:17
cuz my whole goal is that the students,
38:20
enjoy trying to figure
38:22
it out on their own before finding
38:24
out the answer.
38:27
Yeah. Even just reading some of these
38:30
subtle hints, it's not part
38:33
of the idea is that it's looks
38:36
like a murder, but it's actually a
38:38
suicide. I'm
38:40
not seeing stuff that would really
38:42
super call that out. There's
38:44
one down here.
38:47
Yeah. No, wasn't that one?
38:49
Painting is a new thing for him. There it is. He
38:52
apparently was using a brand of paint
38:54
that is toxic. I'm
38:57
not sure that's enough to really help
39:00
you realize. It is a suicide. So I
39:03
guess in some ways I'd say that the whole outline
39:05
is a little bit flawed. This
39:08
is much
39:08
more sophisticated than what I was able to
39:10
get in terms of
39:12
subtlety. Cool.
39:15
So I'll share the prompts
39:18
that I'm using. And actually I think open,
39:20
I added a share. Yep, it did. Can
39:23
you I'll share these as
39:25
well. And
39:26
can you, is you're sharing the prompts and that
39:28
is the showing the syntax, the prompt syntax?
39:30
Okay. All right. That's,
39:33
is that we're
39:34
doing it in the chat. How
39:35
would I, sorry, it's in the chat.
39:38
So that's the outline. And
39:41
that, come
39:45
on. There we go. In fact,
39:47
actually, let me rename these just so I can keep
39:49
them straight too. Outline
39:51
for mystery and then
39:56
generate chapter one and two.
40:00
All right. I
40:03
just put the first one in and the second one
40:05
should be one at
40:07
two.
40:09
Greg, I gotta drop off, but I really appreciate this.
40:12
This was really great to watch you in your element,
40:14
go through and watch your approach and style
40:16
and everything and stuff, and that was really fantastic.
40:18
Thank you.
40:19
It is really great.
40:21
Well, thanks for taking the time to do that. I
40:23
really appreciate it. I hope the you're
40:25
welcome. By the way, if
40:27
I can ask you, do you have, is there any
40:29
book that you recommend reading about
40:32
prompt engineering or
40:34
is it No, I haven't found any
40:36
books yet. I'm actually coming out with
40:38
a course probably next week or
40:40
so. I've been planning this week, but this has
40:42
gotten away from me. Yeah. Thank you all for
40:44
coming.
40:46
Yeah, thank you so much. Really appreciate
40:48
your approach so that people can see
40:50
in real time how things
40:52
progress. That's really great. Yeah. Thank
40:54
you. Thank you so much.
40:56
Thank you for your time. You're welcome. Amazing. Thank
40:58
you. Thank you. Good rest of your week. All
41:00
right. Okay. Bye-Bye. Bye. Okay.
41:03
Bye-bye.
Podchaser is the ultimate destination for podcast data, search, and discovery. Learn More