Dave Thomas: Programming, Publishing & Embracing Change

WEBVTT
00:00:09.390 --> 00:00:17.480
Welcome to Roundabout Creative Chaos, a podcast
00:00:17.480 --> 00:00:19.559
about people who make things and the stories
00:00:19.559 --> 00:00:22.640
behind why they do it. I'm your host, Tammy Coran,
00:00:22.739 --> 00:00:26.559
and this is episode 155, recorded on August 25th,
00:00:26.579 --> 00:00:30.379
2026. Joining me tonight is my co -host, Felipe
00:00:30.379 --> 00:00:34.090
Lasso, and our guest, Dave Thomas. Dave is a
00:00:34.090 --> 00:00:36.630
programmer, author, publisher, and co -author
00:00:36.630 --> 00:00:39.590
of The Pragmatic Programmer. He's also one of
00:00:39.590 --> 00:00:42.329
the original authors of The Agile Manifesto and
00:00:42.329 --> 00:00:45.990
co -founder of The Pragmatic Bookshelf. On this
00:00:45.990 --> 00:00:48.170
episode, we talk about how software development
00:00:48.170 --> 00:00:51.909
has changed, the growing role of AI, and what
00:00:51.909 --> 00:00:53.869
programmers still need to know when machines
00:00:53.869 --> 00:00:57.109
can write the code. We also get into how programmers
00:00:57.109 --> 00:01:00.009
learn, the value of practice, and the uncertain
00:01:00.009 --> 00:01:03.149
future of technical publishing. Are you ready
00:01:03.149 --> 00:01:06.549
to talk tech, music, arts, and zombies? I know
00:01:06.549 --> 00:01:09.870
we are. So grab your crossbow and let's get this
00:01:09.870 --> 00:01:25.739
episode started. Philippe, I don't think we've
00:01:25.739 --> 00:01:29.040
met. We have not, but we have now. We have now
00:01:29.040 --> 00:01:32.140
officially met. I am Philippe, and you are the
00:01:32.140 --> 00:01:35.379
Dave Thomas, and we have met. Welcome, welcome.
00:01:36.219 --> 00:01:39.060
Excellent. Well, thank you for having me. I knew
00:01:39.060 --> 00:01:40.939
I wanted to get you on this show. I just had
00:01:40.939 --> 00:01:42.739
to figure out when I could make it happen, and
00:01:42.739 --> 00:01:46.510
then I had to... hope to the networking gods
00:01:46.510 --> 00:01:49.010
that you would say, yes, and you did. So now
00:01:49.010 --> 00:01:51.349
you're here. Well, of course I will. Of course
00:01:51.349 --> 00:01:56.930
I would. So, excellent. What got you into development?
00:01:57.250 --> 00:01:59.730
I mean, did you grow up with computers? I was
00:01:59.730 --> 00:02:01.430
in school in England, in secondary school in
00:02:01.430 --> 00:02:06.140
England. And in England, you have two bands of
00:02:06.140 --> 00:02:07.900
tests that you take. You take tests when you're
00:02:07.900 --> 00:02:13.020
15 and you take tests when you're 18. And the
00:02:13.020 --> 00:02:15.300
reason for that is that back in the old days,
00:02:15.400 --> 00:02:18.180
and probably to some extent now, a lot of people
00:02:18.180 --> 00:02:20.780
who were going into the trades would take tests
00:02:20.780 --> 00:02:23.020
at 15 and then enter apprenticeship programs.
00:02:23.300 --> 00:02:26.759
And then the tests at 18 were intended to be
00:02:26.759 --> 00:02:31.469
entrance tests for universities. So I was in
00:02:31.469 --> 00:02:36.789
the 18 tests, but you can take them early if
00:02:36.789 --> 00:02:38.370
the school lets you. So you can take them a year
00:02:38.370 --> 00:02:43.349
early and then do more tests. I had, in my second
00:02:43.379 --> 00:02:47.680
to highest year, I had finished all of the tests
00:02:47.680 --> 00:02:49.840
that I was going to take that year. And so there
00:02:49.840 --> 00:02:52.000
was no more lessons to learn. But I still had
00:02:52.000 --> 00:02:54.960
a month of school left. And no, they would not
00:02:54.960 --> 00:02:59.550
let me go home. So to keep us busy, they sent
00:02:59.550 --> 00:03:01.530
us across the road. There was a technical college
00:03:01.530 --> 00:03:04.550
literally across the road. And they were teaching
00:03:04.550 --> 00:03:08.349
the very first course on computer programming
00:03:08.349 --> 00:03:11.689
for that college. So they sent us over there
00:03:11.689 --> 00:03:17.370
just basically to keep us busy. And I got in
00:03:17.370 --> 00:03:22.800
there and it was teaching us basic. And we were
00:03:22.800 --> 00:03:28.000
using the old ASR -33 teletypes, which are this
00:03:28.000 --> 00:03:30.060
typewriter thing that has a big roll of paper.
00:03:30.539 --> 00:03:32.759
And you type and it goes across the paper and
00:03:32.759 --> 00:03:35.120
the paper scrolls off. And at the same time,
00:03:35.139 --> 00:03:39.650
it punches what you type out on paper tape. And
00:03:39.650 --> 00:03:41.879
then... You get that paper tape and you feed
00:03:41.879 --> 00:03:43.900
it back into the machine while you're online
00:03:43.900 --> 00:03:46.360
to a computer and it sends your program up to
00:03:46.360 --> 00:03:48.699
the computer, right? So it sounds medieval and
00:03:48.699 --> 00:03:54.620
it was medieval. And I did that and I totally
00:03:54.620 --> 00:03:58.240
fell in love. I was going to go do math or physics
00:03:58.240 --> 00:04:03.400
at college and I got home and I said to my parents,
00:04:03.520 --> 00:04:05.139
no, I'm not going to do that. I'm going to do
00:04:05.139 --> 00:04:08.669
computer stuff. uh because i just like i said
00:04:08.669 --> 00:04:10.490
i just fell in love with and i don't even know
00:04:10.490 --> 00:04:15.389
i think if you if the reality is that everybody
00:04:15.460 --> 00:04:21.980
is looking for ways to express themselves and
00:04:21.980 --> 00:04:25.199
that sounds kind of corny but you know if you're
00:04:25.199 --> 00:04:28.319
an artist or a dancer or whatever else you have
00:04:28.319 --> 00:04:31.519
obvious outlets for that but for most people
00:04:31.519 --> 00:04:38.759
it's kind of hard uh to find a way of reifying
00:04:38.759 --> 00:04:43.470
what's inside you, making it real. And for me,
00:04:43.529 --> 00:04:46.269
software was my way of expressing myself. And
00:04:46.269 --> 00:04:49.589
I found that I could think of something and then
00:04:49.589 --> 00:04:53.189
express it in code and then show it to other
00:04:53.189 --> 00:04:58.209
people. So it really was a way of breaking down
00:04:58.209 --> 00:05:01.279
that kind of barrier I otherwise had. And I think
00:05:01.279 --> 00:05:04.220
that's why I got into it in the first place.
00:05:04.610 --> 00:05:06.790
Do you remember your first job that you had with
00:05:06.790 --> 00:05:13.540
computers? Well, yeah. I actually got a summer
00:05:13.540 --> 00:05:22.399
job working with a local water authority. And
00:05:22.399 --> 00:05:25.500
in retrospect, they must have had a grant or
00:05:25.500 --> 00:05:28.040
something because it was total make work. There
00:05:28.040 --> 00:05:33.439
was absolutely no benefit to it at all. But they
00:05:33.439 --> 00:05:38.459
had us writing. a program that would predict
00:05:38.459 --> 00:05:44.500
rainfall amounts based on historical data. And
00:05:44.500 --> 00:05:48.600
we were writing this on a really, even at the
00:05:48.600 --> 00:05:52.660
time, a really sad little computer. We were using
00:05:52.660 --> 00:05:57.189
punch cards and it would take, I don't know,
00:05:57.209 --> 00:06:00.730
half an hour to go from a change to seeing if
00:06:00.730 --> 00:06:03.829
it worked or not. And worked on that all summer.
00:06:04.620 --> 00:06:07.279
I have no idea if we even showed the program
00:06:07.279 --> 00:06:09.620
to anybody at the end, but we had a really good
00:06:09.620 --> 00:06:15.930
time writing it. And yeah, from there I went
00:06:15.930 --> 00:06:20.209
to college and really just never looked back.
00:06:20.790 --> 00:06:24.649
So how do you go from that to writing something
00:06:24.649 --> 00:06:29.069
like the Ruby book? What do they call that? The
00:06:29.069 --> 00:06:33.120
Pickaxe book? The Pickaxe book, yeah. That seems
00:06:33.120 --> 00:06:37.670
like a big jump. honestly i was going to ask
00:06:37.670 --> 00:06:40.790
about that like the transition because the the
00:06:40.790 --> 00:06:42.589
software industry has like these these shifts
00:06:42.589 --> 00:06:45.250
right the web stuff comes out and it's like the
00:06:45.250 --> 00:06:47.949
web you have to have a website and blogs and
00:06:47.949 --> 00:06:50.540
the advent of wordpress i mean people made A
00:06:50.540 --> 00:06:52.560
bank with WordPress and being able to sell blogs
00:06:52.560 --> 00:06:54.540
and e -commerce and things like that. Now there's
00:06:54.540 --> 00:06:57.279
like some folks moving to Web 3 .0 and the crypto
00:06:57.279 --> 00:07:00.180
stuff and blockchain. We have the SaaS moving,
00:07:00.259 --> 00:07:02.240
like everything has to be in the cloud as a subscription,
00:07:02.379 --> 00:07:04.600
as a service kind of deal. Now I think everything
00:07:04.600 --> 00:07:06.939
has to be AI, right? Like AI this, we're connected
00:07:06.939 --> 00:07:08.899
to AI, screen recording and all these things.
00:07:09.120 --> 00:07:12.420
So you've been through many, many shifts. To
00:07:12.420 --> 00:07:14.819
Tammy's point, how do we go from punch cards
00:07:14.819 --> 00:07:17.579
to where we're now? Because it's a constant learning,
00:07:17.720 --> 00:07:20.339
right? Something that I often talk about is like
00:07:20.339 --> 00:07:22.680
a doctor always has to be learning, looking at
00:07:22.680 --> 00:07:26.660
research and stuff. But the core aspect of medicine
00:07:26.660 --> 00:07:28.660
doesn't change, which is the fact that humans
00:07:28.660 --> 00:07:31.259
are not evolving at the same pace as technology
00:07:31.259 --> 00:07:33.740
does. So it's not like the hearts of people in
00:07:33.740 --> 00:07:36.939
the 50s and 60s had like different valves and
00:07:36.939 --> 00:07:40.339
veins and chambers. And now like humans have
00:07:40.339 --> 00:07:42.529
20 chambers in there. hearts and so doctors have
00:07:42.529 --> 00:07:45.430
to keep up with that but we have right and so
00:07:45.430 --> 00:07:49.199
to tanya's point i like that question and just
00:07:49.199 --> 00:07:50.879
a follow -up that i have would have afterwards
00:07:50.879 --> 00:07:54.160
is where do you see someone jumping in now right
00:07:54.160 --> 00:07:56.279
they maybe have some of the great books and resources
00:07:56.279 --> 00:07:58.839
that you have and i think more opinion pieces
00:07:58.839 --> 00:08:01.079
or experience -based stuff more than like learn
00:08:01.079 --> 00:08:03.860
ruby from start to finish is great but what would
00:08:03.860 --> 00:08:05.639
a student or a person who wants to start coding
00:08:05.639 --> 00:08:08.540
should even do should they even spend time learning
00:08:08.540 --> 00:08:12.500
the fundamentals algorithms languages what language
00:08:12.500 --> 00:08:16.000
just prompting okay so coming back to Tammy's
00:08:16.000 --> 00:08:18.519
question how did I go from programming to writing
00:08:18.519 --> 00:08:22.660
a book and the answers were by accident totally
00:08:22.980 --> 00:08:28.199
and in the case of the Ruby book because I felt
00:08:28.199 --> 00:08:31.920
it was a story that needed to be told it was
00:08:31.920 --> 00:08:34.299
one of those times where I thought the language
00:08:34.299 --> 00:08:38.529
itself was beautiful and significant And it was
00:08:38.529 --> 00:08:42.269
totally unknown outside Japan because all the
00:08:42.269 --> 00:08:47.629
documentation was in Japanese. So that became
00:08:47.629 --> 00:08:51.190
kind of like a mission book, if you like, to
00:08:51.190 --> 00:08:58.080
do that. In terms of things changing, they do
00:08:58.080 --> 00:09:02.179
and they don't. I have a theory that everything
00:09:02.179 --> 00:09:07.049
in software runs on... varies depending on how
00:09:07.049 --> 00:09:10.169
I look at it, but maybe seven -year cycles where
00:09:10.169 --> 00:09:12.409
we oscillate and we go back and we do it again
00:09:12.409 --> 00:09:16.659
and again and again. And because software developers
00:09:16.659 --> 00:09:20.740
are incredibly bad at knowing their own past
00:09:20.740 --> 00:09:23.720
and knowing the past of development, they are
00:09:23.720 --> 00:09:26.480
very happy to make the same mistakes over and
00:09:26.480 --> 00:09:29.549
over and over again. So yeah, the technology
00:09:29.549 --> 00:09:33.129
has moved on. I'm not convinced that what we
00:09:33.129 --> 00:09:37.600
do with it has changed. so radically that you
00:09:37.600 --> 00:09:43.539
know yes if you stuck a 1960s 70s developer in
00:09:43.539 --> 00:09:47.039
front of visual studio code they would you know
00:09:47.039 --> 00:09:48.600
but they wouldn't have an idea what they're looking
00:09:48.600 --> 00:09:52.389
at uh but if you spent the time to explain to
00:09:52.389 --> 00:09:56.860
them what was going on they would they'd get
00:09:56.860 --> 00:09:59.279
it they would they sometimes wouldn't believe
00:09:59.279 --> 00:10:02.970
it but they would get it um if you look at all
00:10:02.970 --> 00:10:08.929
of the major developments in the field of software
00:10:08.929 --> 00:10:11.570
development, object orientation, functional programming,
00:10:11.710 --> 00:10:13.889
distributed programming, operating systems, programming
00:10:13.889 --> 00:10:18.629
languages, all come out of the 1950s, 1960s,
00:10:18.629 --> 00:10:23.929
and early 1970s. I mean, there's been development,
00:10:24.090 --> 00:10:26.509
but it's been incremental development past then.
00:10:27.409 --> 00:10:32.149
So, yeah, learning the basics and learning old
00:10:32.149 --> 00:10:34.970
school stuff is actually really, really important,
00:10:35.190 --> 00:10:38.879
I think. Now, how do you become a developer nowadays
00:10:38.879 --> 00:10:41.059
when it's so easy just to sit in front of Claude
00:10:41.059 --> 00:10:44.500
and say, write me a program? And I think the
00:10:44.500 --> 00:10:49.490
answer is, okay, if I was given, I don't know
00:10:49.490 --> 00:10:52.169
how much it would cost, $10 million, $20 million
00:10:52.169 --> 00:10:56.830
to create a course for software developers, it
00:10:56.830 --> 00:11:02.210
would be one year apprenticed into a real organization.
00:11:03.269 --> 00:11:06.029
writing code, using AI, and not having the faintest
00:11:06.029 --> 00:11:08.470
idea of what that code does, because that's what
00:11:08.470 --> 00:11:11.710
people do. But you would get an understanding
00:11:11.710 --> 00:11:15.240
of the problems that had had you come back into
00:11:15.240 --> 00:11:19.809
school for a year, and I would teach you not
00:11:20.799 --> 00:11:22.759
all of the fundamentals, because some of them
00:11:22.759 --> 00:11:25.389
are now not relevant, but I would teach you the
00:11:25.389 --> 00:11:29.690
important stuff, and the stuff that you maybe
00:11:29.690 --> 00:11:32.049
were kind of applying during that first year,
00:11:32.090 --> 00:11:35.029
but not knowing you were doing it, or Stuff that
00:11:35.029 --> 00:11:37.090
you should have been applying, but you let the
00:11:37.090 --> 00:11:39.289
AI go in a different direction and it turned
00:11:39.289 --> 00:11:42.529
into a disaster. And so I would motivate the
00:11:42.529 --> 00:11:44.789
first year is I want to get programming. So sure,
00:11:44.830 --> 00:11:46.889
get in there and start programming. Second year
00:11:46.889 --> 00:11:50.009
is I want to start programming correctly. Maybe
00:11:50.009 --> 00:11:52.450
that second year would go to 18 months. I don't
00:11:52.450 --> 00:11:54.370
know. I'd have to sit down and think up. And
00:11:54.370 --> 00:11:58.529
then we'd go back out into industry for six months
00:11:58.529 --> 00:12:02.940
and try to apply what you've learned. And then
00:12:02.940 --> 00:12:05.620
finally come back for like a three -month wrap
00:12:05.620 --> 00:12:09.519
-up in which you kind of like try and prove your
00:12:09.519 --> 00:12:12.500
mastery in the same way that apprentices would
00:12:12.500 --> 00:12:17.000
create a masterpiece to become masters. It would
00:12:17.000 --> 00:12:19.279
be the same with developers. Maybe not quite
00:12:19.279 --> 00:12:22.100
so grand, but you would show how you'd use the
00:12:22.100 --> 00:12:25.960
practical and the theory to do something that
00:12:25.960 --> 00:12:29.600
hadn't been done before. So that's how you should
00:12:29.600 --> 00:12:32.330
be a developer, I think. Given that no one's
00:12:32.330 --> 00:12:35.070
given me the money to go out and do that, then
00:12:35.070 --> 00:12:37.629
I think the alternative is to do that for yourself
00:12:37.629 --> 00:12:40.710
and to say, I want to be a developer and I want
00:12:40.710 --> 00:12:44.450
to be a developer that lasts. Then I understand
00:12:44.450 --> 00:12:46.509
that I'm going to have to teach myself the fundamentals
00:12:46.509 --> 00:12:50.169
at the same time as I'm just butchering code
00:12:50.169 --> 00:12:55.110
using AI. And that takes discipline, but it's
00:12:55.110 --> 00:12:58.559
discipline that pays off. As someone who has
00:12:58.559 --> 00:13:04.509
not only coded, bunch of applications and different
00:13:04.509 --> 00:13:08.110
things and written books about the subject given
00:13:08.110 --> 00:13:11.570
talks about the subject I mean you started a
00:13:11.570 --> 00:13:15.129
publishing company for developers by developers
00:13:15.129 --> 00:13:20.450
which I want to talk about like what made you
00:13:20.450 --> 00:13:24.730
take that leap from being a developer to being
00:13:24.730 --> 00:13:27.690
someone who writes about development someone
00:13:27.690 --> 00:13:31.129
who who has an entire publishing company devoted
00:13:31.129 --> 00:13:35.919
to developers. Remember I said that one of the
00:13:35.919 --> 00:13:39.120
reasons that I got into programming is it helped
00:13:39.120 --> 00:13:48.840
me realize my ideas and make my ideas real. I,
00:13:49.120 --> 00:13:54.539
around about 2000, well, two things happened.
00:13:54.659 --> 00:13:59.889
First of all, Andy Hunt and I, who ended up creating
00:13:59.889 --> 00:14:03.759
Pragmatic, programmers uh or in the mid 90s we
00:14:03.759 --> 00:14:06.000
were consulting and we went out there and we
00:14:06.000 --> 00:14:08.759
talked to companies and if you're a company you
00:14:08.759 --> 00:14:10.539
don't call it a consultant if everything's really
00:14:10.539 --> 00:14:12.399
good you call them in when things are going bad
00:14:12.399 --> 00:14:15.629
and so we would turn up at these companies and
00:14:15.629 --> 00:14:19.450
their software would be a pile of crap and we
00:14:19.450 --> 00:14:22.149
would be able to say things like it would probably
00:14:22.149 --> 00:14:24.870
be a good idea if you tested that before you
00:14:24.870 --> 00:14:29.769
ship it or do you know that when you build the
00:14:29.769 --> 00:14:33.029
product It's different to when Jane builds a
00:14:33.029 --> 00:14:35.289
product. It generates a different piece of code.
00:14:36.450 --> 00:14:41.149
And so all of these kind of like brain -dead
00:14:41.149 --> 00:14:43.070
simple things that people should have been doing,
00:14:43.090 --> 00:14:46.149
they weren't. And for a while, that was nice.
00:14:46.190 --> 00:14:49.210
You could make money. But after a while, it also
00:14:49.210 --> 00:14:52.899
got remarkably frustrating. So we started writing
00:14:52.899 --> 00:14:57.320
down some notes on, you know, before we'd even
00:14:57.320 --> 00:14:59.340
turn up, we'd give them these kind of checklist
00:14:59.340 --> 00:15:01.970
notes of... Maybe you should be doing this, this,
00:15:02.070 --> 00:15:07.210
this, and this. And then we can talk. And my
00:15:07.210 --> 00:15:10.169
wife saw that, Julie saw that, and said, you
00:15:10.169 --> 00:15:11.669
know what, that looks like it could be a book.
00:15:11.769 --> 00:15:13.809
And we said, oh, no, it couldn't. And she said,
00:15:13.870 --> 00:15:17.629
well, why don't you try? So we came up with this
00:15:17.629 --> 00:15:21.960
cunning plan that we would take our notes and
00:15:21.960 --> 00:15:24.139
we would send them to what we viewed as being
00:15:24.139 --> 00:15:27.240
the best technical publisher at the time, which
00:15:27.240 --> 00:15:30.409
was Addison Wesley. And they would reject it.
00:15:30.669 --> 00:15:33.129
But we would get back notes from them to say,
00:15:33.269 --> 00:15:35.629
ah, it was wrong because of this, this, this,
00:15:35.629 --> 00:15:39.519
and this. And then we could improve it. And instead
00:15:39.519 --> 00:15:41.759
of doing that, they totally foiled our plan and
00:15:41.759 --> 00:15:45.799
they accepted it. So that became The Pragmatic
00:15:45.799 --> 00:15:53.389
Programmer. And that was actually quite, it took
00:15:53.389 --> 00:15:55.450
us two years to write that book, the two of us.
00:15:56.909 --> 00:16:01.620
And it wasn't an easy two years. But at the end
00:16:01.620 --> 00:16:04.700
of it. I felt really happy and really proud of
00:16:04.700 --> 00:16:08.899
what we'd done. And we were getting very good,
00:16:08.940 --> 00:16:12.500
not so much feedback as it was clear that we
00:16:12.500 --> 00:16:16.559
were resonating with people and that the ideas
00:16:16.559 --> 00:16:18.899
that were being expressed were kind of like things
00:16:18.899 --> 00:16:21.139
that other people knew but hadn't found a way
00:16:21.139 --> 00:16:24.620
to express before. And so in the same way that
00:16:24.620 --> 00:16:28.039
coding helped me get sort of the inner Dave out,
00:16:28.299 --> 00:16:30.960
I discovered I could do the same with writing.
00:16:31.679 --> 00:16:34.679
and also with speaking. Because one of the things
00:16:34.679 --> 00:16:37.440
about writing a book, back then anyway, was you
00:16:37.440 --> 00:16:40.679
get invited to conferences. And the first couple
00:16:40.679 --> 00:16:42.860
are absolute disasters. But after you find your
00:16:42.860 --> 00:16:46.539
voice, it's again, really good fun. And you can
00:16:46.539 --> 00:16:49.720
actually connect with people in large numbers.
00:16:50.690 --> 00:16:53.470
So that's how it got to that. The reason we actually
00:16:53.470 --> 00:16:56.990
had created the bookshelf is because after the
00:16:56.990 --> 00:17:00.850
big year 2000 buildup, where... you know the
00:17:00.850 --> 00:17:04.369
sky's falling and there were like people were
00:17:04.369 --> 00:17:06.529
hiring people off the street to become programmers
00:17:06.529 --> 00:17:08.690
because everybody was sure they needed you know
00:17:08.690 --> 00:17:13.990
a thousand cobalt developers uh well that didn't
00:17:13.990 --> 00:17:17.309
happen and the industry was oversupplied with
00:17:17.309 --> 00:17:21.089
developers uh undersupplied with money and so
00:17:21.089 --> 00:17:23.670
there was a kind of recession around the end
00:17:23.670 --> 00:17:29.200
of 2000 and consultancy dried up And at that
00:17:29.200 --> 00:17:34.440
time, we'd just written the Ruby book. And we
00:17:34.440 --> 00:17:37.839
thought, okay, so let's think about this book
00:17:37.839 --> 00:17:41.029
writing thing. What else can we do? Came up with
00:17:41.029 --> 00:17:45.670
the idea for the starter kit. And then we issued
00:17:45.990 --> 00:17:49.210
the phrase that should really have been an alarm
00:17:49.210 --> 00:17:52.150
bell for anybody in any industry at any time.
00:17:52.509 --> 00:17:55.509
We said, let's publish it ourselves. How hard
00:17:55.509 --> 00:17:59.750
could it be? And we discovered how hard it would
00:17:59.750 --> 00:18:03.170
be, but only over time. So that's how we ended
00:18:03.170 --> 00:18:05.430
up with the bookshelf. But the bookshelf pioneered
00:18:05.430 --> 00:18:10.470
a lot of different and new things for writers
00:18:10.470 --> 00:18:12.970
and how books are produced and how books are
00:18:12.970 --> 00:18:16.670
published. I think I had never seen, up until
00:18:16.670 --> 00:18:20.789
that point, anyone was even talking about doing
00:18:20.789 --> 00:18:26.329
beta. books or or reader feedback or digital
00:18:26.329 --> 00:18:28.789
first publishing I mean your entire tool chain
00:18:28.789 --> 00:18:33.769
is just in my opinion miles beyond what anything
00:18:33.769 --> 00:18:37.609
else has is out there and has ever been out there
00:18:37.609 --> 00:18:43.210
and and it is very much so geared toward developers,
00:18:43.450 --> 00:18:45.869
right? Some of the practices that you put into
00:18:45.869 --> 00:18:50.289
place. You took it further with DRM -free e -books
00:18:50.289 --> 00:18:54.150
and frequent revisions where it was almost like
00:18:54.150 --> 00:18:59.309
a live version of the book. And I mean, it's
00:18:59.309 --> 00:19:04.049
obvious that you, maybe I'm wrong, from the outside,
00:19:04.190 --> 00:19:06.890
it looks obvious that some of what you developed
00:19:06.890 --> 00:19:10.009
with the bookshelf did come directly from your
00:19:10.009 --> 00:19:14.490
process as a programmer. Is that true? Oh, absolutely.
00:19:14.869 --> 00:19:18.470
Because we had... You had gerbils. Oh, we had,
00:19:18.690 --> 00:19:21.329
yes. We were the first people to have gerbils.
00:19:21.390 --> 00:19:25.569
You're absolutely right. Yep. And boy, I mean,
00:19:25.630 --> 00:19:33.049
the ASPCA. No, anyway. Yeah, I mean, we had no
00:19:33.049 --> 00:19:36.490
idea what we were doing, literally. I mean, the
00:19:36.680 --> 00:19:40.420
tools back in 2000 or 1998 when we were doing
00:19:40.420 --> 00:19:42.910
Pragmatic Programmer. The tool the publishers
00:19:42.910 --> 00:19:46.680
used the most was probably FrameMaker, which
00:19:46.680 --> 00:19:49.599
is a really big, cumbersome, what -you -see -is
00:19:49.599 --> 00:19:54.690
-what -you -get type layout system. Costs an
00:19:54.690 --> 00:19:58.250
arm and a leg, and we couldn't afford that, but
00:19:58.250 --> 00:20:00.250
we still wanted to actually create a book that
00:20:00.250 --> 00:20:04.089
we could look at it as we were writing it in
00:20:04.089 --> 00:20:10.289
the way it would be. And so we used a system
00:20:10.450 --> 00:20:15.519
called LaTeX, which is a kind of, you write plain
00:20:15.519 --> 00:20:19.059
text and it creates actually really beautiful
00:20:19.059 --> 00:20:27.259
laid out print ready PDFs. And so we write Pragmatic
00:20:27.259 --> 00:20:30.859
Programmer using that. And LaTeX is effectively
00:20:30.859 --> 00:20:32.740
a programming language like any other. So you
00:20:32.740 --> 00:20:36.819
just use version control to store all your work.
00:20:38.109 --> 00:20:41.009
And we would have automation that would build
00:20:41.009 --> 00:20:43.369
that. And also so we could see what the final
00:20:43.369 --> 00:20:46.109
product would look like. We had tests, believe
00:20:46.109 --> 00:20:48.690
it or not, so that we could make sure we hadn't
00:20:48.690 --> 00:20:52.289
broken anything. And so we had that process.
00:20:52.630 --> 00:20:54.670
And then we did the same thing with the Ruby
00:20:54.670 --> 00:20:58.250
book, but in spades. In fact, I'm actually currently
00:20:58.250 --> 00:21:01.930
finishing off the sixth edition of Programming
00:21:01.930 --> 00:21:08.349
Ruby. And it does some, in the Ruby book, all
00:21:08.349 --> 00:21:12.319
of those code examples you see, and you see like,
00:21:12.339 --> 00:21:14.220
you know, one plus two, and then there's a little
00:21:14.220 --> 00:21:17.920
arrow and it says the result is three. That's
00:21:17.920 --> 00:21:20.880
actually done every single time you build the
00:21:20.880 --> 00:21:23.599
book. It actually runs that code and shows you
00:21:23.599 --> 00:21:27.410
the output so that you know that what comes out
00:21:27.410 --> 00:21:30.250
in that book is honest because it was actually
00:21:30.250 --> 00:21:34.089
generated as the book was produced. That code
00:21:34.089 --> 00:21:41.559
that does that is 25 years old. It's just fun.
00:21:41.900 --> 00:21:44.480
So we got to this point where, you know, we were
00:21:44.480 --> 00:21:47.500
saying, okay, we need to publish. So we basically
00:21:47.500 --> 00:21:49.400
just went and said, well, how do you do that?
00:21:49.480 --> 00:21:52.960
We didn't know. We had no idea. So we just did
00:21:52.960 --> 00:21:55.559
what came naturally, which was, you know, to
00:21:55.559 --> 00:21:58.779
do it the way programmers would do it. And it
00:21:58.779 --> 00:22:00.859
turned out, like I say, it was different. There
00:22:00.859 --> 00:22:04.920
was a funny story. We were around about 2008
00:22:04.920 --> 00:22:08.880
or so. I went to my one and only publishing conference
00:22:08.880 --> 00:22:14.869
in New York City. And I'd taken my laptop. We
00:22:14.869 --> 00:22:16.430
were working on a book. I can't remember. I think
00:22:16.430 --> 00:22:19.339
it was maybe Joe Armstrong's or something. I
00:22:19.339 --> 00:22:23.400
can't remember. But I was sitting in the audience,
00:22:23.539 --> 00:22:27.960
and there was a woman who was talking about how
00:22:27.960 --> 00:22:32.319
you create a version of your book for the Kindle.
00:22:34.369 --> 00:22:36.970
So she was talking about, well, you go to FrameMaker
00:22:36.970 --> 00:22:39.009
and you get this and export it to this, and then
00:22:39.009 --> 00:22:42.190
you go through this and that. And we've reached
00:22:42.190 --> 00:22:45.009
a point now where we can do that in less than
00:22:45.009 --> 00:22:51.140
a week. And during her talk, without trying to
00:22:51.140 --> 00:22:54.380
be in any way ironic, I actually reformatted
00:22:54.380 --> 00:22:58.009
Joe's book at least two or three times, producing
00:22:58.009 --> 00:23:01.549
both a camera -ready PDF and also an Amazon -ready
00:23:01.549 --> 00:23:05.380
Kindle version of it. just because the tools
00:23:05.380 --> 00:23:07.579
were all automated and they just did that from
00:23:07.579 --> 00:23:11.420
the start. So, yeah, I mean, I think we were
00:23:11.420 --> 00:23:14.779
in many ways, I think we were ahead of the game
00:23:14.779 --> 00:23:17.400
because we cared about the writer and we cared
00:23:17.400 --> 00:23:20.900
about the reader. We paid, I think, still the
00:23:20.900 --> 00:23:26.099
highest royalties of a publisher. So, I mean,
00:23:26.119 --> 00:23:28.460
and that's because we just felt it was fair,
00:23:28.579 --> 00:23:31.720
you know. We were also possibly the worst business
00:23:31.720 --> 00:23:35.809
people that have ever turned publisher. We survived
00:23:35.809 --> 00:23:39.009
fundamentally on word of mouth and goodwill.
00:23:40.779 --> 00:23:45.319
And in retrospect, I mean, that's probably just
00:23:45.319 --> 00:23:47.880
who we were. But in retrospect, we could have
00:23:47.880 --> 00:23:51.279
probably done a lot more to turn it into a grown
00:23:51.279 --> 00:23:56.430
-up business and not just, you know, an evangelical
00:23:56.430 --> 00:23:59.849
business. I want to talk about the book that
00:23:59.849 --> 00:24:02.250
you just released. I didn't know much about it
00:24:02.250 --> 00:24:06.200
until recently. And when I started, again, I
00:24:06.200 --> 00:24:08.220
did all this research on you, even though I've
00:24:08.220 --> 00:24:11.859
known you for years. Anyway, your book, Simplicity,
00:24:12.019 --> 00:24:15.519
spelled with a lowercase s, which at first I
00:24:15.519 --> 00:24:18.420
was like, no. And then I was like, wait a second,
00:24:18.500 --> 00:24:21.420
that makes sense. So tell me about Simplicity,
00:24:21.599 --> 00:24:24.000
the book Simplicity, how it how it came about.
00:24:24.380 --> 00:24:26.920
If you know, because I don't think you had written
00:24:26.920 --> 00:24:32.420
anything new. within the few years before that,
00:24:32.500 --> 00:24:35.980
right? Or did I miss a book? No, I hadn't. No,
00:24:36.039 --> 00:24:39.859
I had not. I think the last thing I wrote was
00:24:39.859 --> 00:24:42.799
the various editions of the Elixir book. So how
00:24:42.799 --> 00:24:46.539
was it returning to a full book, all you? It
00:24:46.539 --> 00:24:49.980
was, actually, it was the hardest thing I've
00:24:49.980 --> 00:24:55.440
ever written. And it was hard because I knew
00:24:55.440 --> 00:24:59.619
what I wanted to say. I could not find a way
00:24:59.619 --> 00:25:06.190
of saying it. So, typically, I can write a book,
00:25:06.230 --> 00:25:09.930
and I'll typically go pretty much beginning to
00:25:09.930 --> 00:25:13.730
end, then go back and just throw away half a
00:25:13.730 --> 00:25:16.430
chapter, rewrite this, whatever. This book, I
00:25:16.430 --> 00:25:19.289
actually did three complete and totally different
00:25:19.289 --> 00:25:25.509
drafts of the book. And I was struggling really
00:25:25.509 --> 00:25:29.690
hard to make sense of what it was I was trying
00:25:29.690 --> 00:25:34.220
to get across. partly because I was trying very,
00:25:34.279 --> 00:25:38.079
very hard not to be prescriptive. So the idea
00:25:38.079 --> 00:25:39.920
behind the book is that we're making software
00:25:39.920 --> 00:25:42.779
too difficult, and it doesn't have to be as difficult
00:25:42.779 --> 00:25:47.970
as we're making it. At the same time, what is
00:25:47.970 --> 00:25:50.910
appropriate in one context is not appropriate
00:25:50.910 --> 00:25:53.710
in a different context. And so anybody that sits
00:25:53.710 --> 00:25:55.829
there and says, this is how you should write
00:25:55.829 --> 00:26:00.210
software is wrong. So I was trying not to be
00:26:00.210 --> 00:26:03.309
prescriptive, but at the same time to say, this
00:26:03.309 --> 00:26:06.750
is how you should do it. And so I couldn't find
00:26:06.750 --> 00:26:12.650
a way of making that work. And eventually I came
00:26:12.650 --> 00:26:16.289
on like a style that I could live with, but it
00:26:16.289 --> 00:26:18.849
still wasn't very good. So for the first time
00:26:18.849 --> 00:26:22.049
ever, actually, writing a book, and I'd probably
00:26:22.049 --> 00:26:25.680
written, if you include editions, probably 20
00:26:25.680 --> 00:26:30.769
by then. I actually worked with an editor, Susanna,
00:26:30.829 --> 00:26:35.509
Susanna Davidson, which was, I should have done
00:26:35.509 --> 00:26:38.230
it from the word go because it really changed.
00:26:39.690 --> 00:26:43.849
It really changed how I looked at it. It's kind
00:26:43.849 --> 00:26:49.049
of like she was acting as an audience and also
00:26:49.049 --> 00:26:51.869
as a conscience at the same time, you know, and
00:26:51.869 --> 00:26:54.910
that kind of very much changed the way I thought
00:26:54.910 --> 00:27:00.440
about the book. That's how it came to be. You're
00:27:00.440 --> 00:27:03.700
right, there was a very long gap. But it was
00:27:03.700 --> 00:27:07.079
kind of something I wanted to say. Yeah, I will
00:27:07.079 --> 00:27:11.619
always strongly encourage people to have an editor,
00:27:11.759 --> 00:27:15.000
even if they are the best writer on the planet,
00:27:15.019 --> 00:27:18.319
even if they are an editor themselves. It's just,
00:27:18.359 --> 00:27:22.450
I don't think I could do anything without an
00:27:22.450 --> 00:27:25.640
editor. Not just for helping to get your thoughts
00:27:25.640 --> 00:27:28.259
together and try and pull apart what it is you're
00:27:28.259 --> 00:27:30.700
trying to say. And, you know, the old adage,
00:27:30.700 --> 00:27:32.819
you're too close to your own words. You know
00:27:32.819 --> 00:27:35.319
what's supposed to be there and it's not supposed
00:27:35.319 --> 00:27:37.460
to be there. And I'm not just talking line editing.
00:27:37.539 --> 00:27:39.680
I'm talking about full -on developmental editing.
00:27:40.200 --> 00:27:43.900
But beyond that, it's always nice to have somebody
00:27:43.900 --> 00:27:46.660
behind you, tapping you on your shoulder and
00:27:46.660 --> 00:27:49.039
asking you, do you have that chapter done yet?
00:27:49.160 --> 00:27:52.730
I'm still waiting. It's a good motivator. Yeah.
00:27:53.609 --> 00:27:57.390
And it's also, I don't think people realize,
00:27:57.589 --> 00:28:01.470
I mean, it probably takes for a decent sized
00:28:01.470 --> 00:28:05.349
technical book, I would guess three to 500 hours
00:28:05.349 --> 00:28:09.980
to write. And when you think of it that way,
00:28:10.019 --> 00:28:12.319
people doing it in their spare time, you know,
00:28:12.319 --> 00:28:17.309
evenings and weekends, that's an incredible commitment
00:28:17.309 --> 00:28:20.569
to make. And if you're doing it totally on your
00:28:20.569 --> 00:28:22.690
own, you're just sitting there in your garret
00:28:22.690 --> 00:28:27.059
writing, right? It's a very lonely business.
00:28:27.950 --> 00:28:32.339
And having someone come along with you and kind
00:28:32.339 --> 00:28:35.180
of like not just doing, you know, is that chapter
00:28:35.180 --> 00:28:37.460
ready, but actually sharing the journey, you
00:28:37.460 --> 00:28:42.569
know? The editors, I mean, our editors at the
00:28:42.569 --> 00:28:45.089
bookshelf were really, really good. And they
00:28:45.089 --> 00:28:48.619
would spend a lot of their time saying, OK, I
00:28:48.619 --> 00:28:50.559
kind of understand what you're trying to say
00:28:50.559 --> 00:28:54.019
here, but, you know, dot, dot, dot, dot. And
00:28:54.019 --> 00:28:58.339
actually working and helping craft the book.
00:28:58.400 --> 00:29:00.279
So you felt you weren't alone. You felt that
00:29:00.279 --> 00:29:02.940
you were actually, you know, you were part of
00:29:02.940 --> 00:29:05.039
a team that was producing it. And I think for
00:29:05.039 --> 00:29:09.349
many people, that is an essential component of
00:29:09.349 --> 00:29:13.109
actually getting done. It's the author -editor
00:29:13.109 --> 00:29:15.809
relationship is one of the most wonderful relationships
00:29:15.809 --> 00:29:20.519
I have ever had. been a part of from both sides,
00:29:20.680 --> 00:29:22.759
right? Because I've played both sides of the
00:29:22.759 --> 00:29:27.180
fence and it's irreplaceable when it comes to
00:29:27.180 --> 00:29:32.509
this industry, I think. So, yeah. And that's
00:29:32.509 --> 00:29:34.089
another thing that you don't get when you're
00:29:34.089 --> 00:29:38.809
looking at AI, right? You're looking at, not
00:29:38.809 --> 00:29:40.750
only are you looking at non -human generated
00:29:40.750 --> 00:29:44.009
content, but you're looking at non -human moderated
00:29:44.009 --> 00:29:47.869
content as well. You were like one of the main...
00:29:48.170 --> 00:29:52.430
behind the whole Agile movement. And what I read
00:29:52.430 --> 00:29:56.150
about that was even more shocking, that you sort
00:29:56.150 --> 00:29:59.309
of, that movement sort of went in a different
00:29:59.309 --> 00:30:03.589
direction than what you all had intended it to
00:30:03.589 --> 00:30:08.069
do, be, and provide for developers. First, for
00:30:08.069 --> 00:30:10.990
our listeners, can you explain what Agile is
00:30:10.990 --> 00:30:13.910
and how you all got together and decided to make
00:30:13.910 --> 00:30:16.529
that thing and then what it turned into now?
00:30:17.180 --> 00:30:20.119
To talk about agile, we have to talk about software,
00:30:20.339 --> 00:30:23.940
which is maybe a bit boring. But back in the
00:30:23.940 --> 00:30:31.339
1980s and 1990s, when software was still being
00:30:31.339 --> 00:30:34.519
kind of worked out, it became obvious that we
00:30:34.519 --> 00:30:38.220
didn't really know how to do it. And projects
00:30:38.220 --> 00:30:41.640
were getting worse and worse and worse. You'd
00:30:41.640 --> 00:30:45.359
commission a project, and the chances of it coming
00:30:45.359 --> 00:30:50.369
in were... you know initially maybe 50 50. um
00:30:50.369 --> 00:30:54.190
and as time went on that number went down not
00:30:54.190 --> 00:30:58.650
up and so by the end of the 90s only one project
00:30:58.650 --> 00:31:03.569
in six actually came in on time and budget so
00:31:03.990 --> 00:31:07.150
you know if you were a company and you said i
00:31:07.150 --> 00:31:09.250
want some new software it was like rolling a
00:31:09.250 --> 00:31:13.150
dice and only if it came up six were you okay
00:31:14.279 --> 00:31:16.960
And the reaction of the industry to this was
00:31:16.960 --> 00:31:22.380
to try to specify more and more what this software
00:31:22.380 --> 00:31:24.559
had to do. So before you started writing software,
00:31:24.839 --> 00:31:27.380
you were supposed to sit down and specify exactly
00:31:27.380 --> 00:31:31.930
what it did, which sounds great in theory, but
00:31:32.440 --> 00:31:35.710
nobody actually knows what they want. You know,
00:31:35.730 --> 00:31:39.450
they may think they do, but our experience is
00:31:39.450 --> 00:31:41.569
that no, they don't really. And if you think
00:31:41.569 --> 00:31:45.369
of yourself, if someone forces you to make detailed
00:31:45.369 --> 00:31:48.289
decisions about some some everyday thing you
00:31:48.289 --> 00:31:51.210
know like how do i how do i make a peanut butter
00:31:51.210 --> 00:31:55.069
sandwich it's not easy to describe you know i
00:31:55.069 --> 00:31:57.009
mean there's all sorts of little small kinks
00:31:57.009 --> 00:31:58.630
that they can get in there they get in the way
00:31:58.630 --> 00:32:02.910
and so by making people focus more and more and
00:32:02.910 --> 00:32:06.630
more on specifying what they had to do they actually
00:32:06.630 --> 00:32:09.650
drove people away from getting what they wanted
00:32:09.650 --> 00:32:17.220
out of this software pretty obvious, like I said,
00:32:17.240 --> 00:32:20.599
round about the turn of the century. And at the
00:32:20.599 --> 00:32:23.839
same time, Andy Hunt and I had written a book
00:32:23.839 --> 00:32:28.339
called Pragmatic Programmer, which did pretty
00:32:28.339 --> 00:32:31.140
well. And that led to us going to conferences
00:32:31.140 --> 00:32:33.220
to talk. And we ended up talking to a whole bunch
00:32:33.220 --> 00:32:36.960
of people at conferences. And a lot of the speakers
00:32:36.960 --> 00:32:39.599
were going through the same kind of thinking
00:32:39.599 --> 00:32:42.059
that we were. And that was that software has
00:32:42.059 --> 00:32:44.180
to be done a different way. This current way
00:32:44.180 --> 00:32:47.480
of trying to specify it to death was just not
00:32:47.480 --> 00:32:53.500
working. So we agreed to meet up and we spent
00:32:53.500 --> 00:32:59.220
three days in actually a ski resort in Utah.
00:33:00.339 --> 00:33:02.680
And we came up with this thing called the Agile
00:33:02.680 --> 00:33:09.630
Manifesto. And it was basically a different way
00:33:09.630 --> 00:33:12.839
of looking at software. And we said, software
00:33:12.839 --> 00:33:16.460
is about communicating. It's about talking and
00:33:16.460 --> 00:33:21.099
interacting with the customers. It's about doing
00:33:21.099 --> 00:33:27.160
and not talking about doing. And it's about reacting.
00:33:27.359 --> 00:33:29.740
It's about saying, we'll start off in this direction,
00:33:29.819 --> 00:33:32.140
but if this direction is not right, we're going
00:33:32.140 --> 00:33:34.819
to pivot. And we're going to enjoy pivoting.
00:33:34.819 --> 00:33:36.559
We're going to set it up in such a way that you
00:33:36.559 --> 00:33:41.779
have to be able to pivot. And that was kind of
00:33:41.779 --> 00:33:44.849
revolutionary at the time. But at the same time,
00:33:44.869 --> 00:33:46.549
I came away from there thinking, hey, this is
00:33:46.549 --> 00:33:50.130
fantastic. We did a good job. And then just returned
00:33:50.130 --> 00:33:55.529
back to doing what I was doing. And to everybody's
00:33:55.529 --> 00:34:00.069
surprise, I think it did the 2001 equivalent
00:34:00.069 --> 00:34:04.920
of going viral. We got... Tens of thousands of
00:34:04.920 --> 00:34:07.500
people signing up on the website, eventually
00:34:07.500 --> 00:34:11.940
hundreds of thousands. And people started listening.
00:34:12.019 --> 00:34:14.000
People started writing projects that way and
00:34:14.000 --> 00:34:20.139
reporting great success. And they were doing
00:34:20.139 --> 00:34:24.260
it basically with no support other than the manifesto
00:34:24.260 --> 00:34:27.239
and a couple of wikis that described how it should
00:34:27.239 --> 00:34:31.199
all work. So that was the start of what became
00:34:31.199 --> 00:34:36.219
Agility. And what you might be talking about
00:34:36.219 --> 00:34:40.260
when you said, you know, not happy is that it
00:34:40.260 --> 00:34:45.460
very rapidly changed tone about five years in.
00:34:45.559 --> 00:34:49.380
Yeah, I had read that it became things where
00:34:49.380 --> 00:34:51.860
you needed certifications and then there were
00:34:51.860 --> 00:34:55.739
frameworks and processes and all sorts of things
00:34:55.739 --> 00:34:58.820
that came about. In fact, I think I read somewhere
00:34:58.820 --> 00:35:03.760
that... you had said something to the effect
00:35:03.760 --> 00:35:08.679
of agile isn't something you do. Agility is how
00:35:08.679 --> 00:35:12.159
you work. And that sort of like everything that
00:35:12.159 --> 00:35:15.320
was going on sort of like dismantled all of that.
00:35:15.980 --> 00:35:19.599
Yeah, that's really true. I mean, I think a lot
00:35:19.599 --> 00:35:22.699
of it was, well, I think there were two things
00:35:22.699 --> 00:35:27.440
that played into that. One was fear and the other
00:35:27.440 --> 00:35:33.280
one was money. The fear is that, And doing things
00:35:33.280 --> 00:35:36.079
in an Agile way requires you to be responsible.
00:35:36.860 --> 00:35:41.039
And you have to basically decide for yourself
00:35:41.039 --> 00:35:43.159
how things should be done and what should be
00:35:43.159 --> 00:35:45.750
done. And for a lot of people, that's scary.
00:35:46.409 --> 00:35:49.289
They would much rather have the comfort of being
00:35:49.289 --> 00:35:53.750
told what to do than just doing it. And so there's
00:35:53.750 --> 00:35:57.090
a large number of people who kind of reacted
00:35:57.090 --> 00:36:00.590
against Agile for that. And then there's another
00:36:00.590 --> 00:36:02.789
group of people who said, wait a minute, the
00:36:02.789 --> 00:36:06.070
software industry is losing literally billions
00:36:06.070 --> 00:36:10.500
of dollars every year. And if this agile can
00:36:10.500 --> 00:36:14.670
help, then people will pay for it. And so they
00:36:14.670 --> 00:36:18.510
came up with various ways of monetizing agile.
00:36:19.750 --> 00:36:25.179
The first way was they took the word agile and
00:36:25.599 --> 00:36:29.929
they turned it into a noun. And because it's
00:36:29.929 --> 00:36:33.110
a noun, you can sell it, right? I mean, you cannot,
00:36:33.250 --> 00:36:36.110
I couldn't sell green, but I could sell a green
00:36:36.110 --> 00:36:39.230
apple. And so you had to turn it into a noun.
00:36:39.329 --> 00:36:43.570
They did. And so it was no longer the manifesto
00:36:43.570 --> 00:36:45.949
for agile software development. It became agile,
00:36:46.210 --> 00:36:48.289
you know, and you'd go out and you'd buy it by
00:36:48.289 --> 00:36:53.139
the pound. And so those two factors combined
00:36:53.139 --> 00:36:56.260
to make agile kind of almost the opposite of
00:36:56.260 --> 00:37:00.530
what it should be. It became processes. And the
00:37:00.530 --> 00:37:02.889
people who were a little bit frightened felt
00:37:02.889 --> 00:37:04.829
happy because they were now sitting inside a
00:37:04.829 --> 00:37:06.550
process and they were doing things the process
00:37:06.550 --> 00:37:11.630
way. What is it about the process that is sort
00:37:11.630 --> 00:37:15.230
of like a detriment to a developer or any kind
00:37:15.230 --> 00:37:18.809
of position, really? Writer, editor, anything
00:37:18.809 --> 00:37:22.059
like that? I don't know if it's a detriment.
00:37:22.900 --> 00:37:25.880
It's kind of a fact of life that you maybe...
00:37:26.250 --> 00:37:29.800
are uncomfortable with. And that's the idea that
00:37:29.800 --> 00:37:33.820
you're never going to get it right. You're always
00:37:33.820 --> 00:37:37.260
going to be iterating towards a solution and
00:37:37.260 --> 00:37:41.090
you've got to be prepared to iterate, which means
00:37:41.090 --> 00:37:46.269
that you have to make it cheap and easy to change
00:37:46.269 --> 00:37:49.969
things. So that if you go off in the wrong direction
00:37:49.969 --> 00:37:52.949
and, you know, it turns out not to be the right
00:37:52.949 --> 00:37:56.030
direction, you can go back and set off in a different
00:37:56.030 --> 00:37:59.920
direction. without losing everything you've ever
00:37:59.920 --> 00:38:04.250
done. Make every step small and make it reversible.
00:38:04.510 --> 00:38:08.739
And if you do that, then it's easy. And you do
00:38:08.739 --> 00:38:13.639
your project. And as you go along, you're gathering
00:38:13.639 --> 00:38:17.119
feedback. And you gather feedback for everything
00:38:17.119 --> 00:38:19.679
that you do, from the smallest decision to the
00:38:19.679 --> 00:38:23.119
biggest. And the idea is then that you use that
00:38:23.119 --> 00:38:27.460
feedback to guide your assessment. of how well
00:38:27.460 --> 00:38:30.159
you're doing, right? So you take a step, you
00:38:30.159 --> 00:38:32.719
look at where you go to. Am I moving in the right
00:38:32.719 --> 00:38:37.820
direction? Am I starting to get closer to value
00:38:37.820 --> 00:38:40.460
or whatever it is that you're measuring? And
00:38:40.460 --> 00:38:43.119
if you are, you keep going. If you're not, you
00:38:43.119 --> 00:38:47.130
undo what you just did. Throw it away and go
00:38:47.130 --> 00:38:50.090
in a slightly different direction. And it's that
00:38:50.090 --> 00:38:52.210
act of throwing away that a lot of people have
00:38:52.210 --> 00:38:54.199
trouble with. You know what they say in writing,
00:38:54.300 --> 00:38:57.099
kill your darlings is very, very difficult to
00:38:57.099 --> 00:39:00.739
do. There's actually an awful lot in common between
00:39:00.739 --> 00:39:05.550
writing and software development. And I think
00:39:05.550 --> 00:39:09.130
a lot of the ideas behind agility, I think if
00:39:09.130 --> 00:39:11.710
anything, writers probably tend to plan, or certainly
00:39:11.710 --> 00:39:14.949
fiction writers tend to plan a lot more than
00:39:14.949 --> 00:39:18.210
developers do, because you have to maintain story
00:39:18.210 --> 00:39:20.130
arcs and this kind of thing that you know from
00:39:20.130 --> 00:39:25.329
the beginning. But you can definitely write in
00:39:25.329 --> 00:39:28.070
an agile kind of way, as long as you're not afraid
00:39:28.070 --> 00:39:31.800
to throw stuff away. Would you say agile, like
00:39:31.800 --> 00:39:33.219
you mentioned, it kind of became an industry,
00:39:33.440 --> 00:39:36.659
right? It started as something of a guideline
00:39:36.659 --> 00:39:39.559
for software development and to meet deadlines,
00:39:39.820 --> 00:39:42.480
to, as the word implies, kind of redundantly
00:39:42.480 --> 00:39:45.440
be agile, be flexible, adapt to the needs. Because
00:39:45.440 --> 00:39:46.880
a lot of the times a client comes to you and
00:39:46.880 --> 00:39:49.760
says, Dave, I need this. But that's like the
00:39:49.760 --> 00:39:52.599
macro 30 ,000 foot kind of view. They don't know
00:39:52.599 --> 00:39:54.699
that we need to land the plane. And there's so
00:39:54.699 --> 00:39:56.780
many small details when you actually start breaking
00:39:56.780 --> 00:39:59.840
down what it is that every little piece of the
00:39:59.840 --> 00:40:02.800
software needs to do. With AI being introduced,
00:40:03.159 --> 00:40:07.150
would you say that even puts more of a wrench,
00:40:07.449 --> 00:40:09.710
throws a wrench at the plans of agile? Because
00:40:09.710 --> 00:40:14.190
now you really have to be agile, but not in the
00:40:14.190 --> 00:40:18.110
sort of noun. industrial sort of become a certified
00:40:18.110 --> 00:40:21.090
agile scrum master kind of way but really these
00:40:21.090 --> 00:40:23.730
tools are rapidly evolving and at the same time
00:40:23.730 --> 00:40:26.730
changing how we develop software that it is probably
00:40:26.730 --> 00:40:28.869
even more i don't want to say obsolete because
00:40:28.869 --> 00:40:31.730
there's good paradigms and guidelines to use
00:40:31.730 --> 00:40:34.909
but the strictness that it's become around like
00:40:34.909 --> 00:40:37.489
the sprints in two weeks and this and estimates
00:40:37.489 --> 00:40:39.289
and retros and all of that that it became sort
00:40:39.289 --> 00:40:41.869
of ceremonious oh yeah this just throws even
00:40:41.869 --> 00:40:44.909
more of a Or wrenching those plans. Yeah, I mean,
00:40:44.949 --> 00:40:51.210
the advent of AI has shown that the ceremonial
00:40:51.369 --> 00:40:55.389
kind of mock agile that people have been doing
00:40:55.389 --> 00:41:00.610
is rubbish. It doesn't work. Never has, really.
00:41:00.889 --> 00:41:04.889
But AI has basically totally knocked that away.
00:41:06.340 --> 00:41:10.420
The interesting thing is, though, that agile
00:41:10.829 --> 00:41:16.960
methods... actually translate pretty much one
00:41:16.960 --> 00:41:21.079
-to -one into the AI world. The difference is
00:41:21.079 --> 00:41:23.539
that the timescales shrink from days to minutes.
00:41:25.539 --> 00:41:30.170
But when I'm using AI to code right now, I am
00:41:30.170 --> 00:41:35.570
constantly working on how to get feedback from
00:41:35.570 --> 00:41:38.570
what the AI is doing, how to monitor what I think
00:41:38.570 --> 00:41:42.510
it's doing. And when I see it going off track,
00:41:42.650 --> 00:41:45.510
I'm stopping. I'm making it roll back. and I'm
00:41:45.510 --> 00:41:46.869
making it go in a different direction, which
00:41:46.869 --> 00:41:49.769
is just what you do when you're manually coding.
00:41:50.570 --> 00:41:54.710
So in a way, I think AI is going to make the
00:41:54.710 --> 00:42:01.159
agile movement become more honest and more aligned
00:42:01.159 --> 00:42:05.300
with the initial values of the manifesto and
00:42:05.369 --> 00:42:08.469
move it away from this kind of agile industrial
00:42:08.469 --> 00:42:11.809
complex. What do you think of the whole AI movement?
00:42:12.210 --> 00:42:14.809
In general. I know that one other thing that
00:42:14.809 --> 00:42:16.510
I learned about you while I was doing all this
00:42:16.510 --> 00:42:20.829
research is you sort of coined the idea of Code
00:42:20.829 --> 00:42:24.809
Kata, which is the programmer doing exercises.
00:42:24.869 --> 00:42:28.349
And as you describe, well, actually, why don't
00:42:28.349 --> 00:42:32.150
you describe what Code Kata is? Yeah, it came
00:42:32.150 --> 00:42:36.809
about, my son used to do karate and they would
00:42:36.809 --> 00:42:40.010
spend hours on the mat just repeating exercises
00:42:40.010 --> 00:42:43.960
over and over again. And these exercises would
00:42:43.960 --> 00:42:50.590
typically be a kind of set of moves that might
00:42:50.590 --> 00:42:53.289
be involved in an actual fight, but they were
00:42:53.289 --> 00:42:56.349
very, very stylized, very ritualized. And you
00:42:56.349 --> 00:42:58.949
would do this over and over and over and over
00:42:58.949 --> 00:43:05.289
again. And the reason was that you had to get
00:43:05.289 --> 00:43:08.969
your body to understand what these various movements
00:43:08.969 --> 00:43:13.369
were without having to think about it. when you're
00:43:13.369 --> 00:43:15.409
actually then competing against somebody else
00:43:15.409 --> 00:43:18.130
things are coming at you so fast you can't stop
00:43:18.130 --> 00:43:20.289
and say oh now he's doing that so i should do
00:43:20.289 --> 00:43:22.110
this and then move my hand like this and whatever
00:43:22.110 --> 00:43:25.710
else so by repeating these exercises over and
00:43:25.710 --> 00:43:29.110
over again you're training your body how to link
00:43:29.110 --> 00:43:31.610
things together without you consciously having
00:43:31.610 --> 00:43:36.559
to do it and so this idea of repetition to to
00:43:36.559 --> 00:43:39.380
learn something is really really powerful so
00:43:40.070 --> 00:43:42.030
I was thinking about that, and I realized that
00:43:42.030 --> 00:43:44.630
I did the same when I was coding, in that there
00:43:44.630 --> 00:43:48.429
are certain little exercises I will tend to do
00:43:48.429 --> 00:43:50.909
when I pick up, for example, a new programming
00:43:50.909 --> 00:43:53.730
language. In order to make sure that I actually
00:43:53.730 --> 00:43:57.610
understand it, I will do the same things I did
00:43:57.610 --> 00:44:00.590
in my previous language, just practice these
00:44:00.590 --> 00:44:02.860
little things. And I'll do them... Sometimes
00:44:02.860 --> 00:44:05.219
many times. And you do it, you throw it away.
00:44:05.340 --> 00:44:06.840
You say, okay, can I do that differently? And
00:44:06.840 --> 00:44:08.360
you do it a different way and you throw it away.
00:44:08.599 --> 00:44:11.039
And all you're doing is you're teaching your
00:44:11.039 --> 00:44:14.260
brain the kind of patterns behind the language,
00:44:14.400 --> 00:44:18.510
the way everything hangs together. So that's
00:44:18.510 --> 00:44:22.110
where Code Carter came from. And I put together,
00:44:22.269 --> 00:44:26.510
I think it was 20 or something, little exercises,
00:44:26.730 --> 00:44:29.989
which were, some of them were very simple. Some
00:44:29.989 --> 00:44:32.730
of them were actually almost impossibly. hard.
00:44:33.469 --> 00:44:40.250
But most of them were things like how you handle
00:44:40.250 --> 00:44:44.230
supermarket discounts, which actually sounds
00:44:44.230 --> 00:44:46.389
really pedestrian, but actually it's got some
00:44:46.389 --> 00:44:48.969
quite tricky things about it. Like if there are
00:44:48.969 --> 00:44:50.949
two discounts on a product, what order do you
00:44:50.949 --> 00:44:53.670
apply them? Makes a difference to the price that
00:44:53.670 --> 00:44:58.710
you actually give at the end. So Carter liked
00:44:58.710 --> 00:45:01.530
that, where I just basically lay out a problem
00:45:01.530 --> 00:45:05.199
and said, There isn't a right solution. Just
00:45:05.199 --> 00:45:08.659
go and do it. And, you know, do it many times.
00:45:08.679 --> 00:45:11.179
Do it for all your different languages. And just,
00:45:11.260 --> 00:45:14.019
you know, learn what you can from it. So that
00:45:14.019 --> 00:45:16.119
was the idea. And quite surprisingly, it kind
00:45:16.119 --> 00:45:18.619
of took off, which is nice. Yeah, but how does
00:45:18.619 --> 00:45:23.059
it tie into, you know, if developers are now,
00:45:23.219 --> 00:45:26.860
a great number of them are turning to AI to have
00:45:26.860 --> 00:45:29.940
them do their work at this point, what is it
00:45:29.940 --> 00:45:33.619
that... The humans will continue to bring to
00:45:33.619 --> 00:45:36.420
the table. If the computers are doing all the
00:45:36.420 --> 00:45:40.179
coding and doing all that kind of work, what
00:45:40.179 --> 00:45:43.360
do we even practice anymore? What we practice
00:45:43.360 --> 00:45:47.440
is actually what we used to practice, but a bit
00:45:47.440 --> 00:45:55.369
more. So it's a mistake to think that AIs can
00:45:55.369 --> 00:46:02.829
do all the coding. They can do a lot of it. Absolutely.
00:46:03.679 --> 00:46:08.030
But left unchecked, their focus is on getting
00:46:08.030 --> 00:46:12.269
something that works and churning out code that
00:46:12.269 --> 00:46:16.469
any human would be quite embarrassed to see.
00:46:17.190 --> 00:46:21.510
And the result of that is that you end up with
00:46:21.510 --> 00:46:25.869
these incredibly big code bases. I mean, the
00:46:25.869 --> 00:46:29.809
team I was working with this week, in the previous
00:46:29.809 --> 00:46:34.139
week, there's... 10 people, 11 people on the
00:46:34.139 --> 00:46:37.400
team. In the previous week, they generated 250
00:46:37.400 --> 00:46:41.559
,000 lines of code, a quarter of a million lines
00:46:41.559 --> 00:46:48.130
of code. And honestly, the functionality they
00:46:48.130 --> 00:46:51.250
delivered from that could probably be done in
00:46:51.250 --> 00:46:58.400
10, 15 ,000 lines. And the difference is that
00:46:58.400 --> 00:47:01.239
AI is very happy at just churning out code and
00:47:01.239 --> 00:47:05.980
churning out code. But not just that. tends to
00:47:05.980 --> 00:47:09.119
look at the problem you give it. You know, like
00:47:09.119 --> 00:47:11.619
I said, people don't know what they, people say
00:47:11.619 --> 00:47:13.300
they know what they want, but they don't really
00:47:13.300 --> 00:47:16.079
know what they need. And it's the same where
00:47:16.079 --> 00:47:18.179
they are. You can tell it what you want, and
00:47:18.179 --> 00:47:20.739
it will go away and slavishly produce what it
00:47:20.739 --> 00:47:24.429
is that you asked for. But will you be happy
00:47:24.429 --> 00:47:28.469
with all it delivers? Maybe, maybe not. And in
00:47:28.469 --> 00:47:31.880
my experience, not. Because a good developer
00:47:31.880 --> 00:47:35.860
is going to interact with you. And you say, I
00:47:35.860 --> 00:47:38.599
want something that's going to add these two
00:47:38.599 --> 00:47:41.989
numbers. And you go, okay, why do you want to
00:47:41.989 --> 00:47:44.969
do that? And it might be because I need to be
00:47:44.969 --> 00:47:47.650
able to give change at my checkout or I need,
00:47:47.650 --> 00:47:49.969
you know, whatever it might be. And by digging
00:47:49.969 --> 00:47:52.389
into the problem, you'll say, oh, you don't really
00:47:52.389 --> 00:47:54.710
want to do that. What you really want to do is
00:47:54.710 --> 00:47:57.650
this, this, and this. And then you iterate on
00:47:57.650 --> 00:48:00.949
that understanding. And that, to me, actually,
00:48:00.989 --> 00:48:03.070
is one of the most fun parts of being a developer.
00:48:03.409 --> 00:48:07.139
is digging into what people actually need as
00:48:07.139 --> 00:48:10.449
opposed to what they think they need. And AI
00:48:10.449 --> 00:48:14.309
is not that good at doing that because AI doesn't
00:48:14.309 --> 00:48:17.210
have, well, first of all, AI is trained to be
00:48:17.210 --> 00:48:19.769
like a little puppy, right? It's to make you
00:48:19.769 --> 00:48:23.849
happy. And so it's never going to say, oh, no,
00:48:23.949 --> 00:48:26.570
you don't want that. You want this, right? It's
00:48:26.570 --> 00:48:29.889
just going to go away and do what you want. But
00:48:29.889 --> 00:48:31.949
also it doesn't have the experience. It doesn't
00:48:31.949 --> 00:48:36.139
say like, I've been working on projects now for
00:48:36.139 --> 00:48:38.360
so many years. I've seen something like this
00:48:38.360 --> 00:48:42.019
before when a customer said this and they really
00:48:42.019 --> 00:48:44.860
meant that. And I can see the parallels between
00:48:44.860 --> 00:48:48.719
that situation and this situation. So I'm going
00:48:48.719 --> 00:48:52.119
to suggest that, no, you don't want that. You
00:48:52.119 --> 00:48:56.480
want something else instead. And right now, AI
00:48:56.480 --> 00:48:59.400
is not going to do that. So personally, I mean,
00:48:59.460 --> 00:49:02.260
if people say, what do you do? I'm a programmer.
00:49:03.889 --> 00:49:06.730
It used to be the case that I literally wrote
00:49:06.730 --> 00:49:11.409
lines of code almost every day. I am still a
00:49:11.409 --> 00:49:14.780
programmer. I haven't written a line of code,
00:49:14.840 --> 00:49:18.920
maybe one or two, but almost nothing for three
00:49:18.920 --> 00:49:22.619
months. But I have turned out some really incredible
00:49:22.619 --> 00:49:27.570
code. And that's because I now have a tool at
00:49:27.570 --> 00:49:30.670
my disposal that lets me do that. But it's like
00:49:30.670 --> 00:49:33.130
a sorcerer's apprentice, right? If you give someone
00:49:33.130 --> 00:49:34.750
who doesn't know what they're doing that tool,
00:49:35.590 --> 00:49:39.130
it's out of control. You actually have to be
00:49:39.130 --> 00:49:41.349
in control of the tool to make it do what you
00:49:41.349 --> 00:49:45.780
want. You would never give a chainsaw to a five
00:49:45.780 --> 00:49:50.900
-year -old. The power of the AI is a bit like
00:49:50.900 --> 00:49:52.539
the power of a chainsaw. It's going to come back
00:49:52.539 --> 00:49:56.880
and bite you if you're not careful. So the reason
00:49:56.880 --> 00:49:59.500
that things like CodeCutter are even more important
00:49:59.500 --> 00:50:05.769
is that You cannot effectively, at least currently,
00:50:06.010 --> 00:50:10.480
you cannot effectively use an AI to do anything,
00:50:10.619 --> 00:50:14.199
actually, if you can't do it yourself, at least
00:50:14.199 --> 00:50:18.389
to some extent. You cannot have an AI help you
00:50:18.389 --> 00:50:21.269
write something unless you know what it is you
00:50:21.269 --> 00:50:23.929
want to say, not if you want it to be good and
00:50:23.929 --> 00:50:28.389
effective. You can't have an AI write code for
00:50:28.389 --> 00:50:31.210
you unless you actually could have written that
00:50:31.210 --> 00:50:35.110
code yourself. and you understand what it is
00:50:35.110 --> 00:50:36.730
it's trying to do and how it's trying to do it
00:50:36.730 --> 00:50:40.760
so you can guide it along the way. And so you
00:50:40.760 --> 00:50:44.630
have to still be able to code if you want to
00:50:44.630 --> 00:50:47.650
use AI to write code. Now, two years from now,
00:50:47.750 --> 00:50:51.260
maybe that won't be true anymore. Two years from
00:50:51.260 --> 00:50:53.280
now, AI might have moved on to the point where
00:50:53.280 --> 00:50:58.789
it is totally capable of having those conversations
00:50:58.789 --> 00:51:01.550
with customers and working out what's needed.
00:51:02.309 --> 00:51:05.849
It's not there yet. You worry about the potential
00:51:05.849 --> 00:51:09.130
feedback loop that it could be creating right
00:51:09.130 --> 00:51:12.949
now. And what I mean by that is you've got people
00:51:12.949 --> 00:51:17.750
that want to use this tool, but they don't understand
00:51:17.750 --> 00:51:20.010
what it is they're making. And then they don't
00:51:20.010 --> 00:51:22.130
bother to check because they think that the AI
00:51:22.130 --> 00:51:25.210
is more than capable and has all the answers.
00:51:25.250 --> 00:51:28.090
So they put in their prompt, it spits back out
00:51:28.090 --> 00:51:31.199
an answer. Maybe they read it over, maybe they
00:51:31.199 --> 00:51:33.260
don't. And then they throw it back out into the
00:51:33.260 --> 00:51:35.699
wild and then more people consume it. And more
00:51:35.699 --> 00:51:38.940
importantly, more AI consumes it. And it's kind
00:51:38.940 --> 00:51:42.280
of, I always say this analogy that it's, it sort
00:51:42.280 --> 00:51:45.900
of reminds me of a VHS copy of a copy of a copy
00:51:45.900 --> 00:51:48.880
of a copy. Eventually you get to the point where
00:51:48.880 --> 00:51:52.699
the copies no longer, you can't watch it because
00:51:52.699 --> 00:51:55.139
it's so bad. And that's what I worry about AI.
00:51:55.219 --> 00:51:58.099
I'm wondering, is that. something that you've
00:51:58.099 --> 00:52:02.380
considered as well, or am I just crazy? Oh, no.
00:52:02.500 --> 00:52:08.519
I think there is so much to worry about in that
00:52:08.519 --> 00:52:12.900
regard. There's the first level effect, and that
00:52:12.900 --> 00:52:18.440
is AIs are really, really great at being confident,
00:52:18.699 --> 00:52:23.599
even when they're wrong. And so we're surrounded
00:52:23.599 --> 00:52:27.119
now by information that AIs are churning out
00:52:27.119 --> 00:52:31.789
that, Most of the time it's true, but quite often
00:52:31.789 --> 00:52:36.610
it's not. And people are not filtering on that
00:52:36.610 --> 00:52:39.710
basis. So everything looks good. If it agrees
00:52:39.710 --> 00:52:42.269
with your biases, then that's the way I'll take
00:52:42.269 --> 00:52:48.690
it. The secondary effect of that is that it's
00:52:48.690 --> 00:52:56.699
getting harder and harder now online to be a
00:52:56.699 --> 00:53:02.599
human and to have your own human output. Because
00:53:02.599 --> 00:53:05.739
you're getting swamped by the sheer volume of
00:53:05.739 --> 00:53:08.719
AI. So if you go on Twitter now, right, I don't
00:53:08.719 --> 00:53:10.300
know what the percentage is, but I would guess
00:53:10.300 --> 00:53:13.679
it's probably 90 % of the posts are AI generated.
00:53:14.539 --> 00:53:21.599
And they are, at best, they are just tedious
00:53:21.599 --> 00:53:25.380
reiteration. And at worst, they are spreading
00:53:25.380 --> 00:53:29.989
misinformation or... uh you know that fake movies
00:53:29.989 --> 00:53:32.190
that make things look weird it's all that kind
00:53:32.190 --> 00:53:34.730
of stuff it just makes it so hard now to be a
00:53:34.730 --> 00:53:39.670
human and actually get noticed um my guess is
00:53:39.670 --> 00:53:42.809
that at some point in the next couple years there'll
00:53:42.809 --> 00:53:46.469
be some kind of breakout of you know coming back
00:53:46.469 --> 00:53:49.730
to being a person away from away from all these
00:53:49.730 --> 00:53:53.349
online machine generated things on that line
00:53:53.349 --> 00:53:56.940
i think It's also very difficult to tell now
00:53:56.940 --> 00:53:59.599
what is AI generated and what isn't. I mean,
00:53:59.619 --> 00:54:02.980
there's some tells in images, in posts, but sometimes
00:54:02.980 --> 00:54:04.980
you look at an image and you're like, is it really
00:54:04.980 --> 00:54:08.579
AI or is it not? It also is making it, to your
00:54:08.579 --> 00:54:10.139
point, very difficult to be a human in the sense
00:54:10.139 --> 00:54:12.960
of what's the incentive of writing a blog post?
00:54:13.139 --> 00:54:15.139
What's the incentive of writing a book if it's
00:54:15.139 --> 00:54:17.300
ultimately going to end up just being used by
00:54:17.300 --> 00:54:19.820
AI? No one's really going to buy it or consume
00:54:19.820 --> 00:54:22.530
it, and it's just... not the juice is not worth
00:54:22.530 --> 00:54:25.489
the squeeze but it used to be right like the
00:54:25.489 --> 00:54:27.829
same thing they say why why get a designer when
00:54:27.829 --> 00:54:29.530
you can have chat GPT just do something like
00:54:29.530 --> 00:54:32.550
Studio Ghibli and I wanted to ask how has that
00:54:32.550 --> 00:54:36.010
impacted the way you do work because you know
00:54:36.010 --> 00:54:38.670
10 -15 years ago and I started doing iOS programming
00:54:38.670 --> 00:54:41.360
and have a platform coding and more coding stuff
00:54:41.360 --> 00:54:43.699
in general books are very prominent i mean the
00:54:43.699 --> 00:54:46.360
pragmatic bookshelf uh a press books a bunch
00:54:46.360 --> 00:54:48.360
of publications were like very very reputable
00:54:48.360 --> 00:54:50.840
and you had tremendously talented individuals
00:54:50.840 --> 00:54:53.440
doing it throughout the years it's been diluted
00:54:53.440 --> 00:54:56.679
you know publications not the product bookshelf
00:54:56.679 --> 00:54:58.719
i think you've kept your quality very high but
00:54:58.719 --> 00:55:00.860
others just diluted that and find anyone who's
00:55:00.860 --> 00:55:02.460
willing to write a book because there's you know
00:55:02.460 --> 00:55:06.099
competition and such but how is business shifting
00:55:06.099 --> 00:55:08.460
with AI and with some of the things that you
00:55:08.460 --> 00:55:10.300
mentioned where it's harder to be a human. And
00:55:10.300 --> 00:55:12.599
to be honest, folks, you're just shifting to
00:55:12.599 --> 00:55:15.059
the model of, hey, Gemini, what's this? What's
00:55:15.059 --> 00:55:17.159
that? How would I learn this? As opposed to taking
00:55:17.159 --> 00:55:19.679
the time to read a book, unless it's a fictitious
00:55:19.679 --> 00:55:23.159
novel, a nice new sci -fi book from the author
00:55:23.159 --> 00:55:25.980
of Project Hail Mary kind of deal. Yeah, I think
00:55:25.980 --> 00:55:29.440
there's two aspects to that, both of which kind
00:55:29.440 --> 00:55:34.269
of amplify each other. You ask about like, how's
00:55:34.269 --> 00:55:37.789
it affected? Like writing books. Well, Tammy
00:55:37.789 --> 00:55:39.789
can tell you. I mean, basically the pragmatic
00:55:39.789 --> 00:55:45.639
bookshelf has shut down pretty much because we
00:55:45.670 --> 00:55:48.809
had, and it's not just us, we actually are doing
00:55:48.809 --> 00:55:53.050
better than many. But over year on year, sales
00:55:53.050 --> 00:55:57.650
are down about 30 % compound because people are
00:55:57.650 --> 00:56:00.230
using AI or largely because people are using
00:56:00.230 --> 00:56:06.670
AI. And so as a result. I don't know what that's
00:56:06.670 --> 00:56:09.530
going to do long term, but I can't help feeling
00:56:09.530 --> 00:56:13.150
that part of the thing about writing a book is
00:56:13.150 --> 00:56:16.630
not just generating a bunch of facts, but it's
00:56:16.630 --> 00:56:19.550
expressing a position, expressing a way of doing
00:56:19.550 --> 00:56:21.670
something, a way of thinking about something.
00:56:22.010 --> 00:56:25.489
And you lose that. But the other scary thing
00:56:25.489 --> 00:56:29.190
is nothing to do with AI. And that is the fact,
00:56:29.230 --> 00:56:31.050
I don't know what it's like where you are, but
00:56:31.050 --> 00:56:34.889
here, I can't remember what the number is, but
00:56:34.889 --> 00:56:37.889
it's like. some ridiculously small percentages
00:56:37.889 --> 00:56:41.969
of seniors in high school can actually read.
00:56:42.130 --> 00:56:48.269
I mean, they have trouble reading more than like
00:56:48.269 --> 00:56:51.269
two sentences. They have trouble keeping attention
00:56:51.269 --> 00:56:56.909
when they read anything. And so long -form writing
00:56:56.909 --> 00:57:04.449
is being replaced by TikTok because kids... Just
00:57:04.449 --> 00:57:06.570
don't. I mean, if you even look at TikTok and
00:57:06.570 --> 00:57:08.570
when they put the captions up, it's not like
00:57:08.570 --> 00:57:11.050
a little paragraph of text, is it? It's one or
00:57:11.050 --> 00:57:12.809
two words at a time. Bang, bang, bang, bang,
00:57:12.829 --> 00:57:18.300
bang. And that's the way they read it. So that's
00:57:18.369 --> 00:57:20.929
the other reason that books are slowly falling
00:57:20.929 --> 00:57:25.980
out of people's consciousness. It's just the
00:57:25.980 --> 00:57:29.360
act of being able to sit down and read a thousand
00:57:29.360 --> 00:57:35.489
words has been lost. At least in the West. I
00:57:35.489 --> 00:57:37.110
don't think that's true in many other places,
00:57:37.130 --> 00:57:40.789
but in the US and probably most of Europe, I
00:57:40.789 --> 00:57:42.590
think that's the case. What do you think was
00:57:42.590 --> 00:57:46.050
the driving force behind that? I could say it
00:57:46.050 --> 00:57:47.989
must have started when they stopped teaching
00:57:47.989 --> 00:57:52.010
cursive, but I think it's a lot deeper than that.
00:57:52.670 --> 00:57:55.769
Yes, not teaching cursive was ridiculous because
00:57:56.059 --> 00:57:58.659
actually, you know what? I think that actually
00:57:58.659 --> 00:58:01.920
had something to do with it because cursive was
00:58:01.920 --> 00:58:04.400
a way of increasing the bandwidth. with which
00:58:04.400 --> 00:58:08.269
you could write. And when you go back to printing,
00:58:08.530 --> 00:58:11.789
which is what most people nowadays do, you're
00:58:11.789 --> 00:58:14.929
slow. And why would you want to be slow? So you
00:58:14.929 --> 00:58:16.570
go to your phone and use your thumbs instead.
00:58:18.079 --> 00:58:20.679
If I was given the time machine and said, okay,
00:58:20.679 --> 00:58:22.639
you can go back and change one thing in history,
00:58:22.900 --> 00:58:28.619
I would go and probably kill Facebook and or
00:58:28.619 --> 00:58:32.690
Twitter as the two things which I think have
00:58:32.690 --> 00:58:36.400
done the most damage to the next generation.
00:58:37.829 --> 00:58:41.030
Enjoying the show? Good. We want to keep it that
00:58:41.030 --> 00:58:44.659
way. If you have a guest suggestion, a story
00:58:44.659 --> 00:58:47.659
to share, or just want to get in touch, send
00:58:47.659 --> 00:58:51.300
a message to interviews at roundaboutcreativechaos
00:58:51.300 --> 00:58:57.889
.com Now, where were we? I want to pick into
00:58:57.889 --> 00:59:02.230
your brain with 13 questions that are based off
00:59:02.230 --> 00:59:06.260
of Marcel Proust. Felipe has not been a co -host
00:59:06.260 --> 00:59:09.400
on this show for many, many, many years. In fact,
00:59:09.460 --> 00:59:12.719
tonight is his first night back after all of
00:59:12.719 --> 00:59:15.960
those years. He used to do the 13 questions.
00:59:16.420 --> 00:59:21.699
And if he's up for it and has access to our shared
00:59:21.699 --> 00:59:24.599
folder, Felipe, you want to take the questions?
00:59:25.420 --> 00:59:27.679
So for those listening who maybe knew, who maybe
00:59:27.679 --> 00:59:30.800
haven't seen me in, I think, about a decade in
00:59:30.800 --> 00:59:36.420
Roundabout. Marcel Proust asked himself these
00:59:36.420 --> 00:59:40.239
questions twice. Once when he was 13, once at
00:59:40.239 --> 00:59:43.139
20. And the goal was not to see, you know, there's
00:59:43.139 --> 00:59:45.679
no right or wrong. It's really just the way,
00:59:45.739 --> 00:59:48.019
it was a way for him to track how much he had
00:59:48.019 --> 00:59:50.719
changed and how much he hadn't, right? It's not
00:59:50.719 --> 00:59:53.059
about work. It's not about being ambitious or
00:59:53.059 --> 00:59:56.440
growing and getting better. It's really more
00:59:56.440 --> 00:59:59.679
about fear, pleasure, loyalty, or potentially
00:59:59.679 --> 01:00:02.190
regret. And so that's the idea kind of behind
01:00:02.190 --> 01:00:06.650
our questions. And maybe this will help you kind
01:00:06.650 --> 01:00:09.530
of get some of those feelings, Dave. But maybe,
01:00:09.570 --> 01:00:13.849
you know, we have you back when you're 20. And
01:00:13.849 --> 01:00:17.530
you can look back at when you were 13 and say,
01:00:17.670 --> 01:00:20.050
oh, this is how much I've changed. And this is
01:00:20.050 --> 01:00:23.530
how much the world has changed with me. So getting
01:00:23.530 --> 01:00:26.010
started with things. Question number one is,
01:00:26.030 --> 01:00:29.179
what is your favorite word? My favorite. Okay.
01:00:29.239 --> 01:00:32.239
Now this is a question where I could answer in
01:00:32.239 --> 01:00:35.420
many different ways because I am, and this is
01:00:35.420 --> 01:00:37.019
going to ruin your entire list of questions.
01:00:37.119 --> 01:00:40.460
I am someone who hates the concept of favorites,
01:00:40.639 --> 01:00:43.659
right? What's your favorite album? What's your
01:00:43.659 --> 01:00:47.199
favorite whatever, right? Because it always depends
01:00:47.199 --> 01:00:49.880
on context, right? There are some times I want
01:00:49.880 --> 01:00:51.460
to listen to something that I listened to when
01:00:51.460 --> 01:00:53.519
I was a kid. There's sometimes I want to listen
01:00:53.519 --> 01:00:55.679
to something brand new and it's always contextual.
01:00:56.840 --> 01:01:00.920
So favorite word, I think there are many reasons
01:01:00.920 --> 01:01:04.480
for a word to be favorite. So I think, for example,
01:01:04.599 --> 01:01:10.780
if you want a word that just feels like a really,
01:01:10.820 --> 01:01:14.239
really good word, it would probably be a word
01:01:14.239 --> 01:01:20.880
like lugubrious or succulent. What does lugubrious
01:01:20.880 --> 01:01:24.550
mean? Lugubrious. Oh, go look it up. It's a great
01:01:24.550 --> 01:01:28.889
word. But I think if you want... The actual word,
01:01:28.969 --> 01:01:33.320
which I think is the best word, or the, yeah,
01:01:33.380 --> 01:01:39.199
the best word there is, it's why. W -H -Y. Ligubrious,
01:01:39.219 --> 01:01:42.079
according to the AI overview, because I forgot
01:01:42.079 --> 01:01:46.579
to hit no AI, Ligubrious is an adjective that
01:01:46.579 --> 01:01:51.119
means extremely mournful, dismal, or gloomy,
01:01:51.199 --> 01:01:55.019
often in an exaggerated or heavy way. Exactly,
01:01:55.119 --> 01:01:59.150
yeah. I love it. Overly expressed. Oh, my gosh.
01:01:59.170 --> 01:02:01.510
I'm going to find a way to use that word. See,
01:02:01.610 --> 01:02:03.530
this is good for me. I don't know about the listeners
01:02:03.530 --> 01:02:06.269
or for you, too, but I'm having a good old time.
01:02:09.070 --> 01:02:12.449
But I love your answer. I love your answer. Now,
01:02:12.670 --> 01:02:15.949
and again, this when we say favorite, I love
01:02:15.949 --> 01:02:17.869
how you interpret it the way you interpret it
01:02:17.869 --> 01:02:22.230
and why is very insightful. And I kind of get
01:02:22.230 --> 01:02:24.289
a sense of where you're coming from. So it doesn't
01:02:24.289 --> 01:02:25.690
necessarily have to be your favorite, but maybe
01:02:25.690 --> 01:02:27.829
it's like. the one that's been on your mind most
01:02:27.829 --> 01:02:29.630
recently, or the one that stuck with you, or
01:02:29.630 --> 01:02:31.110
the one that, like you said, you're in the mood
01:02:31.110 --> 01:02:33.769
of. In that same vein and being antagonistic,
01:02:34.030 --> 01:02:38.780
what is your least favorite word? Yes. Because
01:02:38.780 --> 01:02:44.380
yes is the laziest word I know. Which would you
01:02:44.380 --> 01:02:47.980
say you enjoy or prefer the most? A good book,
01:02:48.119 --> 01:02:51.440
a good movie, listening to some old records,
01:02:51.639 --> 01:02:53.699
or maybe a game. It can be a video game, a board
01:02:53.699 --> 01:02:58.849
game. Not the game. I'm not a big game player.
01:03:00.110 --> 01:03:04.980
Again, it depends on the mood I'm in. I read
01:03:04.980 --> 01:03:11.119
a lot. I really enjoy reading. I get lost in
01:03:11.119 --> 01:03:14.460
books in a way I don't get lost. No, that's not
01:03:14.460 --> 01:03:16.780
quite true. I get lost in books. I also get lost
01:03:16.780 --> 01:03:19.480
in movies, but in a very different way. Movies,
01:03:19.480 --> 01:03:23.949
I get lost in almost like a zen -like, you know,
01:03:23.989 --> 01:03:25.829
you kind of like switch stuff off and it just
01:03:25.829 --> 01:03:29.599
goes into you. with a book, it's more active.
01:03:29.820 --> 01:03:36.900
And so I get, it's more a conversation or a dialogue
01:03:36.900 --> 01:03:41.079
with a book than it is with a movie. Music, I'd
01:03:41.079 --> 01:03:45.260
love music. I don't know very much about it,
01:03:45.340 --> 01:03:51.820
but I just, if I'm on my own, I'll typically
01:03:51.820 --> 01:03:55.619
have music on and who knows what will be playing.
01:03:56.690 --> 01:03:59.610
Who would be your favorite fictional character?
01:04:00.130 --> 01:04:07.239
I'm torn. Maybe Dr. Watson, because he was the
01:04:07.239 --> 01:04:15.130
everyman, but he was smart and he was a foil
01:04:15.130 --> 01:04:18.389
without which Holmes couldn't have been brilliant.
01:04:18.610 --> 01:04:21.670
So he was like, you know, he was a practical
01:04:21.670 --> 01:04:24.389
person. He was a doctor. He served in the army.
01:04:25.769 --> 01:04:28.070
You know, he carried a gun around if he needed
01:04:28.070 --> 01:04:31.150
to. He'd ask stupid questions when he needed
01:04:31.150 --> 01:04:34.909
to. And I think that's actually a role that I,
01:04:35.030 --> 01:04:38.320
you know, without the gun, that's a role that
01:04:38.320 --> 01:04:40.719
I quite enjoy as well. So I think there's that
01:04:40.719 --> 01:04:43.099
aspect to it. There's a whole bunch I really,
01:04:43.119 --> 01:04:45.659
really like, but none of them are kind of like,
01:04:45.760 --> 01:04:50.059
you know. Wow. Interesting, but not memorable
01:04:50.059 --> 01:04:53.130
or not necessarily up there with. Dr. Watson.
01:04:53.210 --> 01:04:54.969
Well, I've also got to tell you, I actually have
01:04:54.969 --> 01:04:58.469
a really, really, really bad recall memory. I
01:04:58.469 --> 01:05:00.869
mean, like clinically bad. I've got a part of
01:05:00.869 --> 01:05:03.750
my brain missing, which is good at doing the,
01:05:03.750 --> 01:05:07.210
you know, what were you reading on the 15th type
01:05:07.210 --> 01:05:10.429
question? I can't do those. And so quite often,
01:05:10.510 --> 01:05:14.190
the only way I can arrive at an answer like that
01:05:14.190 --> 01:05:17.030
is to allow my associative memory just to run
01:05:17.030 --> 01:05:19.289
at it for a couple of minutes. And then it just
01:05:19.289 --> 01:05:22.610
suddenly pops out the other end. Question number
01:05:22.610 --> 01:05:24.250
five is, where is the most interesting place
01:05:24.250 --> 01:05:29.110
you've been to? I think there are a couple of
01:05:29.110 --> 01:05:31.750
answers, depending on how virtuous I'm feeling.
01:05:34.210 --> 01:05:41.170
Possibly the V &A Museum in London. It is interesting
01:05:41.170 --> 01:05:47.590
because it tells the story of an empire through
01:05:47.969 --> 01:05:56.519
artifacts. You can watch how things changed over
01:05:56.519 --> 01:06:02.619
the centuries for that. So that's a place where
01:06:02.619 --> 01:06:05.400
it's really nice to go and just, you find an
01:06:05.400 --> 01:06:07.599
exhibit room and just sit in it for half an hour.
01:06:08.260 --> 01:06:14.719
That's nice. Another place I could spend a lot
01:06:14.719 --> 01:06:20.300
of time in is the Mary Rose in South Hampton
01:06:20.300 --> 01:06:24.730
or Portsmouth, I can't remember. it's uh the
01:06:24.730 --> 01:06:28.230
ship it was one of king henry's ships sank it
01:06:28.230 --> 01:06:32.119
was preserved or half preserved in mud and they
01:06:32.119 --> 01:06:37.679
uh dug it up basically and they went through
01:06:37.679 --> 01:06:40.719
like a 20 -year restoration where they had to
01:06:40.719 --> 01:06:44.219
replace the salt water with some preservative
01:06:44.219 --> 01:06:48.800
and now it's on display like upright and they
01:06:48.800 --> 01:06:53.449
put figures and video displays and stuff in it
01:06:53.449 --> 01:06:55.630
so you can see what people went in which rooms
01:06:55.630 --> 01:07:00.159
and how it all worked and you look at the uh
01:07:00.159 --> 01:07:04.219
the history and you say that uh the british and
01:07:04.219 --> 01:07:08.340
the dutch basically shaped the world in the 17th
01:07:08.340 --> 01:07:13.250
18th century through sailing and it's kind of
01:07:13.250 --> 01:07:16.650
like it paints a picture of you know the captains
01:07:16.650 --> 01:07:18.809
or in the fancy frock coats and everything else
01:07:18.809 --> 01:07:21.110
but you actually look at the ship and you realize
01:07:21.110 --> 01:07:25.510
how how miserable and primitive the conditions
01:07:25.510 --> 01:07:29.789
were in these things. It's stunning when you
01:07:29.789 --> 01:07:36.570
think about how that changed humanity totally.
01:07:37.329 --> 01:07:40.460
So that's really interesting. And I'm sure I'm
01:07:40.460 --> 01:07:42.619
missing out on the really, really, really interesting
01:07:42.619 --> 01:07:45.449
things, but that's what springs to mind. Who
01:07:45.449 --> 01:07:47.809
would you say is the most or one of the most
01:07:47.809 --> 01:07:50.530
interesting people you've met? Could have been
01:07:50.530 --> 01:07:55.710
a mentor, a co -worker, relative? I think probably
01:07:55.900 --> 01:08:00.099
my best friend at high school, I would count
01:08:00.099 --> 01:08:06.469
for that, Toby Bailey, who he and I went from,
01:08:06.469 --> 01:08:09.030
I don't know how it would be counted over here,
01:08:09.110 --> 01:08:13.139
age 12 up to graduation at high school. And we
01:08:13.139 --> 01:08:16.779
were kind of inseparable. And we thought very
01:08:16.779 --> 01:08:21.199
much alike. We enjoyed the same things, but from
01:08:21.199 --> 01:08:25.380
a different perspective. He was phenomenally
01:08:25.539 --> 01:08:30.060
good at mathematics. Went on to do quantum gravity
01:08:30.060 --> 01:08:34.380
at Cambridge, where I was the more practical
01:08:34.380 --> 01:08:38.340
one. So we had this different viewpoint, but
01:08:38.340 --> 01:08:42.159
we kind of like energized each other in that
01:08:42.159 --> 01:08:47.000
way. And so that, he always had interesting things
01:08:47.000 --> 01:08:49.300
to think about or to say that made me think.
01:08:49.460 --> 01:08:53.289
He always had interesting viewpoints. So probably
01:08:53.289 --> 01:08:57.750
him. What is one thing that you own that cannot
01:08:57.750 --> 01:09:01.329
be replaced? So in my head, I've reframed that
01:09:01.329 --> 01:09:04.069
as what would I grab if the house was on fire?
01:09:05.189 --> 01:09:08.680
Exactly the same. Yeah. Well, not necessarily
01:09:08.680 --> 01:09:12.220
because you can own things like a house that
01:09:12.220 --> 01:09:17.380
you can't take with you. But in this case, I
01:09:17.380 --> 01:09:20.739
would say probably, okay, this is going to sound
01:09:20.739 --> 01:09:25.399
a little bit schlocky, but. There is a photograph
01:09:25.399 --> 01:09:27.840
I have on the windowsill in my office, which
01:09:27.840 --> 01:09:31.000
has been beaten to death because it's been in
01:09:31.000 --> 01:09:33.119
a wallet and in pockets and everything else.
01:09:33.199 --> 01:09:36.770
So it's all frayed at the edges. But it's a picture
01:09:36.770 --> 01:09:42.079
when my second son was born of my first son,
01:09:42.159 --> 01:09:45.970
who was like four or five years older, sitting
01:09:45.970 --> 01:09:48.310
holding him. I think it was in the hospital.
01:09:50.270 --> 01:09:56.270
And looking just like kind of... scared and excited
01:09:56.270 --> 01:10:02.399
at the same time. This was before we told him,
01:10:02.439 --> 01:10:05.020
no, he wasn't the third parent, which upset him
01:10:05.020 --> 01:10:08.720
quite a lot. But I think that picture for me
01:10:08.720 --> 01:10:14.140
is, I don't know, that resonates. In the last
01:10:14.140 --> 01:10:17.600
year, is there anything that you can recall that
01:10:17.600 --> 01:10:20.909
you have changed your mind about? As an example,
01:10:20.949 --> 01:10:23.890
we've discussed AI and maybe a year ago you hated
01:10:23.890 --> 01:10:26.310
it. And now you said you haven't written more
01:10:26.310 --> 01:10:29.810
than a dozen lines of code per week or per month
01:10:29.810 --> 01:10:32.050
in three months. Something like that that you
01:10:32.050 --> 01:10:34.569
can remember that you've changed your mind? Well,
01:10:34.689 --> 01:10:36.430
I think, I mean, yeah, I mean, we've already
01:10:36.430 --> 01:10:38.550
talked about AI. That's definitely one of those
01:10:38.550 --> 01:10:42.979
things. I don't know if it's changed my mind
01:10:42.979 --> 01:10:52.130
or whether it's changed my values, but the I'm
01:10:52.130 --> 01:10:54.590
not going to say failure because it's not failed,
01:10:54.729 --> 01:11:01.159
but the cutting back of the bookshelf hit me
01:11:01.159 --> 01:11:07.069
really very hard. And a lot of that was because
01:11:07.069 --> 01:11:12.229
I felt there was a lot of myself in it. And so
01:11:12.229 --> 01:11:18.369
it was my failure, which, I mean, honestly, objectively,
01:11:18.550 --> 01:11:22.119
it probably was. Even so, it was like I was judging
01:11:22.119 --> 01:11:27.359
myself by it. And in the last year, I have learned
01:11:27.359 --> 01:11:30.500
to cut away from that. And I've changed my mind
01:11:30.500 --> 01:11:34.680
about that. And to view that as more of an objective
01:11:34.680 --> 01:11:39.420
thing rather than a subjective one. I think that's
01:11:39.420 --> 01:11:41.800
really positive. If there's anything I can say
01:11:41.800 --> 01:11:45.300
about that from a personal opinion or just, I
01:11:45.300 --> 01:11:47.220
don't want to use the word advice, but it's think
01:11:47.220 --> 01:11:49.300
of it not as failure, but rather as a cycle,
01:11:49.399 --> 01:11:52.350
right? The cycle of a plant, the circle of life,
01:11:52.489 --> 01:11:56.109
Simba Mufasa kind of deal. I think it fulfilled
01:11:56.109 --> 01:11:59.289
its cycle when it needed to in what I would consider
01:11:59.289 --> 01:12:02.930
incredible, with incredible success. And it's
01:12:02.930 --> 01:12:04.630
not due to failure that you had to scale back
01:12:04.630 --> 01:12:06.630
and maybe get close to shutting down, but rather
01:12:06.630 --> 01:12:09.649
just it fulfilled its cycle at the time it needed
01:12:09.649 --> 01:12:12.770
to. And then, hey, just as a door closes, they
01:12:12.770 --> 01:12:16.369
say a door opens. I do personally think, especially
01:12:16.369 --> 01:12:18.590
after having talked to you now, we still have
01:12:18.590 --> 01:12:20.369
a couple of questions to go that you still have
01:12:20.369 --> 01:12:22.829
a ton to offer to us, to our industry, and to
01:12:22.829 --> 01:12:25.050
those who've been big fans and have used resources
01:12:25.050 --> 01:12:28.670
like the bookshelf to learn. So I thank you for
01:12:28.670 --> 01:12:32.750
that very candid and honest answer. If a bit
01:12:32.750 --> 01:12:37.550
lugubrious, which is completely understandable.
01:12:38.770 --> 01:12:43.619
Now, Tammy asked you about your first job. which
01:12:43.619 --> 01:12:45.520
would you say is the worst job you've ever had?
01:12:45.779 --> 01:12:48.380
If any, but maybe if not the worst, just the
01:12:48.380 --> 01:12:51.920
not so great one amongst the greats. Oh, that's
01:12:51.920 --> 01:12:55.399
easy. When I was, I guess I'd have been about
01:12:55.399 --> 01:13:03.640
15. I got a job in a local supermarket and I
01:13:03.640 --> 01:13:06.680
was basically doing warehouse -y kind of stuff.
01:13:08.479 --> 01:13:14.210
And it was... For me, it was kind of like my
01:13:14.210 --> 01:13:16.869
introduction to the world of, or to the real
01:13:16.869 --> 01:13:22.100
world, in a way, of people working. At home,
01:13:22.100 --> 01:13:27.060
we didn't swear that much, if at all. And suddenly
01:13:27.060 --> 01:13:29.119
I'm just like dropped into this warehouse full
01:13:29.119 --> 01:13:36.630
of people swearing like sailors. It was very
01:13:37.079 --> 01:13:43.140
much, it was very rough in that way. But at the
01:13:43.140 --> 01:13:45.119
same time, it was decent money and I kind of
01:13:45.119 --> 01:13:48.170
enjoyed just the mechanical stuff. And then at
01:13:48.170 --> 01:13:52.109
one point, I get not invited, I get told that
01:13:52.210 --> 01:13:56.069
I have to go help out in the delicatessen section.
01:13:57.710 --> 01:13:59.270
Well, it's not really, no, it's actually more
01:13:59.270 --> 01:14:02.750
like the butcher section. It had some cold meats,
01:14:02.810 --> 01:14:07.680
but it was mostly just the cut meat. And I have
01:14:07.680 --> 01:14:12.279
this thing about knives. I really, it's one of
01:14:12.279 --> 01:14:15.739
my big, my big, phobia, fear things is knives.
01:14:16.420 --> 01:14:20.600
And I'm given the job every week of dissecting
01:14:20.600 --> 01:14:26.319
a side of pork into the various cuts, which I
01:14:26.539 --> 01:14:30.239
did through sheer willpower. And so, yeah, that
01:14:30.239 --> 01:14:35.819
was probably my least favorite job. But I can
01:14:35.819 --> 01:14:42.130
still dissect a pig if you want me to. Hey, when
01:14:42.130 --> 01:14:45.359
we get to question 13... That may be a tremendously
01:14:45.359 --> 01:14:48.479
useful skill to have one way or another. Okay.
01:14:49.020 --> 01:14:51.880
Moving on to number 10 and just three away from
01:14:51.880 --> 01:14:55.640
13. So just hold that thought. What is your greatest
01:14:55.640 --> 01:14:59.130
achievement? This is going to sound like I'm
01:14:59.130 --> 01:15:04.609
dodging the question. I think my biggest professional,
01:15:05.079 --> 01:15:07.760
well, the thing I'm happiest about professionally
01:15:07.760 --> 01:15:14.460
is I think I have, I've had an impact. that's
01:15:14.460 --> 01:15:18.560
made a decent number of people's lives just a
01:15:18.560 --> 01:15:23.930
little bit better. So I think that one of the
01:15:23.930 --> 01:15:25.770
scariest things that ever happened to me at a
01:15:25.770 --> 01:15:30.439
conference is when people come up to me. They
01:15:30.439 --> 01:15:34.500
actually come up surprisingly often. People come
01:15:34.500 --> 01:15:38.000
up with me and they say, I liked your talk and
01:15:38.000 --> 01:15:40.119
I listened to it, so now I'm going to go quit
01:15:40.119 --> 01:15:46.899
and change my job. Often they'll write back to
01:15:46.899 --> 01:15:49.060
me and say, yeah, and I got this new job that
01:15:49.060 --> 01:15:51.319
was more likely you were talking about and it's
01:15:51.319 --> 01:15:53.140
so much better and I really enjoy it and all
01:15:53.140 --> 01:15:56.420
this kind of stuff. So I think probably the thing
01:15:56.420 --> 01:16:01.750
I'm quietly happiest about is that, having had
01:16:01.750 --> 01:16:04.770
the opportunity to be able to do that. The flip
01:16:04.770 --> 01:16:07.789
side is, what would you say is your biggest regret?
01:16:09.399 --> 01:16:14.329
Again, professionally, it would probably be...
01:16:14.649 --> 01:16:22.689
I retired from the bookshelf in 2016. It was
01:16:22.689 --> 01:16:28.760
getting to be unpleasant, and I wanted to go
01:16:28.760 --> 01:16:31.779
in one direction, and most of the other people
01:16:31.779 --> 01:16:34.420
wanted to go in another direction. And rather
01:16:34.420 --> 01:16:37.520
than stand and fight, I said, oh, screw it, and
01:16:37.520 --> 01:16:41.800
walked away. And I think probably that's my biggest,
01:16:42.039 --> 01:16:46.869
because if I had sat... and stood my ground and
01:16:46.899 --> 01:16:51.649
gone the way I wanted to go, I think we would
01:16:51.649 --> 01:16:54.670
be in a lot better place today. Second to last
01:16:54.670 --> 01:16:56.810
question before we get back to those very useful
01:16:56.810 --> 01:16:59.050
skills acquired at your job, your first job,
01:16:59.090 --> 01:17:01.949
I think, at 15 at the supermarket. What do you
01:17:01.949 --> 01:17:05.229
fear most? Many different levels, but I guess
01:17:05.390 --> 01:17:09.210
existentially, if I'm looking at it from a kind
01:17:09.210 --> 01:17:14.390
of species point of view, intolerance. I think
01:17:14.390 --> 01:17:18.729
we have a... A society that has grown polarized
01:17:19.140 --> 01:17:25.279
and that refuses to talk or listen to people
01:17:25.279 --> 01:17:30.670
who don't share their views. It's not a political
01:17:30.670 --> 01:17:33.789
thing. It happens everywhere. Left, right, center,
01:17:34.170 --> 01:17:39.560
gay, straight, it doesn't matter. And we've become
01:17:39.560 --> 01:17:46.640
balkanized and we have become angry. That is
01:17:46.640 --> 01:17:49.260
not a good place for an entire world population
01:17:49.260 --> 01:17:53.760
to be. What would Dave Thomas do in a zombie
01:17:53.760 --> 01:17:57.939
apocalypse? And who would you want by your side?
01:17:58.539 --> 01:18:04.079
I would not enjoy a zombie apocalypse. There
01:18:04.079 --> 01:18:08.989
are some people that would. I would not. I honestly
01:18:08.989 --> 01:18:12.569
can't see myself surviving past day half of a
01:18:12.569 --> 01:18:15.729
zombie apocalypse. Even with your pig dissection
01:18:15.729 --> 01:18:19.210
skills? Come on, Dave. Well, I could throw pork
01:18:19.210 --> 01:18:20.909
chops at them, but that's not really going to
01:18:20.909 --> 01:18:29.600
help, you know? No, I don't know. I am not a
01:18:29.600 --> 01:18:34.020
get -out -there -with -a -machete kind of person.
01:18:34.479 --> 01:18:40.800
However, I have a son who would be very handy
01:18:40.800 --> 01:18:42.859
in those circumstances. So if we could get him
01:18:42.859 --> 01:18:45.810
back home again. I'm sure he would protect the
01:18:45.810 --> 01:18:49.590
entire family from any kind of zombie incursion.
01:18:51.130 --> 01:18:55.649
Yeah, I don't know. I don't know. I think I would
01:18:55.649 --> 01:18:58.189
be exactly the same as everybody else. And that
01:18:58.189 --> 01:19:01.329
is you do what you have to for as long as you
01:19:01.329 --> 01:19:04.569
can. I could tell you what not to do. Don't run
01:19:04.569 --> 01:19:08.189
to your nearest Home Depot or big box store and
01:19:08.189 --> 01:19:10.390
try and get all the stuff from there because
01:19:10.390 --> 01:19:12.550
it's going to be picked over. And if it's not
01:19:12.550 --> 01:19:14.880
picked over, you're going to run into some. Pretty
01:19:14.880 --> 01:19:19.000
bad folks that are probably not going to be who
01:19:19.000 --> 01:19:20.899
you want to meet up with at the end of the world.
01:19:21.880 --> 01:19:25.039
So we built a house next door to the house I'm
01:19:25.039 --> 01:19:30.340
currently sitting in. And it has a concrete foundation,
01:19:30.640 --> 01:19:34.960
basically a cellar, which in Texas is very unusual.
01:19:35.770 --> 01:19:39.569
And the person who poured it actually misread
01:19:39.569 --> 01:19:43.619
the plan and initially poured what was supposed
01:19:43.619 --> 01:19:46.989
to be just like... some support columns for a
01:19:46.989 --> 01:19:50.229
deck. He thought it was a solid wall. And so
01:19:50.229 --> 01:19:53.710
we have in the basement of this house, we have
01:19:53.710 --> 01:19:56.329
the regular basement room. And then we made them
01:19:56.329 --> 01:20:00.250
cut a door into this other space, which is, I
01:20:00.250 --> 01:20:03.649
don't know, four feet wide. And it runs the length
01:20:03.649 --> 01:20:07.979
of the house. And we call that our zombie apocalypse
01:20:07.979 --> 01:20:12.550
cellar. Because you can go down there and you
01:20:12.550 --> 01:20:15.270
can barricade yourself in and you'd have your
01:20:15.270 --> 01:20:17.210
supplies and everything else. And you could just
01:20:17.210 --> 01:20:19.149
basically wait there surrounded by 10 inches
01:20:19.149 --> 01:20:23.880
of concrete until, you know, help arrived. And
01:20:23.880 --> 01:20:25.720
obviously help would arrive because that's a
01:20:25.720 --> 01:20:27.659
movie, right? And there would have to be some
01:20:27.659 --> 01:20:32.020
kind of help. I now have this image of you and
01:20:32.020 --> 01:20:34.520
your family sitting in this zombie apocalypse
01:20:34.520 --> 01:20:38.449
shelter underneath your house. surrounded by
01:20:38.449 --> 01:20:42.189
half -dissected pig carcasses. There you go.
01:20:42.409 --> 01:20:44.850
This is not a good image to go to bed on, but
01:20:44.850 --> 01:20:48.069
at least you'll be alive. Well, Felipe would
01:20:48.069 --> 01:20:50.369
say it would be half -dissected gerbil carcasses,
01:20:50.449 --> 01:20:54.210
presumably. There you go. Dave, on my behalf,
01:20:54.369 --> 01:20:55.869
I just want to thank you because I was really
01:20:55.869 --> 01:20:57.489
looking forward to this episode, and I specifically
01:20:57.489 --> 01:21:00.430
told Tammy, before you wrap up Season 13, I want
01:21:00.430 --> 01:21:02.930
to come back for the Dave episode because he
01:21:02.930 --> 01:21:06.050
is someone that I think I would love. Talking
01:21:06.050 --> 01:21:08.470
to, listening to. So I want to thank you so much
01:21:08.470 --> 01:21:10.369
for your time. And I'll kick it back to you,
01:21:10.430 --> 01:21:14.350
Tammy, to wrap us up. Yeah, Dave, I didn't tell
01:21:14.350 --> 01:21:17.210
you this, but you are officially our last guest
01:21:17.210 --> 01:21:20.449
of season 13. And I couldn't think of anybody
01:21:20.449 --> 01:21:23.750
else who I would want to wrap up this season
01:21:23.750 --> 01:21:26.250
with beside you. It's been a while since we've
01:21:26.250 --> 01:21:31.029
talked. And I know that things have shifted tremendously.
01:21:31.970 --> 01:21:34.689
in the publishing world. And we've all felt it.
01:21:34.729 --> 01:21:37.590
We've all felt the absences and the changes and
01:21:37.590 --> 01:21:43.310
the frustration. But I think that overall, you
01:21:43.310 --> 01:21:46.130
know, time marches on, things keep happening.
01:21:46.189 --> 01:21:49.409
And I really can't wait to see what happens next,
01:21:49.449 --> 01:21:52.289
which leads me to my very last question for you,
01:21:52.369 --> 01:21:55.470
Dave. It's a two -part question. What are you
01:21:55.470 --> 01:21:58.670
working on now and where can our listeners go
01:21:58.670 --> 01:22:01.409
to find out more about you and all of the other
01:22:01.409 --> 01:22:05.579
stuff you're working on? So by absolute coincidence,
01:22:06.000 --> 01:22:13.880
I actually accepted a job today. So I am going
01:22:14.359 --> 01:22:18.000
to work with a guy, actually a couple of guys
01:22:18.000 --> 01:22:22.500
that I've known over 20 years, almost 25 years
01:22:22.500 --> 01:22:27.550
actually, who have, who started a business about
01:22:27.550 --> 01:22:34.510
10 years ago where they would put smart sensors
01:22:34.510 --> 01:22:36.890
into things that were shipped around the world.
01:22:37.010 --> 01:22:39.210
And they would do things like measure shock,
01:22:39.409 --> 01:22:42.489
measure temperature, measure humidity, and raise
01:22:42.489 --> 01:22:44.670
alarms if anything went out of spec. So like
01:22:44.670 --> 01:22:47.430
if you were shipping drugs, for example, then...
01:22:47.689 --> 01:22:50.649
there's like a window of time you're not allowed
01:22:50.649 --> 01:22:52.789
to go above a certain temperature for more than
01:22:52.789 --> 01:22:54.529
so many minutes or else you have to throw the
01:22:54.529 --> 01:22:56.229
whole thing away or whatever it might be anyway
01:22:56.229 --> 01:22:59.229
so they started doing that and it's grown and
01:22:59.229 --> 01:23:05.850
grown um and now they're um uh helping like uh
01:23:05.850 --> 01:23:09.970
really big name companies ship things around
01:23:09.970 --> 01:23:12.390
the world uh not just like shipping but for things
01:23:12.390 --> 01:23:16.489
like building data centers which apparently um
01:23:17.529 --> 01:23:20.510
like $500 million worth of stuff goes into a
01:23:20.510 --> 01:23:25.689
data center. So I'm joining them as a kind of,
01:23:25.710 --> 01:23:28.329
in simplicity, I would call it developer without
01:23:28.329 --> 01:23:31.930
portfolio. But my job is to kind of hover and
01:23:31.930 --> 01:23:35.510
make stupid suggestions. And so that's what I'm
01:23:35.510 --> 01:23:40.319
doing now. So I'm really looking forward to that.
01:23:41.260 --> 01:23:46.829
To find me online, I'm just called... Prag Dave,
01:23:46.970 --> 01:23:50.890
just about anywhere. That's P -R -A -G -D -A
01:23:50.890 --> 01:23:54.970
-V -E. Well, Dave, I really appreciated you taking
01:23:54.970 --> 01:23:58.029
the time. I know we ran over by about 15 minutes,
01:23:58.229 --> 01:24:01.409
but... Hey, I'm a software guy. I always run
01:24:01.409 --> 01:24:04.510
over. That's kind of how that works, right? Before
01:24:04.510 --> 01:24:07.689
we... We're agile. Yeah, we're agile. Before
01:24:07.689 --> 01:24:10.810
we hit stop on this little record here button,
01:24:10.970 --> 01:24:13.909
is there anything else that you want to... Tell
01:24:13.909 --> 01:24:15.670
our listeners or anything, did I miss anything?
01:24:15.829 --> 01:24:18.409
Did Felipe not ask you a question you were dying
01:24:18.409 --> 01:24:21.090
for him to ask you? No, you asked me a whole
01:24:21.090 --> 01:24:22.850
bunch of questions I was dying that you wouldn't
01:24:22.850 --> 01:24:27.409
ask me. No, but very enjoyable. Thank you very
01:24:27.409 --> 01:24:29.869
much for the opportunity. You're welcome. If
01:24:35.529 --> 01:24:38.289
any zombies happen to get into your house, go
01:24:38.289 --> 01:24:42.069
for the brains. We insist, do not leave your
01:24:42.069 --> 01:24:45.090
house. Thank you very much for your cooperation.
01:24:56.569 --> 01:25:16.479
Well, folks, that wraps up another episode of
01:25:16.479 --> 01:25:18.920
Roundabout Creative Chaos. You can find us on
01:25:18.920 --> 01:25:22.560
the web at roundaboutcreativechaos .com. We're
01:25:22.560 --> 01:25:25.960
also on Blue Sky, TikTok, Instagram, YouTube,
01:25:26.359 --> 01:25:30.800
LinkedIn, and X at Roundabout Chaos. If you enjoyed
01:25:30.800 --> 01:25:33.199
this episode, please consider leaving a review.
01:25:33.479 --> 01:25:36.500
Maybe even send a friend our way. It's the best
01:25:36.500 --> 01:25:40.420
thing you can do for us and for them. Well, short
01:25:40.420 --> 01:25:43.819
of sending cookies. So until next time, stay
01:25:43.819 --> 01:25:47.460
sharp, keep your brains intact, and whatever
01:25:47.460 --> 01:25:50.779
you do, don't forget rule number one, cardio.