תעתיק 🔗
WEBVTT
NOTE
This file was generated by Descript
00:00:04.116 --> 00:00:07.146
Jonny: Hi everyone, and welcome
to another episode of the Near
00:00:07.146 --> 00:00:09.206
Future podcast with Tom and Jonny.
00:00:09.426 --> 00:00:11.406
This podcast is all about design and ai.
00:00:11.886 --> 00:00:17.796
so I'm weirdly excited, to
introduce our next guest, Matt Webb.
00:00:19.386 --> 00:00:23.946
when you think about design and you
of a certain age, Matt is someone
00:00:23.946 --> 00:00:25.776
that you, will likely think of.
00:00:26.136 --> 00:00:31.506
I remember being a graduate or there
or thereabouts and this weird company.
00:00:31.506 --> 00:00:35.436
Berg was, doing all sorts of
interesting things, including making
00:00:35.436 --> 00:00:39.096
a piece of hard work called Little
Printer that, was effectively a
00:00:39.096 --> 00:00:41.346
receipt printer with a smiley face.
00:00:41.346 --> 00:00:44.766
And it was like really early internet
of things where you could send it
00:00:45.216 --> 00:00:49.441
your Twitter feed or it could send
you the print out, the weather for you
00:00:49.441 --> 00:00:52.056
and you could put it in your pocket
and take it with you on the train.
00:00:52.476 --> 00:00:56.706
I, Matt, I'm sure I'm, missing a hundred
other use cases, but it was just the
00:00:56.706 --> 00:01:02.886
main, it was just beautiful, piece of
hardware and, design that lit everyone up.
00:01:03.036 --> 00:01:03.966
I never got one.
00:01:04.896 --> 00:01:05.976
great injustice.
00:01:06.246 --> 00:01:11.736
I think my, my graduate salary didn't
quite stretch, maybe, I don't know.
00:01:11.736 --> 00:01:18.246
But, since then, Matt, I mean, you've
done a hundred different things.
00:01:18.846 --> 00:01:23.856
I've been following your blog,
a weekly written blog forever, I
00:01:23.856 --> 00:01:27.486
think, I dunno when you started
writing it, but very, early on.
00:01:29.496 --> 00:01:31.236
and it's an incredible read.
00:01:32.106 --> 00:01:34.716
you've been running office hours and
actually the first time we met was
00:01:34.746 --> 00:01:37.536
I booked in for one of your office
hours, half an hour on a Friday
00:01:37.536 --> 00:01:39.876
in late 2023 or whenever it was.
00:01:39.876 --> 00:01:44.826
And we talked about selling a business
versus pivoting and, various existential
00:01:44.826 --> 00:01:47.406
crisis, crises I was having at the time.
00:01:48.396 --> 00:01:55.866
and then since then you've gone into,
get into back into hardware, and I was
00:01:55.866 --> 00:02:03.576
an early signup for Poem/1, which is
your clock, which writes a new haiku,
00:02:04.836 --> 00:02:07.236
telling the time every minute, forever.
00:02:07.566 --> 00:02:10.836
And so Tom and I are both owners of
Poem/1ss and we'll get into that as well.
00:02:11.286 --> 00:02:14.796
And, and now you are running
Inanimate, which we will also get
00:02:14.796 --> 00:02:20.586
into and is very much hardware
focused and, agent Agentic as well.
00:02:20.586 --> 00:02:23.646
So I don't wanna give up away
too much, but it's probably clear
00:02:23.646 --> 00:02:28.146
from the introduction that I'm
a huge fan and, overly gushing.
00:02:28.266 --> 00:02:29.706
So I'll probably continue to do that.
00:02:29.736 --> 00:02:31.686
Apologies to everyone
listening in advance.
00:02:32.076 --> 00:02:35.016
if you feel a little bit sick in
the back of your mouth, I wanna
00:02:35.016 --> 00:02:38.676
hand over to you, Matt, and end
this introduction by asking you.
00:02:40.468 --> 00:02:45.538
When I reflected on everything I've seen
you do over the last however many years,
00:02:45.628 --> 00:02:49.258
it feels like everything has logically
built on the previous thing that you did.
00:02:50.158 --> 00:02:56.578
and it also feels like it's hard to put
you in a box, other than maybe inventor.
00:02:56.578 --> 00:02:57.868
That was the closest word I got to.
00:02:57.868 --> 00:03:03.568
So I'm just wondering if you could
start by just describing how Berg and
00:03:03.568 --> 00:03:09.238
Little Printer got you here and what are
these through lines, through a couple
00:03:09.238 --> 00:03:14.188
of decades of work that have, sort of
stuck with you and taught you more about
00:03:14.188 --> 00:03:16.078
yourself and all of that kind of thing.
00:03:16.078 --> 00:03:17.398
And we'll start there.
00:03:17.548 --> 00:03:18.178
Existential.
00:03:19.198 --> 00:03:20.278
Matt: Yeah, kind words.
00:03:20.278 --> 00:03:21.178
Thank you so much.
00:03:22.798 --> 00:03:24.328
wish I knew how to put it all in order.
00:03:24.328 --> 00:03:27.028
So I'm a, I started in
as a software engineer.
00:03:28.978 --> 00:03:36.088
co-founded a design studio with, Jack
Schulze and Matt Jones joined as well.
00:03:36.088 --> 00:03:38.153
And then that was book.
00:03:39.543 --> 00:03:43.378
And I suppose one of the things it
did was look for new opportunities
00:03:43.378 --> 00:03:47.308
and technology and design to bring
new things into the world, perhaps.
00:03:47.308 --> 00:03:48.898
So it was kind of like product invention.
00:03:49.558 --> 00:03:54.628
And I think that's a, you know, that's,
that, that is something which I've kind
00:03:54.628 --> 00:03:56.518
of taken through my entire career, right?
00:03:56.908 --> 00:03:57.928
where, are the new opportunities?
00:03:57.928 --> 00:03:58.858
What can you do something new?
00:03:59.188 --> 00:04:01.528
And how that works is different.
00:04:01.528 --> 00:04:03.088
It's very kind of context dependent.
00:04:03.088 --> 00:04:07.468
So sometimes that means kind of
making, sometimes that means writing.
00:04:07.678 --> 00:04:11.938
I spent some time as a consultant
editor and Google Research doing,
00:04:12.088 --> 00:04:14.848
you know, editorial to kind of link
together their research in there.
00:04:14.968 --> 00:04:15.298
Product.
00:04:15.568 --> 00:04:18.598
Sometimes that means,
you know, helping teams.
00:04:18.658 --> 00:04:23.368
So I was investing for a
Load full transcript
little bit other
people's money upset with RGA ventures.
00:04:24.148 --> 00:04:26.698
But again, where are the new opportunities
and where do you bring them in?
00:04:26.998 --> 00:04:30.568
And I think operating kind of very
early at that kind of like invention
00:04:30.568 --> 00:04:32.278
stage makes it very hard to go.
00:04:33.358 --> 00:04:34.678
oh, it's about engineering.
00:04:34.738 --> 00:04:39.448
Oh, it's about design, or it's about
you know, strategy or go to market.
00:04:39.448 --> 00:04:42.148
Because at that point everything
touches everything else.
00:04:43.048 --> 00:04:48.928
So personally, I would say I know
enough about most things, badly,
00:04:49.288 --> 00:04:52.798
like enough to be dangerous, but
it's the mix of all of them together,
00:04:52.798 --> 00:04:56.638
which means it's possible to kind of
go, oh, we could do this with that.
00:04:58.048 --> 00:05:03.838
There was a moment about, I don't
know, maybe Jonny, like near to when we
00:05:03.838 --> 00:05:09.328
talked, where I was going through my own
personal, am I in the world, what do I do?
00:05:10.138 --> 00:05:12.178
And I'd really lost my mojo with work.
00:05:13.018 --> 00:05:16.588
And, I think then I took advice
from a friend who was just like,
00:05:16.588 --> 00:05:18.118
really focus on your practise.
00:05:18.868 --> 00:05:22.288
Like what do you like, do you do?
00:05:22.288 --> 00:05:23.068
How do you do it?
00:05:23.788 --> 00:05:27.118
And I started thinking about what
areas I wanted to work in rather
00:05:27.118 --> 00:05:28.468
than kind of what role I wanted.
00:05:29.488 --> 00:05:32.218
And practised for me really changed
a few years ago where I realised
00:05:32.218 --> 00:05:36.508
that what I care about is I care
about human computer interaction.
00:05:37.258 --> 00:05:40.348
care about small groups of people,
I care about the physical world.
00:05:41.188 --> 00:05:42.418
And so I was just like, you know what?
00:05:42.418 --> 00:05:45.118
I will apply my practise just
to investigating those things.
00:05:45.148 --> 00:05:49.978
And I, changed the way I worked
to be much, much more hands-on.
00:05:51.268 --> 00:05:52.678
And that was five or six years ago.
00:05:52.678 --> 00:05:53.278
I've not looked back.
00:05:53.968 --> 00:05:54.568
Best thing I did.
00:05:56.148 --> 00:06:02.838
Jonny: Yeah, a lot of, maybe design ons
or whatever we're called now, would've,
00:06:02.898 --> 00:06:08.208
you know, ended up in directing huge
teams of people and, you know, driving,
00:06:08.208 --> 00:06:12.408
big change across multi-billion dollar
companies and going in that direction.
00:06:12.768 --> 00:06:16.728
And it really, I think I noticed that
point in your career where you were like,
00:06:16.728 --> 00:06:20.118
no, I just want to build shit basically.
00:06:20.118 --> 00:06:22.098
I just want to like stay curious.
00:06:22.728 --> 00:06:23.088
and
00:06:27.258 --> 00:06:30.588
I think I've gone through a
similar experience basically where,
00:06:31.638 --> 00:06:36.648
coming out of exiting a startup
and going, well, I haven't had a
00:06:36.648 --> 00:06:39.138
design title for like eight years.
00:06:40.068 --> 00:06:42.678
do I want to be a designer now or
do I want to do something else?
00:06:43.038 --> 00:06:48.318
or do I want to be a founder mainly
or, and ending up trying not to worry
00:06:48.318 --> 00:06:53.658
too much about that and start just
thinking about what do I actually enjoy.
00:06:54.648 --> 00:07:00.408
and on acts not facts, which is
your consulting vehicle, website,
00:07:00.438 --> 00:07:03.378
you have those, the Venn diagram
of those three things, don't you?
00:07:03.858 --> 00:07:07.278
I think you say it quite openly, like,
these are the things I care about if
00:07:07.278 --> 00:07:08.298
you want to talk about those things.
00:07:08.298 --> 00:07:08.778
Great.
00:07:10.188 --> 00:07:14.148
so was there a, some
nervousness around that?
00:07:14.148 --> 00:07:20.118
Like what if I reject these obvious
career paths or this, this expected way
00:07:20.118 --> 00:07:24.348
of doing things like judgement or like
lack of income or any of that stuff.
00:07:24.348 --> 00:07:26.058
Was that part of the equation for you?
00:07:28.462 --> 00:07:35.602
Matt: Think I was very lucky making that
decision about refocusing because I had,
00:07:35.782 --> 00:07:39.442
I'd just come out of a period working
with a company where I was, you know,
00:07:40.972 --> 00:07:44.872
doing products and design, but a lot
of it, it kind of started with coaching
00:07:44.872 --> 00:07:48.862
designers and I was definitely kind of
on that track and that didn't work out.
00:07:49.282 --> 00:07:52.792
You know, the company up and
I was like, what do I do?
00:07:54.142 --> 00:08:01.452
And the chat GPT moment just happened
and I've been on the GPT-3 beta
00:08:02.152 --> 00:08:06.442
some time and that space gave me the
real opportunity to kind of dig in.
00:08:07.372 --> 00:08:12.502
And my sort of like, my meta observation
was, I remember like my first kind
00:08:12.502 --> 00:08:17.512
of notes about GPT-3 was this thing
is capable of like creative ideas
00:08:17.692 --> 00:08:20.272
or rather it's not capable, but
it's possible to use it together.
00:08:20.422 --> 00:08:21.982
And that's new in computing.
00:08:21.982 --> 00:08:24.832
And I looked there and I was
like, computing's not the same.
00:08:25.702 --> 00:08:27.592
Jobs aren't gonna exist
in the same kind of way.
00:08:29.122 --> 00:08:31.852
I don't know how I'm gonna
navigate my way through this.
00:08:32.722 --> 00:08:35.212
I know is that we are
imagination bottlenecked.
00:08:35.962 --> 00:08:38.902
We're not gonna sort this out by standing
in front of a whiteboard or a Figma.
00:08:39.652 --> 00:08:41.812
only way you figure that out is
to get your hands dirty and to
00:08:41.812 --> 00:08:42.922
see what the material tells you.
00:08:44.122 --> 00:08:49.402
And so I was very lucky in that,
I had this kind of
need to figure
00:08:49.402 --> 00:08:53.242
it out and this new technology had
kind of arisen and that the only
00:08:53.242 --> 00:08:54.772
way through it was to start makings.
00:08:55.552 --> 00:09:01.402
So the first thing I made was, there's a
radio show I really love, BBC in our time.
00:09:02.332 --> 00:09:04.012
I made the first podcast
trip back in the day.
00:09:04.402 --> 00:09:07.282
So I like kind of deep affection
for it and I kind of wanted a
00:09:07.282 --> 00:09:08.422
better way of navigating it.
00:09:09.172 --> 00:09:14.932
And I decides to make a website
file and you know, this was like
00:09:14.932 --> 00:09:18.322
right at the beginning, kind of, you
know, around the same time as chat
00:09:18.322 --> 00:09:19.582
GPTI was like, oh, you know what?
00:09:19.582 --> 00:09:21.322
I could make this using GPT-3, right?
00:09:21.322 --> 00:09:22.612
Like use it for the data extraction.
00:09:22.942 --> 00:09:24.502
Turns out I was one of the
first people to do that.
00:09:24.502 --> 00:09:26.362
So it kind of got hacking used
and all those kind of things.
00:09:26.812 --> 00:09:31.492
But was the moment where I was
just like, oh, this technology out
00:09:31.492 --> 00:09:33.532
of its many different attributes.
00:09:34.582 --> 00:09:40.492
unblocks me in of making work
in a way that used to kind of
00:09:40.582 --> 00:09:41.902
require kind of a larger team.
00:09:43.072 --> 00:09:46.432
And then I started carrying on making
more things like that and ended up
00:09:46.432 --> 00:09:49.942
developing opinions about how we
work and collaborate with AI tools.
00:09:51.202 --> 00:09:54.292
And that let me investigate the other
things in my then, right, like kind of
00:09:54.292 --> 00:09:58.222
the physical space bit of multiplayer
then being able to kind of have a practise
00:09:58.222 --> 00:10:02.122
of just like making, I had a rule for
myself, which is one gif every week, Like
00:10:02.122 --> 00:10:05.392
investigate, communicate, share each week.
00:10:05.872 --> 00:10:09.292
And after like a month of this,
somebody was like, Hey, can
00:10:09.292 --> 00:10:12.412
you come and do that for us?
00:10:13.852 --> 00:10:17.842
And that was a whole other opportunity
to start learning and building and things
00:10:17.842 --> 00:10:19.222
just really kind of cascaded from there.
00:10:20.148 --> 00:10:23.298
Jonny: A month of kind of
curiosity, pure curiosity, and then
00:10:24.408 --> 00:10:25.908
like a green shoot of, hold on.
00:10:25.908 --> 00:10:27.078
I could just do this.
00:10:27.798 --> 00:10:28.428
maybe,
00:10:30.618 --> 00:10:31.998
yeah, that's, really fascinating.
00:10:32.058 --> 00:10:36.948
You mentioned, a gif every week,
you write a blog post every week.
00:10:38.418 --> 00:10:44.568
How important has writing been to you as
a way to, I suppose, think through either
00:10:44.568 --> 00:10:48.228
these problems or these directions you
want to head or even just the work itself?
00:10:49.486 --> 00:10:52.936
Matt: Yeah, I think, I mean,
you'll kind of see, if you read
00:10:52.936 --> 00:10:55.786
my blog, it's often very scrappy.
00:10:55.786 --> 00:10:57.346
There are typos everywhere.
00:10:58.126 --> 00:11:01.816
but one of the things I use it for
is to get my, like I start with kind
00:11:01.816 --> 00:11:05.536
of something that I believe or a
hunch, and I try and unpack it and
00:11:05.536 --> 00:11:07.246
explain why I think that's important.
00:11:08.146 --> 00:11:10.246
process of linear, linear,
00:11:12.616 --> 00:11:16.036
helps me put ideas in order and kind of
give them words and a bit of structure.
00:11:17.176 --> 00:11:22.936
And happens then is I think about
the next thing, and that takes me
00:11:22.936 --> 00:11:24.226
to an idea I didn't have before.
00:11:25.186 --> 00:11:29.656
So for me, is not just a kind of a way,
or at least in my own blog of kind of
00:11:29.656 --> 00:11:32.626
documenting or communicating, but it
is the process of having new ideas.
00:11:33.436 --> 00:11:35.356
You know, lots of people
do that with sketchbooks.
00:11:35.506 --> 00:11:38.506
Lots of people do it with their hands, you
know, other people do it in conversation.
00:11:40.636 --> 00:11:41.236
it's writing.
00:11:42.196 --> 00:11:46.426
benefit of writing like that is
there are no replies on my blog,
00:11:47.446 --> 00:11:50.026
which means that nobody can kind
of go, oh, that was interesting.
00:11:50.026 --> 00:11:51.706
And then just like bang
out a few words, right?
00:11:52.396 --> 00:11:57.466
only way if so motivated they can kind
of like respond is to click on this
00:11:57.466 --> 00:12:01.996
link called on Office Hours they have
to book a call with me, which means they
00:12:02.026 --> 00:12:03.886
have to give up 30 minutes their time.
00:12:04.006 --> 00:12:04.246
Right?
00:12:05.026 --> 00:12:09.286
that means that the conversations,
I get through that by giving
00:12:09.286 --> 00:12:10.216
up replies in the blog.
00:12:10.216 --> 00:12:12.586
The conversations I get through
are incredibly high value.
00:12:13.606 --> 00:12:18.496
They're so, you know, I've done
maybe 400, 450 calls like that.
00:12:18.496 --> 00:12:19.696
We've done office hours over the years.
00:12:20.656 --> 00:12:24.736
Talk with people from all over the
worlds, all kinds of backgrounds.
00:12:25.576 --> 00:12:27.736
And they're always
incredible conversations.
00:12:28.186 --> 00:12:29.956
And so that becomes part of
my process as well, right?
00:12:30.106 --> 00:12:34.306
that's kind of part of having the feelers
out and understanding like kind of what
00:12:34.306 --> 00:12:36.886
is going on with new ideas and culture.
00:12:41.129 --> 00:12:43.889
Tom: Wow, that's a love,
love that approach.
00:12:46.079 --> 00:12:49.409
have to ask the question, since we are
an AI and design podcast, is, does AI
00:12:49.409 --> 00:12:51.119
feature in your writing process at all?
00:12:51.299 --> 00:12:59.489
Have you found ways that, it can
elevate or challenge, or
mirror anything
00:12:59.519 --> 00:13:02.489
in how you work through your ideas?
00:13:02.564 --> 00:13:04.454
Matt: feature one exception.
00:13:04.634 --> 00:13:05.234
I'll tell you something.
00:13:05.504 --> 00:13:12.854
So I tried once seeing, could
I involve AI in helping me
00:13:12.854 --> 00:13:14.234
structure my thoughts into write?
00:13:15.224 --> 00:13:20.264
And it came up with a turn of phrase that
was so perfect, I couldn't let it go.
00:13:21.494 --> 00:13:25.394
And that terrified me because when
I, you know, like I said, like I
00:13:25.394 --> 00:13:29.354
write in order to use the words that
come up to have more ideas, right?
00:13:29.684 --> 00:13:32.714
Like those are Lego bricks.
00:13:33.464 --> 00:13:37.904
And when I tried to lean on that
idea, that beautiful turn of
00:13:37.904 --> 00:13:39.134
phrase, it wouldn't take my weight.
00:13:39.854 --> 00:13:41.594
It was like I was
gaslighting myself, right?
00:13:41.594 --> 00:13:45.824
It was like there was this idea that
sounded like it was from me, but it just
00:13:45.824 --> 00:13:50.384
didn't, yeah, it couldn't bear any weight.
00:13:50.384 --> 00:13:51.374
I couldn't use it for anything.
00:13:52.634 --> 00:13:53.804
And it's scary, right?
00:13:53.804 --> 00:13:57.104
To almost kind of have these things
in my notes that feel like, oh,
00:13:57.764 --> 00:14:00.314
like my thoughts, but they're not.
00:14:00.794 --> 00:14:02.474
So it well away.
00:14:03.584 --> 00:14:09.074
so I don't let, I don't let
AI agents into my notes.
00:14:09.104 --> 00:14:10.094
I'm terrified of that.
00:14:11.204 --> 00:14:14.354
until very recently, I
wouldn't let them write emails.
00:14:15.134 --> 00:14:17.534
Well, that's something which has changed.
00:14:18.404 --> 00:14:22.244
And the one exception
there is, is transcription.
00:14:23.414 --> 00:14:28.334
So one of the ways I love writing
blog post now is to go for a walk
00:14:28.394 --> 00:14:34.694
and to talk into my watch 30 minutes
and then come back and turn that
00:14:34.694 --> 00:14:36.434
transcription into an actual blog post.
00:14:36.674 --> 00:14:37.904
I have a little workflow around that.
00:14:38.534 --> 00:14:42.764
kind of address it to a sort of a, a
fake assistant, because then you can
00:14:42.764 --> 00:14:44.294
kind of make outta band notes, right?
00:14:44.294 --> 00:14:47.564
So I call her Diane's out
of, I was obsessed with Twin
00:14:47.564 --> 00:14:48.554
Peaks when I was a kid, right?
00:14:48.554 --> 00:14:50.114
So like Dale Cooper talking to Diane.
00:14:50.834 --> 00:14:53.714
So I'll be like, you know, Diane
move that idea like to the previous
00:14:53.714 --> 00:14:57.464
section so you can kind of have this
kind of like conversation, which is
00:14:57.494 --> 00:15:01.214
both transcription and outta band
instructions about editing simultaneously.
00:15:01.604 --> 00:15:04.154
And then I'll take those words
and then turn it to a blog post.
00:15:04.844 --> 00:15:07.094
And those are, that's a
really fun way of writing.
00:15:07.094 --> 00:15:11.024
But apart from that, I'm like,
you know, won't live it in it's.
00:15:13.004 --> 00:15:15.104
Jonny: That's more the analogy of the,
00:15:15.209 --> 00:15:15.569
Tom: Great.
00:15:16.034 --> 00:15:20.354
Jonny: assistant with a, a notepad
while you dictate into your,
00:15:22.424 --> 00:15:25.544
into the air as you're wandering around
your office, and then they come back
00:15:25.544 --> 00:15:29.534
for, and say, this is what you wrote,
and then you can sort of review that.
00:15:30.494 --> 00:15:34.394
I found very similar writing, which is,
00:15:36.704 --> 00:15:37.574
it's almost too good.
00:15:37.574 --> 00:15:40.454
Like I'm not as good at
writing as it in some way.
00:15:40.454 --> 00:15:43.814
Like, it doesn't feel authentic.
00:15:45.224 --> 00:15:54.074
so yeah, I completely agree and I, dunno
how much, product design or sort of,
00:15:54.284 --> 00:15:57.194
visual design or that kind of stuff that
you're doing at the moment, but have
00:15:57.194 --> 00:16:00.494
you pushed into anything else with ai?
00:16:00.494 --> 00:16:06.464
Have you found AI to, be similar in
other aspects of your creativity as well?
00:16:06.464 --> 00:16:11.504
Do you do not trust it to sort
of complete something that is, I
00:16:11.504 --> 00:16:14.744
suppose more on the design side
rather than the, the writing side.
00:16:16.110 --> 00:16:18.120
Matt: I use, we use AI a lot, right?
00:16:18.120 --> 00:16:23.340
So, bit of context, you know, early
stage startup called Inanimate.
00:16:23.340 --> 00:16:25.200
I'm co-founder, co-founder Daniel.
00:16:27.420 --> 00:16:34.860
we have a kind of a small team of
partners, freelancers, employee, and
00:16:34.860 --> 00:16:38.490
we're putting AI into the real world, and
we're doing that via physical devices.
00:16:38.490 --> 00:16:41.580
So we're making hardware, we're
gonna be bringing out some stuff
00:16:41.580 --> 00:16:46.800
towards the, you know, towards the
end of the year, which are some very
00:16:46.800 --> 00:16:51.330
early prototypes to kind of get our
interactions and our technology out
00:16:51.330 --> 00:16:52.740
there in the world onto people's desks.
00:16:53.460 --> 00:16:56.100
just short run, you know,
not even products, right?
00:16:56.760 --> 00:16:59.400
And, you know, that's, that, that's
a lot of different streams to be
00:16:59.400 --> 00:17:00.870
running at the, the same time.
00:17:00.870 --> 00:17:02.760
And there's, you know, there's
a limited amount of time,
00:17:03.510 --> 00:17:04.770
course, Claude code, right?
00:17:04.800 --> 00:17:06.060
Like tonnes and tonnes of that.
00:17:06.390 --> 00:17:06.840
That's fine.
00:17:06.990 --> 00:17:07.890
You know, taken for granted.
00:17:08.280 --> 00:17:15.900
It's been wonderful watching our embedded
engineer, using an AI plugin he developed
00:17:16.080 --> 00:17:18.900
with PCB software to review PCB layout.
00:17:19.260 --> 00:17:20.610
That's kind of a use
useful tool.
00:17:21.510 --> 00:17:27.720
interesting is that one of the parts
of my process is that, about two thirds
00:17:27.720 --> 00:17:33.540
of the way through design work, will,
just like make the landing page for the
00:17:33.540 --> 00:17:35.880
projects it's a really interesting filter.
00:17:35.880 --> 00:17:38.430
You've kind of, you've kind of understood
what the thing is, you've been making
00:17:38.430 --> 00:17:42.300
it, you know, but there's a bit where
you have to kind of go, is this gonna
00:17:42.300 --> 00:17:43.770
tell the story to other people, right?
00:17:43.770 --> 00:17:47.790
Like, not just kind of what the design is,
but how it's kind of understood physical
00:17:47.795 --> 00:17:52.680
things with kind of software behaviours
and making that landing page is.
00:17:54.734 --> 00:17:56.924
It's not gonna be the landing page, right?
00:17:56.924 --> 00:18:00.314
So it's kind of redundant work to make
it good, but there is something useful
00:18:00.314 --> 00:18:04.034
in kind of like helping to think out
the information hierarchy the messaging
00:18:04.034 --> 00:18:08.114
hierarchy and how to tell the story
and which bits you're gonna highlight,
00:18:08.534 --> 00:18:12.284
where it's useful to go through the
process because it what it tells you.
00:18:13.874 --> 00:18:16.514
And I tried for the first time the
other day to go, well, what if I could
00:18:16.514 --> 00:18:18.014
do this via transcription, right?
00:18:19.094 --> 00:18:24.284
So I got a coffee, went and sat in the
park, I talked about this collection
00:18:24.434 --> 00:18:30.224
for putting together and just everything
on my mind, like 40 minutes and came
00:18:30.224 --> 00:18:40.034
back and with, GPT in that case to
make a microsite to explain these
00:18:41.144 --> 00:18:43.364
using, do you know, tldraw the,
00:18:43.956 --> 00:18:44.256
Jonny: Yep.
00:18:44.294 --> 00:18:45.464
Matt: whiteboarding app, right?
00:18:45.914 --> 00:18:50.354
They have, tldraw Offline, which is a
version which has of agents skills that
00:18:50.354 --> 00:18:54.279
agents are really good apps, like seeing
it and drawing on it and iterating.
00:18:55.544 --> 00:19:00.554
And I had it make me the microsite based
on my transcription in tldraw offline.
00:19:01.034 --> 00:19:03.764
Then I covered it in post-it notes
and got it to read the post-it
00:19:03.764 --> 00:19:04.934
notes and do another iteration.
00:19:06.104 --> 00:19:09.584
And it was useful I think, not because
this is anything which is gonna go
00:19:09.584 --> 00:19:16.274
in front of people, but because it,
it helped organise my own thoughts
00:19:16.274 --> 00:19:17.504
and gave me something to bounce off.
00:19:18.914 --> 00:19:20.834
And that's been a kind of
an interesting application.
00:19:22.388 --> 00:19:25.568
Jonny: Like AI as editor
rather than writer.
00:19:26.498 --> 00:19:27.308
is that reductive?
00:19:28.398 --> 00:19:29.538
Matt: AI is rubber duck.
00:19:30.420 --> 00:19:30.660
Jonny: Yeah.
00:19:30.738 --> 00:19:31.578
Matt: rubber duck phrase?
00:19:31.638 --> 00:19:31.998
Right.
00:19:32.040 --> 00:19:32.280
Jonny: Yeah.
00:19:32.298 --> 00:19:34.758
Matt: know, the, kind of, you know,
you're trying to figure out a difficult
00:19:34.758 --> 00:19:36.168
bug, so you need someone just to talk to.
00:19:36.498 --> 00:19:38.208
I think there's a little
bit about that, right?
00:19:38.292 --> 00:19:38.502
Jonny: Yeah,
00:19:38.748 --> 00:19:39.018
Matt: Yeah.
00:19:40.362 --> 00:19:42.732
Jonny: if you were rambling in
the park to yourself then that
00:19:42.732 --> 00:19:45.312
would be probably ably insane.
00:19:45.762 --> 00:19:49.092
Whereas if you're rambling at least
to something that might have an
00:19:49.092 --> 00:19:52.782
opinion about it, or, I mean maybe
if you had an actual rubber duck in
00:19:52.782 --> 00:19:54.732
the park, then would that be useful?
00:19:54.732 --> 00:19:55.092
I dunno, but.
00:19:56.486 --> 00:19:56.876
Matt: I do.
00:19:56.876 --> 00:20:00.026
I do wonder what people think when
they see me walking around and
00:20:00.026 --> 00:20:02.156
talking, because like at a certain
point you're like, that's fine.
00:20:02.156 --> 00:20:04.946
They're on a phone call and they've
got AirPods on, and if they're
00:20:05.336 --> 00:20:08.156
sitting opposite, they'd be like,
that guy's been talking nonstop.
00:20:08.306 --> 00:20:09.356
Like 20 minutes.
00:20:09.836 --> 00:20:13.376
Like, let the other
person have a bit there.
00:20:13.376 --> 00:20:15.206
Come on, what is
00:20:15.380 --> 00:20:15.680
Jonny: This.
00:20:17.216 --> 00:20:17.906
Matt: No, it's great.
00:20:17.906 --> 00:20:22.316
And there is something, there's something
great about like, you know, I was
00:20:22.376 --> 00:20:27.446
debugging some firmware recently and I had
the piece of hardware plugged into my Mac
00:20:27.446 --> 00:20:32.066
mini at home and I was the debugging just
on my phone, just like typing in, like try
00:20:32.066 --> 00:20:34.286
these things out and walking in the sun.
00:20:35.006 --> 00:20:37.976
there is something wonderful about kind
of like walking in the sun, debugging
00:20:37.976 --> 00:20:42.146
firmware, walking in the sun, thinking
about like how a site is gonna work and
00:20:42.146 --> 00:20:45.686
how the messaging hierarchy works and
how that impacts the product design,
00:20:46.676 --> 00:20:48.986
which is, a different way of working.
00:20:48.986 --> 00:20:51.716
I think it provides a looseness,
which I think is really important.
00:20:55.593 --> 00:20:59.523
Tom: I think so much of what you're saying
feels like it's built around this tight
00:20:59.523 --> 00:21:05.223
feedback loop of making something, writing
something, getting input from others, sort
00:21:05.223 --> 00:21:09.933
of reducing the friction and how you that
into the world and iterate on it quickly.
00:21:10.293 --> 00:21:16.023
I'm curious in relation to that,
and I'm curious about your energy
00:21:16.263 -->
00:21:17.793
throughout this whole experience.
00:21:17.793 --> 00:21:21.513
I feel like there's been a big change
in, for a lot of us, working with AI
00:21:21.513 --> 00:21:26.373
in the last few years of things feeling
increasingly tiring or exhausting
00:21:26.373 --> 00:21:28.203
because you're expending so much
more energy and a lot of what you're
00:21:28.203 --> 00:21:32.973
just describing there sounds like
very, mentally intensive activities.
00:21:33.063 --> 00:21:36.663
Like the thing that used to be the
break to go to the park is now a 40
00:21:36.663 --> 00:21:39.873
minute walking, production session.
00:21:39.873 --> 00:21:43.173
And then presumably you get back to your
laptop afterwards and, take that and
00:21:43.173 --> 00:21:44.673
start moulding it into something else.
00:21:44.823 --> 00:21:49.743
Like, is there any, are there any changes
to how you balance or how you create
00:21:49.743 --> 00:21:56.223
space for letting your brain absorb
what is happening without constantly
00:21:56.223 --> 00:22:00.513
making, or is that part of your,
personality to just producing and then
00:22:00.603 --> 00:22:02.103
it sort of works itself out in the wash?
00:22:02.174 --> 00:22:05.114
Matt: that, you know, you sort
of get that a lot with coding
00:22:05.114 --> 00:22:07.094
agents especially, right?
00:22:07.364 --> 00:22:11.204
And I wonder how pilots deal with this.
00:22:11.204 --> 00:22:15.464
Well, you know, airline pilots, you
know, fly by wire, you know, all
00:22:15.464 --> 00:22:19.184
the rest takes over and they only
get involved in difficult bits.
00:22:19.454 --> 00:22:22.514
But at least if you're flying a
plane, they don't kind of say, well
00:22:22.514 --> 00:22:24.914
that boring kind of eight hours in
the middle, you've now got to go and
00:22:24.914 --> 00:22:26.264
do the difficult landings elsewhere.
00:22:26.954 --> 00:22:30.584
is an element when you're working with
half a dozen agents at once, you are only
00:22:30.584 --> 00:22:34.219
involved in the really intellectually
difficult bits as opposed to the kind
00:22:34.219 --> 00:22:38.954
of the very pleasant sort of like, you
know, doing the knitting in the middle.
00:22:40.304 --> 00:22:41.474
It's exhilarating, isn't it?
00:22:43.424 --> 00:22:44.114
this whole world.
00:22:44.864 --> 00:22:48.374
And you definitely need to there are
moments where you just kind of take a
00:22:48.374 --> 00:22:53.684
step back and just go, I just need to
like think about this, build my own
00:22:53.684 --> 00:22:57.254
mental models, relax a bit with it.
00:22:58.124 --> 00:23:00.494
but you know, it's also a startup world.
00:23:00.494 --> 00:23:04.004
This is not the time to be kind
of, you know, pitching back.
00:23:04.009 --> 00:23:07.244
I've got some, you know, we've got some
very kind of definite opinions about
00:23:07.244 --> 00:23:10.244
what we're building and it's a matter
of like making it work at the moment.
00:23:13.734 --> 00:23:18.594
Jonny: we should switch onto,
Poem/1 and inanimate in more depth.
00:23:18.594 --> 00:23:22.224
But before we do that, just to
sort of round out this, this first
00:23:22.224 --> 00:23:27.414
section, you have carved out space
for yourself to explore in this way.
00:23:28.104 --> 00:23:31.104
Tom and I both actually have over
the last year, part of what we're,
00:23:31.164 --> 00:23:35.514
we've been doing intentionally is
allowing ourselves to play with this
00:23:35.514 --> 00:23:36.714
technology and get a feel for it.
00:23:37.044 --> 00:23:42.354
A lot of, people working full time,
you know, as a designer inside an
00:23:42.354 --> 00:23:44.814
organisation, which sets the parameters
around what they can do a little
00:23:44.814 --> 00:23:48.834
bit more, don't necessarily have as
much of the like free thought, you
00:23:48.834 --> 00:23:51.474
know, zero to one type work to do.
00:23:52.344 --> 00:23:59.784
Do you have, I suppose, do you have any,
thoughts or advice as maybe, but just
00:23:59.844 --> 00:24:04.374
around what everyone should be picking
up right now, if they're not already
00:24:04.374 --> 00:24:09.594
to just help inform their worldview
around what's coming or like what,
00:24:09.804 --> 00:24:11.214
how they should change their practise?
00:24:12.506 --> 00:24:16.826
Matt: I think there's, I would say
companies are, I've used this phrase
00:24:16.826 --> 00:24:18.896
before, imagination, bottleneck, right?
00:24:19.346 --> 00:24:21.986
So, you know, for the last like
kind of 10, 20 years we've been
00:24:21.986 --> 00:24:25.046
working with digital and we all
kind of know what that looks like.
00:24:25.076 --> 00:24:27.926
We know what good looks like, we
know what can be done with a team
00:24:27.926 --> 00:24:29.576
of six in three months, right?
00:24:29.576 --> 00:24:32.756
Like everything has got a kind of a,
a shape to it that we understand and
00:24:32.756 --> 00:24:37.196
we figure out the best thing to do by,
know, doing wide frames and standing in
00:24:37.196 --> 00:24:44.846
front of sopping, post-its, that isn't
the case with these new technologies.
00:24:45.896 --> 00:24:50.786
We don't know the best ways to
design them and, we have to be
00:24:50.786 --> 00:24:53.366
really sensitive to kind of what
people feel when they're using them.
00:24:53.666 --> 00:24:55.556
and these are often new
feelings that people having.
00:24:55.556 --> 00:24:59.396
You know, Tom, you mentioned
one about the, overwhelm, right?
00:25:00.056 --> 00:25:02.786
And there are other kind of new
feelings that like users have as well.
00:25:03.476 --> 00:25:05.336
And the only way to figure
those out is to make things.
00:25:06.146 --> 00:25:10.016
So, you know, where this kind
of reminds me of the very early
00:25:10.016 --> 00:25:12.146
days of smartphones taking off.
00:25:12.236 --> 00:25:16.226
It's like we do not know what is
possible, what's easy, what's difficult,
00:25:16.226 --> 00:25:17.516
you know, our intuitions are wrong.
00:25:18.356 -->
00:25:21.596
that's at a corporate level like
that, that has to happen somehow.
00:25:21.956 --> 00:25:25.916
Any leaders out there like make time
for that sort of like build and share.
00:25:26.786 --> 00:25:31.856
But on the personal practise,
of our investors is Betaworks.
00:25:31.886 --> 00:25:36.116
And earlier this year we were with
Betaworks in New York with the
00:25:36.116 --> 00:25:38.846
rest of the cohort of, founders.
00:25:38.936 --> 00:25:43.646
Another, you know, another
bunch of, startup founders in
00:25:43.646 --> 00:25:44.846
our kind of camp they call it.
00:25:45.086 --> 00:25:47.636
And it was inspiring to see how
those people worked because they
00:25:47.636 --> 00:25:52.166
were embracing collaborating with
agents, like at a level I hadn't seen.
00:25:52.646 --> 00:25:56.186
And I realised at that point that
I need to invest more in my own
00:25:56.186 --> 00:25:58.496
tools and my own working practises.
00:25:58.496 --> 00:26:02.276
And I kind of, I would say that on
average I would probably invest,
00:26:02.336 --> 00:26:06.866
I dunno what, like a day, a month,
less into kind of, oh, you know,
00:26:06.926 --> 00:26:08.576
try a new note taking app, right?
00:26:08.666 --> 00:26:10.046
You know, that would be it.
00:26:10.316 --> 00:26:13.466
And I consciously said, this needs
to be about a day, a week or more.
00:26:14.966 --> 00:26:23.126
I need to be going, how do I, you know,
how do I use like, well, know, I don't
00:26:23.126 --> 00:26:28.946
know, like my desktop to drag my browser
to like, know, make the kanbans for
00:26:28.946 --> 00:26:32.066
me to, to run an autonomous fashion.
00:26:32.066 --> 00:26:36.356
Like I need to learn these things
and to be really kind of like about
00:26:36.506 --> 00:26:41.216
finding points where I can use
more automation and intelligence.
00:26:41.246 --> 00:26:44.396
And I'm not gonna find that
by just hoping it comes up.
00:26:44.876 --> 00:26:46.376
It needs to be a deliberate effort.
00:26:47.456 --> 00:26:50.216
And that was, you know, I, made
that kind of conscious shift and
00:26:50.216 --> 00:26:51.626
that was like really valuable.
00:26:52.496 --> 00:26:55.646
So I would say just like, it's not about
kind of exactly what tools to use, but
00:26:55.646 --> 00:27:00.656
just increase the time you give yourself
to figure out what to do because that is
00:27:00.656 --> 00:27:05.306
an investment that will compound really
quickly over a period of like months.
00:27:09.065 --> 00:27:09.875
Tom: True do.
00:27:09.935 --> 00:27:15.455
I'm curious, do you have any hunches
on whether that is a reflection
00:27:15.455 --> 00:27:19.205
of this current moment with AI
or that is a practise we will
00:27:19.205 --> 00:27:21.365
continue to be doing years from now?
00:27:22.414 --> 00:27:26.764
Matt: Think it's a difference today
based on how it was yesterday.
00:27:28.414 --> 00:27:30.124
I wonder what it will be like tomorrow.
00:27:30.124 --> 00:27:33.304
I dunno whether the exponential
is gonna, like, flatten off a
00:27:33.304 --> 00:27:37.204
bit or curve up or stay the same.
00:27:37.204 --> 00:27:37.444
Right.
00:27:38.884 --> 00:27:44.554
I think I am a, it's really difficult kind
of, I, like to say like, I'm a believer
00:27:44.554 --> 00:27:47.524
in kind of like investing in tools and,
you know, this is the man who's had the
00:27:47.524 --> 00:27:49.414
same way of writing for the last 20 years.
00:27:49.684 --> 00:27:52.594
So, you know, or you know, five or six.
00:27:52.594 --> 00:27:57.364
Anyway, it's very easy to spend time
going, gonna think about new processes
00:27:57.364 --> 00:28:00.754
and just like swap from one piece
of notetaking software to another.
00:28:01.474 --> 00:28:03.724
So there's got to be some kind of
discipline at least to go like,
00:28:03.994 --> 00:28:05.194
how am I gonna do this differently?
00:28:06.514 --> 00:28:08.734
But I, wonder, that's
an interesting question.
00:28:08.734 --> 00:28:10.474
I wonder what it would be
like in another two years.
00:28:14.531 --> 00:28:18.551
Tom: I think a lot of these conversations
have that, that question has been
00:28:18.551 --> 00:28:22.631
circling around like what is a
specific thing as a result of the
00:28:22.631 --> 00:28:27.011
state of the technology right now,
folks from larger organisations.
00:28:27.011 --> 00:28:30.341
There's more of a question of how much
is of this is a temporary infrastructure
00:28:30.341 --> 00:28:33.791
to introduce now and then a year from now
that will just be established and the team
00:28:33.791 --> 00:28:35.621
will be set up to work in this new way.
00:28:35.651 --> 00:28:38.831
And we're just sort of on the edge
of this transition is happening.
00:28:39.491 --> 00:28:44.171
But I think it, hearing how you work
and your own practise, it feels like
00:28:44.171 --> 00:28:50.921
it is fundamental to how you are, you
will continue to iterate in that space
00:28:50.921 --> 00:28:52.451
regardless of where the technology is.
00:28:52.461 --> 00:28:56.711
It, sounds like you, that sort of
feedback loop that is so central to
00:28:56.711 --> 00:29:00.131
how you work is critical doing this.
00:29:00.131 --> 00:29:03.911
So I can imagine you being who would
invest more in that space than others.
00:29:04.654 --> 00:29:05.524
Matt: always more to learn.
00:29:05.584 --> 00:29:10.594
I think one of the ways I think about it
is that these AI tools provide what I'd
00:29:10.594 --> 00:29:13.294
call like universal basic agency, right?
00:29:13.294 --> 00:29:18.784
There are a bunch of projects that I've
not done before because there are certain
00:29:18.784 --> 00:29:23.464
pieces to them that I'm not capable of
doing and that blocks the entire project,
00:29:24.814 --> 00:29:29.584
and AI just gets me over the hump with
something that might be quite small that
00:29:29.584 --> 00:29:31.204
somebody else would find very, easy.
00:29:32.374 --> 00:29:35.644
But if I wanted to do that project, it
would be a
question of enrolling someone
00:29:35.644 --> 00:29:38.824
else In this kind of picture in my
head, I'll give you an example, right?
00:29:39.274 --> 00:29:42.904
So for years I had this idea that
I wanted an arrow on my phone that
00:29:42.904 --> 00:29:44.314
would points to Centre of the Galaxy.
00:29:45.064 --> 00:29:45.844
Ridiculous idea.
00:29:48.782 --> 00:29:52.862
About 10, 15 years ago, I used to be
able to do this in my head, right?
00:29:52.952 --> 00:29:55.982
You could kind of like visualise
it, the earth has got kind of an
00:29:55.982 --> 00:29:58.532
axial tilt, goes around the sun.
00:29:58.532 --> 00:30:01.022
You can think about what time
of year the solar system itself
00:30:01.022 --> 00:30:03.872
has tilted, and then there's the
kind of the centre of the galaxy.
00:30:04.622 --> 00:30:06.782
And so the earth's rotating,
the earth's go around the sun.
00:30:06.782 --> 00:30:08.672
It's all these different
tilts, time of day.
00:30:09.152 --> 00:30:13.382
And you know, the central black hole
is 26,000 light years away straight
00:30:13.382 --> 00:30:17.522
through the constellation Sagittarius,
Why do you need to know that?
00:30:17.732 --> 00:30:18.182
Don't know.
00:30:18.272 --> 00:30:20.732
That's what I thought about at the bus
stop when I was waiting in is LinkedIn.
00:30:21.632 --> 00:30:23.132
over the years I forgot how to do this.
00:30:23.132 --> 00:30:23.822
And I was like, you know what?
00:30:23.822 --> 00:30:25.742
It'd be really good to have
a phone app to do this.
00:30:26.012 --> 00:30:29.162
I spent like two years trying to persuade
people to like, help me build that.
00:30:29.162 --> 00:30:30.572
I couldn't build phone apps, right?
00:30:30.782 --> 00:30:34.742
I can draw an arrow and find of say
how it wants to behave, couldn't do it.
00:30:35.582 --> 00:30:39.482
And that was, I talked to a friend and
he was like, yeah, I'll totally like, you
00:30:39.482 --> 00:30:43.652
know, I'll totally help you build this
app and, you know, never got round to it.
00:30:44.612 --> 00:30:48.242
And then one day I was like, oh, this
chat GPT might be able to do this, right?
00:30:48.242 --> 00:30:51.422
So I spent a day kind of like
copying and pasting code between
00:30:51.422 --> 00:30:52.982
the chat GPT window and Excode.
00:30:53.822 --> 00:30:55.772
was sort of previ coding.
00:30:56.582 --> 00:30:57.842
And it got me over the hump.
00:30:57.962 --> 00:31:01.292
There was a problem about how to
combine rotations, which requires a
00:31:01.292 --> 00:31:03.082
mathematical technique called quaternions.
00:31:03.932 --> 00:31:04.982
had to learn that as well.
00:31:05.822 --> 00:31:09.722
but it got me over the hump of being
able to, you know, I can't make apps,
00:31:09.858 --> 00:31:10.008
Jonny: foot.
00:31:11.192 --> 00:31:12.242
Matt: but then it worked, right?
00:31:12.717 --> 00:31:16.082
And so then I released this app and I'm
like, this is like just for me, right?
00:31:17.432 --> 00:31:18.332
Absolutely
blew up.
00:31:19.322 --> 00:31:23.822
And I started receiving emails from
people who are like, oh, you know what?
00:31:23.822 --> 00:31:25.532
I'm gonna, a really
difficult place in my life.
00:31:26.522 --> 00:31:29.312
I sit outside in the middle of the night
and I smoke a cigarette, and I think
00:31:29.312 --> 00:31:32.432
about my father and I look at your app
and it, you know, it suits me to know
00:31:32.432 --> 00:31:36.152
that there is one fixed point that we've,
you know, we can regard as like steady.
00:31:36.152 --> 00:31:39.542
I'm just, you know, imagine waking
up in the morning with an email
00:31:39.542 --> 00:31:42.452
like that in your inbox and it turns
out that all of that was worth it.
00:31:43.772 --> 00:31:47.822
Would, I have been able to persuade
anyone to make this very, very speculative
00:31:47.822 --> 00:31:50.672
thing with me, for me to put their time?
00:31:50.792 --> 00:31:51.362
Probably not.
00:31:52.292 --> 00:31:59.072
So what happens is like your, as you
know, your capabilities grow, your
00:31:59.222 --> 00:32:04.082
kind of desire to make more grows and
you can make more things which are
00:32:04.112 --> 00:32:06.122
based on even more esoteric hunches.
00:32:06.122 --> 00:32:08.192
I mean, you know, Poem/1, right?
00:32:08.192 --> 00:32:11.102
Like case in point, there's no way I
could have done that my end before.
00:32:12.422 --> 00:32:15.452
And so I think, yeah, like
just on learning more tools.
00:32:16.322 --> 00:32:19.352
it will be interesting to see where
this goes, but I, the thing I'm excited
00:32:19.352 --> 00:32:24.092
about is the individuals and designers
and the artists who can now put their
00:32:24.092 --> 00:32:29.012
kind of pre-verbal imaginations into
the world in a way they couldn't before.
00:32:32.977 --> 00:32:35.257
Tom: And, excellent segues Poem/1 as well.
00:32:35.467 --> 00:32:42.037
I wanted off this, this next segment with
a, a little story, which sets the scene
00:32:42.037 --> 00:32:43.717
for how this conversation came about.
00:32:44.077 --> 00:32:45.937
we were at a dinner recently.
00:32:46.747 --> 00:32:48.367
I was lucky to sit next to you.
00:32:48.427 --> 00:32:49.267
We were chatting.
00:32:49.597 --> 00:32:53.227
I think that was a good conversation
around what is boring and sort of trying
00:32:53.227 --> 00:33:00.757
to define the state of boredom and how to
achieve that great dinner conversation.
00:33:01.117 --> 00:33:02.587
Much more interesting than it sounds.
00:33:03.577 --> 00:33:08.857
but at dinner I also brought along
my Poem/1, my, physical device with
00:33:08.917 --> 00:33:13.927
the, e ink display, which, shows you
a new poem every minute for eternity.
00:33:14.557 --> 00:33:17.137
And you, held this device.
00:33:17.137 --> 00:33:19.237
I was a little bit nervous about
showing you, I thought I would come
00:33:19.237 --> 00:33:22.657
across as either a fanboy or just
some weird person from the internet
00:33:22.657 --> 00:33:24.367
who's brought this physical device.
00:33:25.147 --> 00:33:27.697
wasn't sure what kinda reaction
it was gonna get on the
00:33:27.697 --> 00:33:28.507
fence about even doing it.
00:33:28.507 --> 00:33:31.117
And then you were studying this,
this device looking closely at
00:33:31.117 --> 00:33:33.607
it, felt like for, a long time.
00:33:33.607 --> 00:33:38.047
And then, eventually we had the discussion
about the, the screen not being at the
00:33:38.047 --> 00:33:40.177
level of quality that you were hoping for.
00:33:41.107 --> 00:33:46.747
and, the, the original device was
great, in my opinion, you have sent me a
00:33:46.927 --> 00:33:48.457
replacement and the new is even better.
00:33:49.087 --> 00:33:52.957
But I think there's something about
this, that, that set the scene for a
00:33:52.957 --> 00:33:59.797
conversation around how to use AI with
the aspects of Poem/1 that help you the
00:33:59.797 --> 00:34:05.017
things that you are either have limited
capacity for or you don't enjoy doing.
00:34:05.347 --> 00:34:07.867
And that was sort of the seed of
the conversation that I thought
00:34:07.867 --> 00:34:11.437
would be so interesting to have you
on talk about through your work.
00:34:12.097 --> 00:34:15.997
But perhaps preface that, before
we get into that, maybe you could
00:34:15.997 --> 00:34:20.977
just give us a bit of an overview
of Poem/1 as a project for you.
00:34:21.997 --> 00:34:23.467
what was the impetus for it?
00:34:23.467 --> 00:34:26.677
You kind of set the scene a little bit
with ChatGPT and, sort of seeing its
00:34:26.677 --> 00:34:31.327
creative potential, but I'm curious what
went from there to actually having this
00:34:31.597 --> 00:34:35.707
idea for a device and putting it out into
the internet and getting lots of people
00:34:35.707 --> 00:34:38.167
to support that, help us fill that gap.
00:34:38.178 --> 00:34:45.138
Matt: kind of a ridiculous
story of, many twists and turns.
00:34:45.858 --> 00:34:47.718
I'll give the quick
00:34:49.818 --> 00:34:54.918
but I want to say something about
the prehistory of the AI clock it
00:34:54.918 --> 00:34:59.238
was called then, it came from what I
would call the reverse Turing test.
00:34:59.898 --> 00:35:06.738
So I was building a prototype with
a client where this is very, early
00:35:06.738 --> 00:35:08.388
again, like GPT-3 days, right?
00:35:08.778 --> 00:35:09.768
G PT 3.5.
00:35:11.148 --> 00:35:16.818
and the user needs to be able
to speak via WhatsApp to an ai.
00:35:17.568 --> 00:35:20.298
And I made something just to kind of
connect these things up, like what the
00:35:20.298 --> 00:35:24.138
prompts are, like, what it was about
doesn't really matter, it's very important
00:35:24.138 --> 00:35:27.198
in any project to get to end as quick
as possible so you can feel what it's
00:35:27.198 --> 00:35:30.948
like to have something which is not
human text appearing in your pocket.
00:35:31.008 --> 00:35:31.188
Right?
00:35:31.968 --> 00:35:32.508
that's unusual.
00:35:32.508 --> 00:35:33.288
That's gonna feel different.
00:35:33.288 --> 00:35:33.903
So you're like, what's that feel like?
00:35:35.028 --> 00:35:36.768
And made this p.
00:35:39.374 --> 00:35:42.134
You know, something comes back
on the phone and I'm like,
00:35:42.194 --> 00:35:44.324
I don't even, I built this.
00:35:44.384 --> 00:35:46.574
I don't believe that's ai, right?
00:35:46.634 --> 00:35:48.764
Like at, 'cause at the time you
would default to being human text.
00:35:48.944 --> 00:35:50.174
I'm like, I'm gonna demo this.
00:35:50.174 --> 00:35:54.914
And they're gonna think either
it's automated or they are
00:35:54.914 --> 00:35:56.834
going to, type something in.
00:35:56.834 --> 00:35:58.994
It's gonna come back and they're
gonna go, Matt is just there
00:35:58.994 --> 00:36:00.014
at the other end just typing.
00:36:00.434 --> 00:36:00.674
Right.
00:36:01.514 --> 00:36:04.814
And I've experienced this before
where I was like, I was evaluating
00:36:04.814 --> 00:36:08.774
a company which could, they claimed
they could send back iMessages.
00:36:09.524 --> 00:36:12.254
was like, how do I know it's not
just like the founder doing that kind
00:36:12.254 --> 00:36:15.284
of YC thing, like things that don't
scam, like literally typing things in.
00:36:16.094 --> 00:36:19.214
I asked it to say the alphabet backwards
and it came back really quick and I'm
00:36:19.214 --> 00:36:20.204
like, oh, okay, that's a computer.
00:36:21.614 --> 00:36:27.554
So for this prototype, I was like,
how do I prove this is actually an ai?
00:36:28.424 --> 00:36:29.234
So I was like, I know what I'll do.
00:36:29.234 --> 00:36:31.634
It's like proof of life, like, you
know, the kind of people that hold up
00:36:31.634 --> 00:36:32.924
the newspaper to go, this is happening.
00:36:33.344 --> 00:36:34.874
I'll get us to do
something about the time.
00:36:35.534 --> 00:36:38.444
And, the thing that I can't do really
quickly is make a poem about that.
00:36:38.444 --> 00:36:42.164
So I'll make a poem about the
time the user type something in.
00:36:42.164 --> 00:36:42.854
So they would go, hello?
00:36:43.544 --> 00:36:44.864
It would send back a poem about the time.
00:36:45.314 --> 00:36:47.234
So that was the thing
I did with the client.
00:36:47.594 --> 00:36:51.794
I came downstairs one morning, I have
an e ink clock on my shelf that, is an
00:36:51.794 --> 00:36:55.294
open source project from the Connected
Environments programme at, UCL.
00:36:57.164 --> 00:36:59.954
Been there for a few years and I
was like, what if I put that poem
00:37:00.044 --> 00:37:03.639
on stack thing on that clotter?
00:37:04.724 --> 00:37:09.734
So I did that and it was weird because
this is, you know, we don't really
00:37:09.734 --> 00:37:15.314
have a culture I think of like changing
texts in the world in English, right?
00:37:15.764 --> 00:37:18.584
Or even text as decoration,
like live, laugh, love, right?
00:37:18.584 --> 00:37:19.514
Like that's kind of maybe it.
00:37:19.574 --> 00:37:19.724
Right.
00:37:20.324 --> 00:37:24.644
And this text was strange because
it was really it was kind of, kind
00:37:24.644 --> 00:37:26.234
of a get up and go kind of vibe.
00:37:27.224 --> 00:37:31.094
it would lie about the
time a, a rhyme work.
00:37:33.090 --> 00:37:36.870
there was this kind of like
aspirational nature to it, and
00:37:36.870 --> 00:37:37.950
I was like, this is strange.
00:37:37.980 --> 00:37:41.400
So I sort of posted 'em on
Twitter, completely blew up.
00:37:42.360 --> 00:37:45.930
Everyone started retweeting it,
ended up very quickly in New York
00:37:45.930 --> 00:37:48.930
Times as an example of like, kind
of what people are building with ai.
00:37:49.200 --> 00:37:50.190
And I'm like, you know what?
00:37:50.670 --> 00:37:52.230
I want to get this into
more people's hands.
00:37:53.250 --> 00:37:57.630
So, I started gathering people's
email addresses, and at the time
00:37:57.630 --> 00:37:58.740
I was like, I'm gonna make a kit.
00:37:58.740 --> 00:38:02.010
I was like, I tried to make a
kit, and I was like, couldn't
00:38:02.010 --> 00:38:03.150
guarantee that people get a piece.
00:38:03.150 --> 00:38:08.820
It's like, okay, I will try and give it
away to someone, and all I want in return
00:38:09.150 --> 00:38:11.160
is one of these clocks to put on my shelf.
00:38:12.090 --> 00:38:15.390
And so I talked to a retail company
that would do some product development
00:38:16.260 --> 00:38:19.350
a manufacturer got in touch as well
because they'd seen this kind of thread
00:38:20.100 --> 00:38:22.920
they'd seen the newspaper and they
were like, okay, we'll, do it right.
00:38:23.250 --> 00:38:26.610
And then for whatever reason, they had a
change of heart, like right at the moment
00:38:26.610 --> 00:38:31.230
where we were going to go into, like kind
of the industrial design development.
00:38:31.230 --> 00:38:36.900
There were mockups and everything
and, that fell apart just like wery.
00:38:36.960 --> 00:38:37.530
They changed their mind.
00:38:38.340 --> 00:38:39.720
at that point I'm a bit like, come on.
00:38:39.720 --> 00:38:41.910
I was so close to having
this thing on my shelf.
00:38:41.910 --> 00:38:42.060
Right.
00:38:43.410 --> 00:38:46.980
And I remember looking back at my, I made
some kind of notes about how to bring
00:38:46.980 --> 00:38:51.180
it into the world you know, option A is
like a kit, option B is like, somebody
00:38:51.180 --> 00:38:52.890
else makes it, option C is Kickstarter.
00:38:52.950 --> 00:38:56.220
And I'd like, I'd crossed it out
and I was like, never do that.
00:38:56.940 --> 00:38:58.200
This is a bad idea.
00:38:59.430 --> 00:39:08.580
And I then ran into a friend Alex from
who's, founded Approach, which is an
00:39:08.580 --> 00:39:12.990
industrial designs agency, he was like,
you know what, we can help with that.
00:39:14.790 --> 00:39:17.070
And it gave them something, you
know, it was great to work with
00:39:17.070 --> 00:39:19.620
'em because
they really, like, they
work with the contract manufacturer.
00:39:19.620 --> 00:39:20.820
I know.
00:39:20.820 --> 00:39:25.740
Well, know much better now, but new
of them, they got to show off some of
00:39:25.740 --> 00:39:31.050
their thinking about AI and we worked
really closely, on the subjects.
00:39:31.140 --> 00:39:34.290
I wrote to another friend, Tom Armitage,
to help out on the firmware as well.
00:39:35.070 --> 00:39:35.670
together on that.
00:39:36.960 --> 00:39:39.720
And then it, so it's been,
you know, a bit of a kind of a
00:39:40.080 --> 00:39:41.340
team there to make it happen.
00:39:42.060 --> 00:39:44.460
then yes, as you know, eventually
I was just like, I wanna
00:39:44.460 --> 00:39:45.330
see this thing in the world.
00:39:46.410 --> 00:39:47.160
Take it to kick start.
00:39:50.190 --> 00:39:53.310
I thought that was gonna
be a very quick process.
00:39:54.330 --> 00:39:56.940
It's just an ePaper display in a box.
00:40:00.540 --> 00:40:03.300
what could possibly be
difficult about that
00:40:05.580 --> 00:40:06.075
years past?
00:40:08.558 --> 00:40:10.028
You eventually have it on your shelf.
00:40:11.498 --> 00:40:16.778
that was my evenings and weekends,
for a good period of time.
00:40:16.838 --> 00:40:20.318
And, you know, as you mentioned
about the screen, one of the things
00:40:20.318 --> 00:40:24.968
I really obsessed about was the
Matt finish on this ePaper screen.
00:40:26.318 --> 00:40:32.858
Getting that right, the QC to kind
of get that happening, really tough.
00:40:33.248 --> 00:40:34.898
I've got it here.
00:40:35.618 --> 00:40:39.008
And, but that, and that finish matters.
00:40:39.458 --> 00:40:42.548
And when I saw yours at dinner, I'm
just like, something is not right here.
00:40:43.328 --> 00:40:46.628
didn't come all this way to have
somebody with a clock where they've
00:40:46.683 --> 00:40:52.718
forgotten to put the screen film
on and, needs to make good on that.
00:40:54.548 --> 00:40:54.998
Jonny: Amazing.
00:40:55.988 --> 00:40:56.258
I,
00:40:57.033 --> 00:40:57.393
Tom: Amazing.
00:40:57.398 --> 00:41:02.078
Jonny: I got my Poem/1, I think I posted
about this, but even though I knew exactly
00:41:02.078 --> 00:41:06.728
how it was gonna work, and I'd seen all of
your updates and all that kind of thing,
00:41:07.088 --> 00:41:12.998
I held it in my hand and, you know, it,
it took two, two years, I dunno well over
00:41:12.998 --> 00:41:17.228
a year to get from me supporting that
Kickstarter to me holding the clock in
00:41:17.228 --> 00:41:21.758
my hand and all of the, the pain that
you went through, I saw all of that.
00:41:22.568 --> 00:41:26.138
but what also had happened is the
technology underpinning it had changed
00:41:26.138 --> 00:41:31.748
so dramatically that what had maybe
been extraordinary when you shared it on
00:41:31.808 --> 00:41:34.718
Twitter when you first built this thing.
00:41:36.758 --> 00:41:38.888
the, the technology itself
isn't extraordinary.
00:41:38.888 --> 00:41:40.928
In fact, it sort of feels like,
00:41:44.660 --> 00:41:47.600
I don't know if retro is unfair,
but like, it, sort of feels
00:41:47.600 --> 00:41:51.140
lo-fi compared to everything else
you can do with this stuff now.
00:41:51.890 --> 00:41:55.640
So it had a, a different impact
on me than I thought it would when
00:41:55.640 --> 00:41:57.080
I placed my Kickstarter order.
00:41:59.450 --> 00:42:02.420
and as such, it sort of feels
more meaningful, it feel,
00:42:02.480 --> 00:42:03.770
feels like a time capsule.
00:42:05.090 --> 00:42:07.940
and it, because it doesn't have to
be a demonstration of extraordinary
00:42:07.940 --> 00:42:13.040
technology, it can just be what it is,
which is a beautiful thing to, look at
00:42:13.880 --> 00:42:20.360
and a hilarious and sometimes overly s
mushy like rhyming coupler that appears
00:42:20.360 --> 00:42:23.930
on your, on your, device every minute.
00:42:24.260 --> 00:42:31.550
So yeah, I, kudos 'cause tho those
people didn't order a clock and
00:42:31.550 --> 00:42:32.540
haven't been following along.
00:42:33.140 --> 00:42:35.360
Won't have seen what it took to get there.
00:42:35.360 --> 00:42:38.990
But it is a beautiful, I've just
moved house, so I dunno where mine is.
00:42:38.990 --> 00:42:40.430
Otherwise I'll be
brandishing mine as well.
00:42:40.430 --> 00:42:42.530
But, it's in a box still, I think.
00:42:42.530 --> 00:42:42.740
But,
00:42:44.174 --> 00:42:45.884
Matt: Retro, retro is the right word.
00:42:45.884 --> 00:42:51.014
I mean, very, strange to have made like
the world's first retro AI product, right?
00:42:54.284 --> 00:43:00.464
it's AI from a more innocent time,
because, you know what, it do?
00:43:00.524 --> 00:43:02.654
It shows a rhyming couplet
for every minute a day.
00:43:02.654 --> 00:43:05.504
That's 500,000 poems a year, right?
00:43:05.564 --> 00:43:09.494
That is, you know, the highest poetry
velocity of any consumer appliance, right?
00:43:09.494 --> 00:43:11.894
Like, I dare you to find, one else.
00:43:11.894 --> 00:43:16.004
But at the same time, it was far
enough into thinking about AI that,
00:43:16.904 --> 00:43:24.974
the clocks show the same, poem time in
the day, same time for that day, right?
00:43:24.974 --> 00:43:25.424
It's different.
00:43:25.514 --> 00:43:25.814
Explain.
00:43:26.564 --> 00:43:30.584
and that's because I, you know, like even
at that point I'm like, I'm not sure I can
00:43:30.584 --> 00:43:36.134
justify data centres spinning up with the
carbon footprint and the water for like,
00:43:36.134 --> 00:43:41.744
you know, something this frivolous single
person and weirdly, that had a kind of a,
00:43:42.524 --> 00:43:45.374
it does provide a sort of an extra layer.
00:43:45.614 --> 00:43:48.824
The fact that, you know, are 700 clocks
in the world, which we're all kind of
00:43:48.824 --> 00:43:51.614
showing the same permit at the same time.
00:43:51.764 --> 00:43:56.324
Is that when you look at it, somebody
else might be
looking at the same thing.
00:43:57.824 --> 00:44:00.314
And there's a kind of a, somebody
called it an invisible social
00:44:00.314 --> 00:44:02.054
network there, which I really love.
00:44:02.114 --> 00:44:06.524
There is this kind of like,
possibility of communion someone else.
00:44:07.364 --> 00:44:07.964
And it, yeah.
00:44:08.294 --> 00:44:13.424
It's, you know, one of the, one of the
other problems about it being older is
00:44:14.564 --> 00:44:15.974
there's a whole kind of poetry machine.
00:44:15.974 --> 00:44:17.414
I mean, I'm the poet here, right?
00:44:17.414 --> 00:44:21.164
I kind of live with this on my,
kitchen for, you know, a year and was
00:44:21.164 --> 00:44:23.084
like tuning the, machine behind this.
00:44:23.504 --> 00:44:26.594
but it's built around the model,
which is very soon to be turned off.
00:44:27.254 --> 00:44:29.834
And every time I try and update
the machine, so like, move
00:44:29.834 --> 00:44:33.404
away from that model, the newer
models are just too domesticated.
00:44:34.154 --> 00:44:35.414
I can't get the variety in.
00:44:36.244 --> 00:44:36.874
Jonny: Four Zero.
00:44:36.934 --> 00:44:37.204
No.
00:44:37.574 --> 00:44:38.309
Matt: real kind of challenge there.
00:44:38.969 --> 00:44:39.189
yes.
00:44:40.244 --> 00:44:43.904
Jonny: Yeah, famously, famously romantic.
00:44:44.942 --> 00:44:47.342
Matt: Yeah, it's really tricky.
00:44:47.762 --> 00:44:52.382
So yeah, this is, now the, active
sort of like, what do I do?
00:44:52.682 --> 00:44:53.492
What do I do about that?
00:44:53.702 --> 00:44:53.852
Right.
00:44:53.852 --> 00:44:57.422
You know, having gone through, you
know, the design and the manufacturing
00:44:57.422 --> 00:45:01.562
and the sourcing and then fulfilment,
and that's where the agents come in,
00:45:01.562 --> 00:45:02.732
dealing with FedEx and all the rest.
00:45:02.732 --> 00:45:07.382
Right now I'm onto, well, how do
I, like, how do I keep this live?
00:45:07.382 --> 00:45:10.112
I'm, you know, I've got some
more in the warehouse on Shopify.
00:45:10.502 --> 00:45:11.432
All those kind of things.
00:45:12.272 --> 00:45:16.022
people will still be able to get their
hands on them, but the, the poetry is the
00:45:16.022 --> 00:45:17.342
thing that kind of really occupies me.
00:45:17.342 --> 00:45:19.172
And that was its own
journey as well, right?
00:45:19.172 --> 00:45:23.462
Like kind of to kind of make something
which was frivolous to begin with
00:45:24.272 --> 00:45:29.552
then to kind of really go through
like, you know, who is, the poet here?
00:45:30.362 --> 00:45:31.202
What is my role?
00:45:32.492 --> 00:45:33.902
The machine is not the poet.
00:45:33.902 --> 00:45:34.202
Right?
00:45:34.202 --> 00:45:37.172
Like that, you know, I kind of
say like that, you know, chap,
00:45:37.172 --> 00:45:38.942
GPT, composers, poems, or clock.
00:45:38.942 --> 00:45:40.112
Like, it's, me.
00:45:40.172 --> 00:45:43.982
I have, I stand, you know, I have to stand
behind this it has
to be intentional.
00:45:44.462 --> 00:45:46.817
And so that's, you know, that,
that was unexpected as well.
00:45:48.428 --> 00:45:49.058
Jonny: fascinating.
00:45:49.058 --> 00:45:56.408
Matt, I'd love to learn about what
Poem/1 then took you onto, and it
00:45:56.408 --> 00:45:58.988
feels like there is some connective
tissue between that and inanimate.
00:46:00.518 --> 00:46:04.388
but, in your de describe the, the
journey of this like art project,
00:46:04.388 --> 00:46:11.648
this, as you say, frivolous, you
know, beautiful, but probably not,
00:46:12.998 --> 00:46:15.188
infinitely scalable thing that you built.
00:46:16.268 --> 00:46:20.708
and how you sort of either, either
parlayed that or, moved that into
00:46:20.738 --> 00:46:22.178
inanimate and what is inanimate.
00:46:24.072 --> 00:46:27.702
Matt: I mean, one of the wild things about
doing poem is like realising what can be
00:46:27.702 --> 00:46:30.462
done with a small group of people now.
00:46:30.522 --> 00:46:34.392
You know, last time I did, know,
manufacturing with Burger, it was
00:46:34.392 --> 00:46:39.972
like that studio was full of people
from many different disciplines.
00:46:39.972 --> 00:46:42.312
And while we were making little Printer,
I think there must have been like
00:46:42.312 --> 00:46:46.872
kind of 15, 30 people kind of in the
studio and the kind of the core team.
00:46:47.502 --> 00:46:50.052
Like so lucky to be working
with so many incredible people.
00:46:50.952 --> 00:46:55.062
And the idea of doing something which
was, you know, like, like I said,
00:46:55.062 --> 00:46:58.542
there are kind of like people I worked
with that kind of like more on my own.
00:46:59.262 --> 00:47:00.492
It's, kind of ridiculous, right?
00:47:00.972 --> 00:47:06.192
And part of that is like the China
ecosystem, the Shenzhen ecosystem
00:47:06.197 --> 00:47:10.002
and the UK system ecosystem are like
way more developed than they were.
00:47:12.192 --> 00:47:13.302
to kind of work together like that.
00:47:14.202 --> 00:47:17.082
some of it is just kind of
making s smart decisions about,
00:47:19.572 --> 00:47:22.692
putting, you know, how what the supply
chain is behind different things
00:47:22.692 --> 00:47:23.952
like that, that's more sophisticated.
00:47:24.822 --> 00:47:29.862
and I think one of the, you know, aspects
of doing Poem/1 was it just kind of said
00:47:29.862 --> 00:47:34.332
to me, it is now possible to think in
credibly about hardware in the world
00:47:35.562 --> 00:47:36.882
from a kind of a startup perspective.
00:47:36.882 --> 00:47:39.522
And maybe kind of, it hasn't
been from the last like 10 years.
00:47:40.242 --> 00:47:45.222
I think the other thing it made me
really feel is like, it is magical
00:47:45.522 --> 00:47:51.462
to have something which is kind of
like which isn't kind of inside an
00:47:51.912 --> 00:47:54.642
app window in the world around you.
00:47:55.602 --> 00:47:56.382
How does that work?
00:47:57.012 --> 00:48:01.962
And since the beginning of doing poem,
I really started asking myself like,
00:48:03.097 --> 00:48:10.152
how do you have AI in the world in a way
where, you know, regular people use it.
00:48:11.262 --> 00:48:14.052
And there are all kinds of like
really interesting issues when you
00:48:14.052 --> 00:48:15.102
start digging against that Right.
00:48:15.162 --> 00:48:17.082
You know, interaction.
00:48:17.142 --> 00:48:20.082
How are you gonna interact
with, you know, in the world?
00:48:20.532 --> 00:48:25.842
How do you have really low later,
you know, kind of inside the 150
00:48:25.842 --> 00:48:28.452
milliseconds, kind of like cognitive
loop of kind of instant in.
00:48:28.752 --> 00:48:31.932
How do you do that when inference is
at least a kind of a network half away?
00:48:34.302 --> 00:48:35.712
how do different things
kinda work together?
00:48:36.192 --> 00:48:37.482
how do you generate more interfaces?
00:48:37.752 --> 00:48:40.752
And it was on those kind
of questions where I.
00:48:42.888 --> 00:48:46.368
I started developing some technology
and some opinions about the
00:48:46.368 --> 00:48:48.648
interactions would start answering that.
00:48:49.068 --> 00:48:56.538
And I think this would've gone nowhere
except for an old trends Daniel, came
00:48:56.538 --> 00:48:59.388
over for dinner because he was kind of
thinking about kind of similar issues.
00:49:00.408 --> 00:49:03.018
we started talking about like, okay,
but what does this mean in the world?
00:49:03.888 --> 00:49:05.088
How would we actually do this?
00:49:05.088 --> 00:49:06.378
Like, how would that come into the world?
00:49:07.098 --> 00:49:08.808
And outta those
conversations came Inanimate.
00:49:09.468 --> 00:49:11.208
We were like, we think we
can actually do this, right?
00:49:12.018 --> 00:49:16.578
we can turn the room around us
into being a, ui, which increases
00:49:16.578 --> 00:49:17.538
the agency of people in it.
00:49:18.648 --> 00:49:23.358
but the way we'll start is with
objects that people understand as
00:49:23.358 --> 00:49:26.868
products that they kind of allow,
you know, can be understood and
00:49:26.868 --> 00:49:28.218
get invited into people's lives.
00:49:28.458 --> 00:49:31.668
And we'll start in a quite
humble way and we'll take it, you
00:49:31.668 --> 00:49:32.988
know, sort of like step by step.
00:49:33.738 --> 00:49:35.848
And was the, origin of Inanimate.
00:49:35.868 --> 00:49:39.408
And we saw, you know, raised
a bit of exploration money and
00:49:40.128 --> 00:49:41.058
been the story of the last year.
00:49:42.138 --> 00:49:47.148
But equal parts design, technology,
you know, product design, understanding
00:49:47.148 --> 00:49:50.418
how people understand things and
and kind of knowing what is now
00:49:50.418 --> 00:49:51.718
possible with the Shenzen ecosystem.
00:49:53.854 --> 00:49:55.894
Jonny: So what can I just
ask about exploration money
00:49:55.894 --> 00:49:57.634
that has piqued my curiosity.
00:49:59.344 -->
00:50:04.564
as much as you can share, like what was
your story that you told the investors
00:50:04.564 --> 00:50:08.644
of the angels or institutions or
whoever helped with exploration money?
00:50:08.644 --> 00:50:14.614
Like what was the, were you
able to tell a story of a grand
00:50:14.614 --> 00:50:16.564
vision and an enormous business?
00:50:16.564 --> 00:50:20.554
Or was this actually just, there's
something interesting here and,
00:50:21.634 --> 00:50:23.254
we would love to, explore it.
00:50:23.254 --> 00:50:26.284
Like what, was the, how
did you go about that?
00:50:27.140 --> 00:50:33.590
Matt: so we were, we pretty much did
this kind of digging our own until
00:50:33.590 --> 00:50:35.720
there were a few technology prototypes.
00:50:37.070 --> 00:50:41.240
when we start using some of that, and one
of the core bits here is a of software
00:50:41.240 --> 00:50:46.220
that we open source called Resident, which
lets you sort of hot reload, new behaviour
00:50:46.220 --> 00:50:47.480
into products kind of on the flight.
00:50:48.650 --> 00:50:53.540
And between that a couple of other
explorations, we were like, all
00:50:53.540 --> 00:50:55.100
products are going to work this way.
00:50:55.250 --> 00:50:58.130
Like, it just became clear like
you start using these things, it's
00:50:58.274 --> 00:50:58.394
Tom: It.
00:50:59.450 --> 00:51:00.860
Matt: this is how things
are going to work.
00:51:01.580 --> 00:51:06.860
And at that point, you know, where if you
meet the right investors who are prepared
00:51:06.860 --> 00:51:10.970
to kind of go, not only, you know what
this kind of feels like, you know, you,
00:51:10.970 --> 00:51:14.210
might be on something here, but also you
are the, people who could pull this off.
00:51:15.410 --> 00:51:17.270
That was what, that was how
the pre-seed came about.
00:51:17.750 --> 00:51:21.020
And we had to have, we have a great
group of investors involved in that.
00:51:21.680 --> 00:51:26.060
so it was very, early prototypes,
but enough to kind of go, yeah, okay.
00:51:26.060 --> 00:51:28.280
There are some kind of core assumptions
here which are kind of born out.
00:51:29.780 --> 00:51:29.870
and
00:51:30.036 --> 00:51:30.276
Jonny: But
00:51:30.320 --> 00:51:30.500
Matt: go.
00:51:30.560 --> 00:51:30.800
Yeah.
00:51:31.836 --> 00:51:35.316
Jonny: I'm a, early, I'm a, I've
just started exploring resident.
00:51:35.376 --> 00:51:39.126
I have my little M5Stick here, which
I dunno if, those people that are
00:51:39.126 --> 00:51:44.016
watching can say it makes a, it's
meant to say hello Matt, but this,
00:51:44.646 --> 00:51:47.616
this hardware doesn't, have a real
speaker, so it just beeps at you.
00:51:49.566 --> 00:51:50.016
but yeah,
00:51:50.440 --> 00:51:50.620
Matt: It's
00:51:50.646 --> 00:51:51.996
Jonny: excited to keep exploring.
00:51:52.660 --> 00:51:53.140
Tom: It shows on the
00:51:53.226 --> 00:51:53.526
Jonny: Yeah.
00:51:53.946 --> 00:51:54.846
it does say it on the screen.
00:51:54.876 --> 00:51:55.416
Yes.
00:51:56.766 --> 00:51:57.216
okay.
00:51:57.366 --> 00:52:01.836
Well I realised we skipped over, one
topic and we will cover that last
00:52:01.896 --> 00:52:03.426
before we get to quickfire questions.
00:52:03.426 --> 00:52:05.556
So, Tom, I'll hand over to you for.
00:52:08.802 --> 00:52:09.132
Tom: Yeah.
00:52:09.582 --> 00:52:10.512
A gentech fulfilment.
00:52:10.632 --> 00:52:13.062
So Matt, you, well there was so
many different directions we could
00:52:13.062 --> 00:52:17.082
take this conversation in and Poem/1
led Oh, so many doors to open.
00:52:17.292 --> 00:52:21.942
But that dinner conversation the other
night around, specifically around how
00:52:21.942 --> 00:52:25.572
you use Claude agents to help with
the fulfilment process around Poem/1.
00:52:26.562 --> 00:52:32.352
if you could share us a, a bit about
that wonderful world and, how has that
00:52:32.352 --> 00:52:37.302
experience been fulfilling Kickstarter
orders in this new era of AgTech help.
00:52:38.932 --> 00:52:41.302
Matt: fascinating things about, you
know, there, there's this kind of
00:52:41.302 --> 00:52:43.792
line like, you know, they call it
hardware 'cause it's hard, right?
00:52:44.212 --> 00:52:46.522
And weirdly, I think that like
connected hardware is not,
00:52:46.912 --> 00:52:49.612
you know, it's not difficult.
00:52:51.052 --> 00:52:54.442
you know, I meet engineers who are
dealing with things like the sort
00:52:54.442 --> 00:52:58.312
scale of the internet or the kind of
the, depth of computer science that
00:52:58.312 --> 00:53:02.062
like, I've got no way of kind of
like, you know, even comprehending.
00:53:03.532 --> 00:53:05.902
But the thing with hardware
or with product is there's
00:53:05.902 --> 00:53:07.282
just lots of different bits.
00:53:07.282 --> 00:53:08.632
It's very, broad.
00:53:09.772 --> 00:53:14.332
And one of those very, very broad
pieces is when you've got your products
00:53:14.332 --> 00:53:17.512
in the world, you've got to get 'em
into people's house, fulfil, right?
00:53:17.602 --> 00:53:19.102
You know, the factory has
made a bunch of things.
00:53:19.102 --> 00:53:21.022
They've put 'em on pallets,
they've gone into a warehouse,
00:53:22.132 --> 00:53:23.362
they've got to arrive somewhere.
00:53:25.072 --> 00:53:33.082
That to me, is as fascinating as
the ID as the word wrapping the
00:53:33.082 --> 00:53:36.232
firmer or the interactions around
what happens when we hit the depart.
00:53:38.092 --> 00:53:42.772
And I'd kind of everything else in
the process, I touched a little bit,
00:53:43.522 --> 00:53:44.872
but I hadn't thought about fulfilment.
00:53:45.022 --> 00:53:46.612
I was just like, that is
a piece of the process.
00:53:47.362 --> 00:53:48.142
I will not think about it.
00:53:48.142 --> 00:53:51.232
Turns out is a really good
ecosystem for this, right?
00:53:51.232 --> 00:53:54.982
So my, my manufacturing is in
Dongguan, which is a factory, about
00:53:54.982 -->
00:53:56.422
an hour's drive to North Shenzhen.
00:53:57.172 --> 00:53:58.762
over the water from Shenzhen is Hong Kong.
00:53:58.822 --> 00:54:02.992
Hong Kong's a really good place to
ship from it's a, you know, it's
00:54:02.992 --> 00:54:05.302
a, you know, very open borders.
00:54:05.302 --> 00:54:08.602
You can get there from anywhere in the
world, including from mainland China.
00:54:09.592 --> 00:54:13.882
and it turns out there are warehouses
which specialised in crowd funding,
00:54:14.362 --> 00:54:19.522
fulfilment and like, you know,
but later for the Shopify, you
00:54:19.522 --> 00:54:21.112
know, that kind of shipping.
00:54:21.442 --> 00:54:22.852
So I was like, you know,
jobs are good one, right?
00:54:22.852 --> 00:54:24.562
Like kind of put the
options in there back a kit.
00:54:24.562 --> 00:54:28.042
Everyone kind of gives their details
and I'll just kind of upload a big kind
00:54:28.042 --> 00:54:29.902
of CFD and Floship and send everything.
00:54:31.976 --> 00:54:32.876
Should be easy.
00:54:34.376 --> 00:54:35.246
Was not easy.
00:54:36.416 --> 00:54:41.006
The first hundred clocks, which, you
know, they're kind of the early test,
00:54:41.336 --> 00:54:45.416
they went out fine about half of the uk,
half of the us, some scattering elsewhere.
00:54:45.416 --> 00:54:49.046
Even though I made people promise I would
not shit for anywhere else in the world.
00:54:49.286 --> 00:54:52.736
People moved in the meantime,
fine, all went out fine.
00:54:53.216 --> 00:54:59.636
So then I ship out another 600, wake up
one morning bank holiday weekend, wake up
00:54:59.636 --> 00:55:08.906
one morning I've got like 15 messages from
my warehouse who are like, know, has got a
00:55:08.906 --> 00:55:11.486
question about this package that went out.
00:55:13.226 --> 00:55:14.456
can you just tell us what it is?
00:55:14.456 --> 00:55:16.046
And I'm a wrote couple of lines.
00:55:16.646 --> 00:55:22.646
And then one of them linked through to a
page on FedEx's website where there was
00:55:22.646 --> 00:55:24.476
this kind of like big long form fill in.
00:55:25.526 --> 00:55:26.756
It's like, this is weird.
00:55:26.966 --> 00:55:28.676
And I sort of grind my way through that.
00:55:28.676 --> 00:55:31.706
It's like the family was out for
the weekend, I'm like, sunny, right?
00:55:31.706 --> 00:55:33.986
So I just go and grind
through those things.
00:55:34.616 --> 00:55:38.426
Just I'm finishing like another
10 or 15 or it's like, okay,
00:55:40.076 --> 00:55:41.186
this is now getting tricky.
00:55:42.806 --> 00:55:47.696
So I think, okay, I'm gonna start,
write a little called skill for this.
00:55:48.866 --> 00:55:49.616
Drag my browser.
00:55:49.706 --> 00:55:50.636
I've never done that before.
00:55:50.786 --> 00:55:52.976
You know, been a bit cautious
about these kind of things.
00:55:53.966 --> 00:55:55.646
and I'll give it all the
background information.
00:55:55.646 --> 00:55:59.666
I'll write a little Claude skill where
it can,
drive my browser and just kind of
00:55:59.666 --> 00:56:01.946
like fill in those things for me, right?
00:56:01.946 --> 00:56:04.466
So what I'll do is I'll
just give it the URL.
00:56:04.466 --> 00:56:10.106
I'll go there, I'll wake up next
morning, another hundred I queries
00:56:10.166 --> 00:56:14.351
from FedEx and the warehouse of
the right, all slightly different.
00:56:16.913 --> 00:56:19.148
I'm like, okay, that's a problem.
00:56:19.568 --> 00:56:20.408
I can't deal with this.
00:56:20.618 --> 00:56:20.978
Now.
00:56:22.148 --> 00:56:25.898
I am not the kind of person that likes
AI agents in my notes or in my email.
00:56:26.558 --> 00:56:30.338
find it like psychologically
destabilizer, but at the same time,
00:56:30.338 --> 00:56:33.518
I do not have time to respond to a
hundred different requests for FedEx.
00:56:34.058 --> 00:56:36.698
Especially filling in
this very dedicated PDF.
00:56:36.698 --> 00:56:41.708
They gave me where they were like,
it's just, you know, a declaration
00:56:41.708 --> 00:56:46.988
form about a clock where all the
questions are we think it's a watch.
00:56:47.438 --> 00:56:48.128
Is it a watch?
00:56:48.128 --> 00:56:48.863
No, it's a watch.
00:56:48.968 --> 00:56:49.928
How big is the movement?
00:56:50.108 --> 00:56:52.358
Ah, sounds like the movement.
00:56:52.508 --> 00:56:54.728
The movement is the, bit of
the thing that makes it go that
00:56:54.728 --> 00:56:56.198
you look right, not the case.
00:56:56.918 --> 00:56:58.508
that's really small boxes.
00:56:58.958 --> 00:56:59.408
PCP.
00:56:59.408 --> 00:56:59.618
Right?
00:57:00.068 --> 00:57:00.878
It's really small.
00:57:00.938 --> 00:57:02.138
That means it must be a watch.
00:57:02.228 --> 00:57:02.708
Are we right?
00:57:03.518 --> 00:57:04.983
It's like not a watch, it's a clock.
00:57:05.138 --> 00:57:06.458
How many jewels does it have?
00:57:06.788 --> 00:57:07.748
It doesn't have any jewels.
00:57:08.108 --> 00:57:08.468
Right.
00:57:08.648 --> 00:57:11.738
And so they, I think they
thought I was smuggling watches.
00:57:13.058 --> 00:57:17.528
And so I built, ended up building a
whole bunch of board skills to like
00:57:17.798 --> 00:57:22.388
deal, like read my email, bring out
all these things, and then people
00:57:22.388 --> 00:57:25.568
start getting the clocks and they start
asking me questions, support questions,
00:57:27.068 --> 00:57:28.268
forgotten about all of that as well.
00:57:28.268 --> 00:57:28.418
Right.
00:57:29.108 --> 00:57:31.958
And you think there's a few of them, but
there's a kind of a scale problem, which
00:57:31.958 --> 00:57:35.078
is just like people have very innocuous
questions or something doesn't arrive.
00:57:36.728 --> 00:57:40.118
or like perhaps something is broken
or the firmware's not quite right.
00:57:40.538 --> 00:57:44.798
There was a new networking problem
that arose between the test units and
00:57:44.798 --> 00:57:47.258
the new ones just be tracked down.
00:57:47.258 --> 00:57:48.338
So I'm tracking all of these.
00:57:48.728 --> 00:57:49.928
I'm like, I can't do this.
00:57:52.104 --> 00:57:53.454
MIT train plot can help.
00:57:53.454 --> 00:57:58.194
Again, the problem with agents is
you give 'em a problem, they grind
00:57:58.194 --> 00:58:00.354
every single problem into dust, right?
00:58:00.354 --> 00:58:02.334
They don't know where the
bottom is, and they try and work
00:58:02.334 --> 00:58:04.704
everything up from first principles.
00:58:06.144 --> 00:58:10.614
At least this is my experience,
and they can work a bit
00:58:10.614 --> 00:58:12.444
differently in different cases.
00:58:13.494 --> 00:58:17.964
So want that, know, I don't want it
to kind of like accidentally miss
00:58:17.964 --> 00:58:23.469
an email one time, reply to somebody
twice, you know, decide one answer
00:58:23.604 --> 00:58:24.954
for one person networking problem.
00:58:24.954 --> 00:58:27.444
Another one for somebody who's
maybe got kind of like the screen
00:58:27.444 --> 00:58:28.464
finishes and what they wanted.
00:58:28.824 --> 00:58:29.754
I need policies for that.
00:58:30.834 --> 00:58:36.414
So what I started doing started thinking,
what does my agent need to work?
00:58:38.034 --> 00:58:39.384
let me, actually just show you, right?
00:58:39.434 --> 00:58:40.994
This is my ops screen.
00:58:41.294 --> 00:58:46.424
So this is a website which is not meant
for humans, it's just a picture for me,
00:58:48.704 --> 00:58:50.174
to, see what the agent is seeing.
00:58:51.314 --> 00:58:54.554
So my Claude communicates with
the backend website via an MCP.
00:58:55.424 --> 00:58:56.534
says MCP served for this.
00:58:56.534 --> 00:58:57.249
That's the way it interacts.
00:58:57.249 --> 00:59:01.604
This is a nont interactive website,
this is basically like a cache of
00:59:01.604 --> 00:59:04.054
all the data, coming in from Floship.
00:59:04.084 --> 00:59:08.594
My warehouse BackerKit, which kind
of manages, you know, backer surveys
00:59:08.594 --> 00:59:10.244
and collecting fulfilment reads.
00:59:10.244 --> 00:59:11.444
My email caches it there.
00:59:11.924 --> 00:59:15.014
And then it triages issues
which are appearing in email.
00:59:15.794 --> 00:59:18.644
it has a kind of a basic CRM,
which is basically a linking table.
00:59:18.644 --> 00:59:22.934
The idea is that like there's very,
little beta kept here, but it's here
00:59:22.934 --> 00:59:24.824
to maintain state for the agent.
00:59:25.694 --> 00:59:27.974
And I'm not gonna go into any
of these because I don't wanna
00:59:27.974 --> 00:59:30.255
kind of reveal kind of PII.
00:59:31.184 --> 00:59:38.174
but Tom, we've got this one right, which
is, this is a very, ugly screen, right?
00:59:38.174 --> 00:59:39.254
But it's not meant for me.
00:59:39.974 --> 00:59:44.594
This is a log of my, like
pulling out from my email.
00:59:46.544 --> 00:59:49.454
these are kind of pull them from Floship
so I can see kind of what the orders are
00:59:49.454 --> 00:59:50.834
and I can kind of
track 'em through here.
00:59:51.584 --> 00:59:56.414
there's a kind of an LLM written kind of
description about like what the, going on.
00:59:56.414 --> 01:00:00.734
And I can kind of run this every
couple of days and triage, new issues.
01:00:01.364 --> 01:00:05.564
I also have, you know, I have a kind
of a stats page over here where I
01:00:05.564 --> 01:00:09.014
can see where clots are being turned
on, you know, if people are upgrading
01:00:09.014 --> 01:00:10.094
their firmware and things like that.
01:00:12.614 --> 01:00:19.634
this is interesting, I think because it's
the first time I've built something where
01:00:19.664 --> 01:00:25.784
the important bit is I'm building new
ground troupe, and then the agent itself
01:00:25.784 --> 01:00:27.284
is figuring out how to operate on that.
01:00:28.004 --> 01:00:30.074
But it's not having to figure out
things through first principles.
01:00:30.524 --> 01:00:34.724
So the agent skill started
as, read this via the MCP.
01:00:36.494 --> 01:00:38.054
see if you can identify issues.
01:00:39.704 --> 01:00:45.974
If you have a way of dealing with it
in a document called the SOP Stand
01:00:45.974 --> 01:00:48.104
Operating procedures, then do that.
01:00:48.134 --> 01:00:48.824
Just apply that.
01:00:48.824 --> 01:00:50.294
Ask me to confirm before
you send anything.
01:00:50.594 --> 01:00:54.704
If you don't, then propose what our
policy should be, approve it, add
01:00:54.704 --> 01:00:56.324
it to the SOP, and then apply it.
01:00:57.044 --> 01:00:59.894
And over time the SOP has grown.
01:01:00.224 --> 01:01:03.794
So there are now a number of very
detailed policies about how to
01:01:03.794 --> 01:01:07.244
deal with different issues, which
don't have to be figured out again.
01:01:07.244 --> 01:01:08.834
And it's that kind of like compounding
01:01:09.424 --> 01:01:10.024
Jonny: Self-healing.
01:01:10.744 --> 01:01:11.044
Yeah.
01:01:11.594 --> 01:01:15.794
Matt: Yeah, I couldn't have
run this stage of the process
01:01:16.454 --> 01:01:18.254
without Claude doing this for me.
01:01:18.914 --> 01:01:22.514
I also couldn't have done it had I
just let the agent loose on my email
01:01:22.514 --> 01:01:24.524
and loose on the warehouse APIs.
01:01:25.304 --> 01:01:28.364
there is something about, and I think
this is a kind of an interesting pattern
01:01:28.364 --> 01:01:33.194
to kind of go, I mean I think about
this sometimes as like context plumbing,
01:01:33.674 --> 01:01:35.294
but it's structured context, plumbing.
01:01:35.684 --> 01:01:39.014
How do you make sure things flow to
the same place so that doesn't become
01:01:39.014 --> 01:01:42.944
a new source of truth, but is just
pulling together source of truth from
01:01:42.944 --> 01:01:47.414
elsewhere and then have the agent
operate on that, but in a way where
01:01:47.414 --> 01:01:49.724
it has kind of compounding processes.
01:01:50.114 --> 01:01:52.424
And that's something I'll be
carrying into kind of other projects.
01:01:52.424 --> 01:01:52.784
Definitely.
01:01:53.904 --> 01:01:56.544
Jonny: Feels like a, a, a,
an agent in that sense is a
01:01:56.544 --> 01:01:58.164
human, but just so much faster.
01:01:58.164 --> 01:01:59.964
And moreau autonomous in a sense.
01:01:59.964 --> 01:02:04.284
Like we've been building dashboards to
look at data for however long because if
01:02:04.284 --> 01:02:08.124
you were to look at the raw data, you'd
make the wrong inference or you'd, you
01:02:08.124 --> 01:02:11.634
know, I wouldn't be able to write the
sequel in a way that I would be confident
01:02:11.634 --> 01:02:13.884
enough to know that the data was right.
01:02:13.974 --> 01:02:21.294
So I need a data scientist or, a data
analyst to write, to make sure that
01:02:21.294 --> 01:02:25.314
the dashboard is the correct data
and then I can infer a decision or
01:02:25.314 --> 01:02:28.944
make a decision about a new product
direction or whatever it is from that
01:02:28.944 --> 01:02:30.114
dashboard, because that's trusted.
01:02:30.354 --> 01:02:36.414
And it feels like what you've done here
is basically, had to do that because you
01:02:36.414 --> 01:02:43.344
have this incredibly fast AI employee that
will just make all the wrong decisions
01:02:43.344 --> 01:02:45.294
incredibly quickly and not question them.
01:02:45.684 --> 01:02:47.904
So you have to force them into
this box where they're only
01:02:47.904 --> 01:02:49.014
looking at this one thing.
01:02:49.824 --> 01:02:54.024
so I suppose it's in that sense, it's a
human nature amplified or something, but.
01:02:54.570 --> 01:02:57.660
Matt: Yeah, I think one of the
interesting things I suppose is that
01:02:57.660 --> 01:03:00.300
I've moved from dashboards to reports.
01:03:01.230 --> 01:03:04.860
that kind of overview is really
kind of for me to trust that the
01:03:04.860 --> 01:03:06.450
data is flowing in the right way.
01:03:06.750 --> 01:03:07.050
Right.
01:03:07.050 --> 01:03:10.230
And if like the email is kind of
like the email such as broken, I can
01:03:10.230 --> 01:03:14.250
it like blow away the cache, change
the way that comes in, you know,
01:03:14.370 --> 01:03:15.990
that's all very deterministic, right?
01:03:16.590 --> 01:03:19.680
But for getting a kind of an overview
about what's going on, I don't
01:03:19.680 --> 01:03:21.210
rely on anything deterministic.
01:03:21.780 --> 01:03:25.500
I get the agent to look at everything
and there's another skill which
01:03:25.500 --> 01:03:26.760
is give me my daily report.
01:03:27.630 --> 01:03:33.360
And what that does is it can intelligently
prioritise it can take into account kind
01:03:33.360 --> 01:03:35.880
of like the severity of the issue, how
long it's been kind of sitting there
01:03:35.880 --> 01:03:40.080
is with a customer who's been kind
of messed around before, and so on.
01:03:40.230 --> 01:03:44.760
And you know, this, this item's been stuck
in Belgium for like six weeks, right?
01:03:44.820 --> 01:03:49.140
It's kind of
happened or to go kind
of, oh, there are two parcels which are
01:03:49.140 --> 01:03:50.700
meant for this particular destination.
01:03:50.790 --> 01:03:52.110
Do they have a kind of a commonality?
01:03:52.890 --> 01:03:56.400
are a whole bunch of, deliveries
being caught up in Oakland, right?
01:03:56.400 --> 01:03:59.520
And I was able to kind of identify
that, but it's that intelligence that
01:03:59.520 --> 01:04:03.090
wouldn't have surface fire, a dashboard,
which kind of like where the structure
01:04:03.090 --> 01:04:06.960
presupposes the information, where
what I can do instead because I have
01:04:06.960 --> 01:04:10.890
this ground tree, this give the agent
the soup and go, this is the output
01:04:11.190 --> 01:04:14.145
report I want about how you feel about
n is gonna be different day to day.
01:04:15.660 --> 01:04:17.760
And that's been a really
interesting process as well.
01:04:18.300 --> 01:04:20.640
'cause I think the tools
that we use to interact with
01:04:21.330 --> 01:04:22.410
agents like this is different.
01:04:22.530 --> 01:04:27.600
It reminds me kind of, of how I imagine
offices worked in the 1950s, right?
01:04:28.980 --> 01:04:32.640
it's kind of, it's reports based to
somebody at a desk who's responsible
01:04:32.640 --> 01:04:33.840
for this information going through.
01:04:34.740 --> 01:04:37.530
And I wonder like how much we
have to learn from kind of like
01:04:38.670 --> 01:04:43.260
fashioned information, people
flowing offices instead of computers.
01:04:43.260 --> 01:04:45.300
Like how we collaborate with these things.
01:04:51.108 --> 01:04:57.198
Tom: Yeah, many, so many examples there
that you, of really interesting agentic
01:04:57.228 --> 01:05:01.848
use cases that are also beyond code.
01:05:01.848 --> 01:05:06.648
It feels like coding, agentic
coding is one true use case that has
01:05:06.648 --> 01:05:09.678
really been nailed, but most people.
01:05:09.858 --> 01:05:14.088
But I think some of the examples
you gave there are this, almost like
01:05:14.088 --> 01:05:19.038
operating system for how to even
build the foundation of a company
01:05:19.038 --> 01:05:23.688
or a product or a team on top of, so
lots of really inspiring things there.
01:05:23.838 --> 01:05:28.968
I know we are well over time, so we
should probably switch to, final, quick
01:05:28.968 --> 01:05:30.438
fire questions if that's okay with you.
01:05:30.498 --> 01:05:35.298
Unless there's anything else, any last
thoughts you wanna share on that topic?
01:05:35.508 --> 01:05:37.068
AgTech fulfilment overall.
01:05:37.106 --> 01:05:39.326
Matt: be remiss if I didn't
put in a couple of plugs.
01:05:39.331 --> 01:05:39.631
Is that okay?
01:05:40.699 --> 01:05:40.989
Jonny: Okay.
01:05:42.782 --> 01:05:43.082
Tom: for it.
01:05:43.220 --> 01:05:47.510
Matt: plug number one is there are
gonna be more blocks for sale because
01:05:47.510 --> 01:05:50.300
I didn't, I overproduced the batch.
01:05:51.320 --> 01:05:54.560
so there are
some, which didn't go
to Kickstarter backers, and that is
01:05:54.560 --> 01:05:55.910
waiting for me to set up the Shopify.
01:05:55.910 --> 01:06:00.260
So please, if you go to Poem Town, there
is a link on that page to Substack.
01:06:01.040 --> 01:06:03.710
sign up with a, you'll find out
first about when to get those clocks.
01:06:04.460 --> 01:06:08.030
And if you wanna follow along with
inanimate, we have some things coming up.
01:06:08.960 --> 01:06:14.180
go to news tech and subscribe,
like, and subscribe.
01:06:14.180 --> 01:06:17.000
As I say, like we're beginning to be a bit
more vocal about what we're doing right.
01:06:17.450 --> 01:06:20.150
And, that's, an interesting story as well.
01:06:20.150 --> 01:06:21.860
So I hope people enjoy
following along with that.
01:06:23.248 --> 01:06:23.728
Jonny: Fantastic.
01:06:24.178 --> 01:06:26.848
I'll have to have you back in
a year or, something and, see
01:06:26.848 --> 01:06:27.838
what's happened with Inanimate.
01:06:28.283 --> 01:06:28.733
It'd be awesome.
01:06:28.874 --> 01:06:29.444
Matt: Would love to.
01:06:30.104 --> 01:06:32.294
Now, I've been looking forward
to these quickfire questions.
01:06:32.396 --> 01:06:33.086
Jonny: Oh God.
01:06:34.334 --> 01:06:36.584
Matt: on, 10 hooks about this.
01:06:40.002 --> 01:06:41.112
Tom: Okay, let's kick off.
01:06:41.112 --> 01:06:43.332
Maybe we could do a couple each,
Jonny, and then wrap up there.
01:06:43.992 --> 01:06:47.952
So I, we'll kick off with, when
a GI takes over everything.
01:06:48.342 --> 01:06:50.217
What will you do with all
of your free time Matt?
01:06:50.442 --> 01:06:50.892
Matt: You know what?
01:06:50.892 --> 01:06:55.692
Probably exactly what I'm being I'm doing
now, it will just feel like a very kind
01:06:55.692 --> 01:07:01.722
of like pursuit, bit like kind of like
hiding in the woods and doing the things
01:07:01.722 --> 01:07:02.812
that the machines could be doing anyway.
01:07:06.032 --> 01:07:06.272
Jonny: Them,
01:07:06.368 --> 01:07:08.138
Matt: put, like, why do
people still play chess,
01:07:08.292 --> 01:07:08.742
Tom: Of that.
01:07:08.798 --> 01:07:09.458
Matt: question, right?
01:07:09.458 --> 01:07:09.638
Like,
01:07:10.112 --> 01:07:10.412
Jonny: yeah.
01:07:10.448 --> 01:07:11.828
Matt: are way better at
chess and like, what?
01:07:11.828 --> 01:07:12.728
Why did it, because it's fun.
01:07:13.958 --> 01:07:17.648
So I think I get very fulfilled out
of what I'm doing at the moment.
01:07:18.218 --> 01:07:18.983
So yeah, probably the same.
01:07:20.176 --> 01:07:23.746
Jonny: Feel like that gives you,
hope for the human condition or that
01:07:24.436 --> 01:07:27.526
we will find fulfilment generally
as a species because it's not just
01:07:27.526 --> 01:07:29.176
all about efficiency at all costs.
01:07:31.306 --> 01:07:33.046
Matt: Yeah, I'd like to
be optimistic about that.
01:07:33.046 --> 01:07:33.351
I hope so.
01:07:33.872 --> 01:07:34.172
Jonny: Yeah.
01:07:34.382 --> 01:07:34.682
Yeah.
01:07:35.012 --> 01:07:41.192
so my, quick fire question is, if
there's one product that you wished
01:07:41.192 --> 01:07:45.852
you'd created physical, digital
in the world, what would that be?
01:07:47.914 --> 01:07:51.544
Matt: I am, this is not the
product, it's the creative.
01:07:52.924 --> 01:07:55.024
there's the product friend friend.com,
01:07:55.264 --> 01:07:58.024
which is the AI pendant
where you speak to it.
01:07:58.354 --> 01:08:00.154
And now with version two,
it kind of speaks back.
01:08:01.234 --> 01:08:06.814
And, I think of the, there's a few
kind of overlooked areas and ai
01:08:06.814 --> 01:08:10.444
and one of them is kind of how we
use AI as a kind of a, a companion.
01:08:11.104 --> 01:08:14.434
Like it's only been done in the kind of
a, a game, like a playful way so far,
01:08:14.434 --> 01:08:16.804
but nobody's taken it really seriously.
01:08:16.804 --> 01:08:22.384
Friend is doing that and their
recent creative, which was all you
01:08:22.384 --> 01:08:26.014
know, analogue v analogue sound,
and tiny vignettes, which showcased
01:08:26.014 --> 01:08:28.024
the product but don't centre it.
01:08:28.294 --> 01:08:32.464
They're more of the lives of
people around it is just inspiring.
01:08:32.464 --> 01:08:35.134
It's absolutely beautiful
and absolutely pitch perfect.
01:08:38.000 --> 01:08:40.220
Jonny: It must because they spent
a lot of money on friend.com.
01:08:40.580 --> 01:08:42.020
That's, my main memory of friend.
01:08:42.020 --> 01:08:43.010
I don't have a friend.
01:08:43.010 --> 01:08:46.100
I mean, if I have some
friends, I don't have a friend.
01:08:46.760 --> 01:08:47.930
Always a very difficult thing to say.
01:08:47.930 --> 01:08:51.885
I assume when you're talking
about, Tom, over to you.
01:08:55.738 --> 01:08:59.968
Tom: And my last question, you've
given us a tonne of examples today of
01:09:00.028 --> 01:09:04.648
things that you have built recently,
but I have to ask you what the silliest
01:09:04.648 --> 01:09:08.543
thing is that you've prototyped like
with ai, purely because you could.
01:09:10.594 --> 01:09:15.964
Matt: There have been so many ridiculous
projects, but they often, they predate ai.
01:09:16.744 --> 01:09:22.054
I think my, one of the hardest projects
was I made a vending machine to sell books
01:09:22.174 --> 01:09:23.614
that would tweet when it sold the book.
01:09:25.684 --> 01:09:26.314
was great.
01:09:26.314 --> 01:09:29.854
I took it on tour around like various,
kind of like lobbies in London that
01:09:29.854 --> 01:09:31.684
was in like it was in Google for that.
01:09:32.404 --> 01:09:37.234
And, it would they would stock
books that were recommended by
01:09:37.234 --> 01:09:38.314
people working in the building.
01:09:39.874 --> 01:09:41.584
It weighed 270 kilogrammes.
01:09:41.584 --> 01:09:42.814
It was almost impossible to move.
01:09:44.074 --> 01:09:50.284
My house was full of vending
machine parts for years, and I
01:09:50.284 --> 01:09:55.144
was
like, do not pick up hobbies,
which like won't sit through doors.
01:09:56.074 --> 01:09:58.324
It's a, a ridiculous tube.
01:09:58.324 --> 01:10:01.294
But that was pre ai, so I'm not allowed,
I'm not allowed to have that one.
01:10:01.294 --> 01:10:06.274
But yeah, like the in the string of
very many ridiculous perceived selling.
01:10:07.848 --> 01:10:08.178
Jonny: Amazing.
01:10:08.178 --> 01:10:09.858
I think you're allowed that.
01:10:09.858 --> 01:10:12.138
I don't think, I don't think
there's an AI constraint on this.
01:10:12.138 --> 01:10:13.578
I think you're allowed
your vending machine, Matt.
01:10:16.173 --> 01:10:17.283
Tom: Yeah, definitely.
01:10:17.283 --> 01:10:19.533
Although I am curious to ask the
follow up question of if you're
01:10:19.533 --> 01:10:23.883
doing that today, is there any
aspect of that AI helped you with?
01:10:24.813 --> 01:10:26.943
AI unlocked the possibility
of doing another
01:10:27.240 --> 01:10:28.410
Jonny: Forward, move the vending machine.
01:10:28.773 --> 01:10:29.403
Tom: around London?
01:10:29.580 --> 01:10:31.290
Matt: Yeah, it would've
made technology much easier.
01:10:31.290 --> 01:10:34.650
So vending machines internally have
a protocol, which is a bit like USB
01:10:35.190 --> 01:10:39.330
called, MDB, I think was invented by
Coca-Cola, Japan in the mid 1970s.
01:10:39.690 --> 01:10:42.450
And the technology spec is like 300 pages.
01:10:42.540 --> 01:10:45.570
It was very difficult to integrate with.
01:10:46.200 --> 01:10:48.330
I think that it would make much easier.
01:10:49.530 --> 01:10:51.360
but you know, it wouldn't
help with the other problem.
01:10:51.360 --> 01:10:55.440
The vending machine, which was his thing
to tell you about vending machines, right,
01:10:55.740 --> 01:10:57.870
is they're fridges basically, right?
01:10:58.440 --> 01:11:01.950
So if you turn the refrigeration
element off, 'cause it's noisy, what
01:11:01.950 --> 01:11:04.500
happens is it just gets and warm.
01:11:06.330 --> 01:11:09.420
The glue to hold my shelf talkers
on, which is the technical name for
01:11:09.420 --> 01:11:12.660
the cards underneath the books that
say like why they're good started
01:11:12.660 --> 01:11:15.720
melting and all the shelf talkers
would fall off after about three days.
01:11:16.620 --> 01:11:20.220
Finding a formulation of
adhesive to stop them falling off
01:11:24.570 --> 01:11:25.650
took me weeks,
01:11:28.110 --> 01:11:28.830
side the seller tape.
01:11:28.860 --> 01:11:30.870
If you're interested the best stuff,
01:11:31.752 --> 01:11:32.022
Jonny: Go.
01:11:32.070 --> 01:11:33.810
Matt: to a dozen others
before I landed on that.
01:11:34.620 --> 01:11:35.730
AI couldn't have helped with that.
01:11:36.210 --> 01:11:39.175
So there's, you know, there's utility
for us humans yet, that's what we can do.
01:11:41.514 --> 01:11:42.144
Jonny: The meat layer.
01:11:42.834 --> 01:11:43.104
Yeah.
01:11:44.253 --> 01:11:44.763
Tom: Love that.
01:11:44.913 --> 01:11:45.153
Love
01:11:45.324 --> 01:11:45.594
Jonny: right.
01:11:46.104 --> 01:11:47.094
We should, finish there.
01:11:47.424 --> 01:11:49.884
Matt, thank you so much for joining us.
01:11:49.884 --> 01:11:56.694
This is, a monster edition and so glad it
was learned so much and really enjoyed it.
01:11:56.694 --> 01:11:59.064
So, maybe we'll have you back.
01:11:59.124 --> 01:12:00.444
I would love for that to be true.
01:12:00.809 --> 01:12:07.134
but in the meantime, best of luck
with inanimate and and thank you.
01:12:08.692 --> 01:12:09.752
Matt: Jonny Tom, thanks for having me.
01:12:09.917 --> 01:12:10.447
Been a pleasure.