# Founder review: the first reading of the book

**Date:** 22 August 2026 · **Source:** voice memo, transcribed by Otter.ai, reproduced
verbatim below (per the house rule: the founder's voice is source material and is not
edited) · **Status:** Review **r001** on [the reviews register](../reviews/index.html),
normalised into nine numbered items with the agent's comments and proposals at
[/reviews/r001.html](../reviews/r001.html). No content changes were made from this memo
directly: per the review workflow, the agent comments and proposes first, and changes
land only after agreement, item by item.

**What it says, in one paragraph:** the verdict on the first reading is strongly
positive, and the consequence is a decision to invest: the book becomes positioning for
the wider work, developed organically at scale. The requested changes: retitle to
**Fractal Semantic Graphs: Meaning Through Connectivity**, because the fractal concept
permeates everything and must anchor the book; make every chapter carry an interesting
idea that challenges the reader, going back to source documents where concepts were
diluted; add an homage to the lineage (the Semantic Web, graph theory, graph databases,
ontologies of ontologies and taxonomies of taxonomies); correct the overstatement that
we do not use databases (the truth is no live databases: serverless and ephemeral
engines, including in-browser SQLite and RDF querying in the vaults, are used and
demonstrated); add many screenshots, especially of the vaults, which deserve a section
of their own; add case studies (SGit itself, IssuesFS and its issues-of-issues
ontology consensus model, EmailFS and EmailFS Lite, and the agentic team's own
organisation as a graph); and run all of this through the review workflow being built,
including a written review of the review, a stated approach, a list of what more is
needed, and an agentic team with a researcher on this project.

---

## Verbatim transcript

Dinis Cruz 0:01

Okay, so this is a review of the my first reading of meaning to connectivity book that
we created from the graphs website, and and you know I guess my first comment is it's
fucking brilliant, right? Like it's amazing, right? Basically, it's really cool to take
a lot of that research, a lot of those ideas, a lot of those concepts, and basically a
package it on a website that will allow agents to consume it, but b you know create a
book the way we're creating it and with a workflow. And basically, what this convinced
me is actually it's actually worth me now spending a bit of time just cleaning up this
book because I think this book could also be a really good positioning for a lot of the
projects that we're working on, and a lot of the concepts that I've been promoting can
actually be explained here in terms of reality, and in terms of also something I've
tried before, but I think now we can scale, which is a much more organic way to to
develop a book, right? Especially now that we have, you know, these amazing editors and
this amazing ability to scale a lot of the the parts of the creation of the book that
was were a bit harder before. So, so a lot of the key concepts, and it's incredible,
like even the first version, how coherent it is. But there's a number of things that
I'm missing. So I want to just list them, sort of as a stream of consciousness. I took
a bunch of screenshots, which I think will support some of these. But I think I can
just list them all here now. Some of these ideas. So okay. So I think one. I love the
concept, the name, meaning to connectivity. Of course, that's the thing. But I feel
that we we need the other part, which is also the part that I start talking. Which also
I think it's an important one to anchor what's going on here, right? And and basically
is the what's it called? The the fractal semantic graphs. So I actually think the the
name of the book because I was thinking we have to have graphs in the name, right? But
actually the name of the book really should be fractal semantic graphs, and then colon,
and then meaning to connectivity, because this fractal concept is a concept that we'll
see permeates throughout a lot of the stuff that exists, right? And a lot of the things
I've been doing because it's and again it's one of the key concepts that I want to
describe. So I think that's important. Now the other thing that I feel this book should
be about is this book should be about sharing the concepts and the core ideas and
screenshots and evidence that we have that I have created, but in a way that every
chapter, every page should have an interesting idea, should challenge the reader,
should have you know, because a lot of stuff I talk about are stuff that the
interesting concepts, but they haven't got a lot of exposure, right? So we we have
enough backing materials, and I feel there's a lot of documents that I've created
already that have a little bit of been distilled and been dilute a little bit. So let's
go back to them to make sure that some of these concepts really shine through, right?
And this is where again now the diff is very important. And in fact, what you want,
what I would like you to do is this is a good example of you should write a review of
this, then you should write the approach that you would take to do these fixes,
including what else you need and what more information, so we can have a conversation.
We can even have a I can have a chat, a full-blown chat session, right? With you know a
new version of Claude, right? Just about this, so that we can create the brief for you
to make the changes, because part of what we're developing here, like I mentioned my
last memo, is the workflow. I think that's what we want to nail here. We want to nail
what does this workflow does that actually allows us to create a really powerful book
here, right? And so going back to the concept, so the fractal element is very important
because you'll see that just about everything I have talks about it. And one of the I
think the most interesting concepts that I arrived at is this idea of you know graphs
of graphs of graphs of some ontologies of ontologies and taxonomies of taxonomies of
taxonomists, right? Because we should you know this is basically saying that, and I
think it's important that in this graph we pay homage, right, and acknowledgement to
the amazing work that was done before. Like I was massively inspired by semantic
graphs. I think semantic graphs are amazing. The and then I think I wrote about it.
Right, there's a document that. And again, if you need documents, please list them now
in a sort of hey, we need more information here and here and here, or let's do more
research here and here and here, and then let's fire up an agent to do that. Again,
that should be part of the workflow, right? Like almost like a research project. In
fact, we should be setting up the agentic team on this project, which would actually
include the researcher, right? So, so yeah. So we should acknowledge semantic graphs.
We shouldn't acknowledge the amazing work of graph theory, right? You know, and you
know a lot of these, again, I'm not saying that graph databases are important. They are
amazing, but I just build that our serverless approach, the the the graph structure
where everything is a graph, and you hyperlink things and you connect stuff and you you
sort of extract meaning through through connectivity. Right, that that whole zooming is
very important, and I think we have a lot of examples where that is in action, and and
I think, and that's what I think we should have. And I think the the the book, in a
way, should have these sort of really powerful examples of the stuff that I've been
developing, like you know the the s the sgit, which is fundamentally you know, an
encrypted version control, you know, graph-based structure, right? That everything is
encrypted on the server side, but uses graphs as a way to think about systems, and it's
actually a really nice way to explain how to think about a graph, right? But in in
reality, like the way that Git S git is structured, we can actually use it to explain
what that graph looks like because again, when you talk about graphs or graphs or
graphs, it's interesting to then think that we we can navigate through that sort of
SGIT environment, right? And I give it a whole bunch of other properties, right? What
we have there, right? So, so the other thing that's important to acknowledge here is we
can make a lot of questions and we can actually do a lot of follow-ups with the
individual agents that we have, right? And it's important to to capture, you know,
those agents to, for example, to show the way that we have even the main project, the
way it is organised. If you think it's a graph, right? Like if you think about like
even the first, you know, version of this, where I asked the librarian of that project
to do a research, to go and do the documents and find the first set of documents that
we then used that we then created for this, that's that's a graph in itself, right? And
and even like the way we we keep now both the what's it called we keep both the this
the web pages and the the print PDFs in synchronisation. That again it's a graph,
right? Actually, another really important project to read and to map out, and I think
I'll make it into a separate website because I think it deserves to be independent. And
also remember that the reason I'm also doing a lot of these things is so that it's
easier for you to also find the key evidence for what's what exists and what I have
created. So, yeah. So one of the projects I was talking about is this project called
Issues FS, which actually was one of the first interesting ones to really connect this
sort of fractal element because you can have issues of issues of SS of issues of SS of
issues of SS, and and and it really has a lot of cool documents that explains how you
can create ontologies of ontologies and how they all sort of you know relate to each
other and how you can have an ontologies define a top level who then gets overreaching
at lower level and then eventually we arrive at the consensus right so I think that
deserves a whole section of itself and the whole explanation and and also how it works
especially the fact that we created it so that it works just of a CLI works off a file
system works of different places, and I think that's that's a really cool case study to
talk about because it actually shows how that thing can work in action. Another really
good example is the email FS that I created, and the email FS again, and especially the
email FS Lite, which we created in some of the projects, are really powerful because it
shows again the power of creating these graph-like structures that allow things to
communicate and things to be linked. Another So another thing that is really missing
from here is is screenshots, right? Like we really need to do screenshots of
screenshots and top a lot of screenshots on the side, and and we need to basically show
the put them in the book, right? Because I think that will really bring a lot of things
to life. I feel that there's a lot of concepts that, unless you slap a screenshot in
there, it won't make any sense, right, to people, right. And again, a screenshot is
worth 1000 words in these days, and and in this particular situation, I think having
lots of screenshots would really make a massive difference here, right, and how
everything fits together. So, yeah. So let's let's add them, because because again,
that's part of the workflow, right? It's part of our figuring out the best way to
publish this. So if you if you take a look at the great work that the Skit or AI has
done in the vaults, for example, you see that so especially there's a section there
that describes what the vaults do, and I think we should be showing that. For example,
every time we talk about a vault, we should have a screenshot of it, and I think we
should actually have a whole section just on the vaults because they they are real-life
examples of a lot of these graphs, right? So that's really powerful. Another mistake
that was made in some of the comments is when we say we don't use databases. I use, in
fact, I've used graph databases, right? In fact, most basically, I don't. What tends to
happen is you. What I mean is that I don't have live graph databases, and that's
important. So, for example, one of the vaults, and let's find it, has a really cool
example of a full-blown Spark and RDF sort of query, and even writing MySQL. There's
one even has a SQLite on the browser, but it's a good example of we load stuff, we
delete stuff, we load on demand. So it's almost like you have these ephemeral. It's
more about serverless and ephemeral infrastructure, and again, ephemeral is a great
example of of riverless, and yeah. So we start putting all this together, and then do
it. Let's do a review, and then let's do the next version.

Transcribed by https://otter.ai