Dear Pjotr,
On 2026-06-02 06:13, Pjotr Prins wrote:
We simply have to get smarter about models. Economics will drive
that and an energy crisis helps (in that way). Having 8+
competing initiatives world wide does not help either, or maybe it
does :)
Economics is too narrow a focus, given wider dynamics.
The history of enshittification from these players is strong.
Looking at their sudden rush to sell shares in their ventures doesnt
bode well -- Winter may be around the corner.
Its evident by recent history regarding major supply chain holdups
concerning Hormuz.
The functionality of information society can not afford to be dependent
on the peculiar habits of global leaders.
The revelations about the Fast16 cyberweapon make very clear how
impeccable the OS stack needs to be.
Its entirely possible that there could be war among Atlantic allies in
our lifetime (seemingly from a tantrum).
Considering the pernicious effect backdoors could have for today's
ecosystem (akin to TETRAs weakening of the TEA1 cipher), there are
numerous playbooks which a nefarious nation state (or well funded
adversary) could deploy.
Whatever happens, its important for Guix to be a source of integrity.
Automating package writing in Guix is a boon for *our* projects. It
*is*
saving us significant time with better outputs. I get things done I
could not do before, both in terms of my scientific effort and my
software engineering effort. I need to use AI for my work. There is no
other way (any more). We are planning a massive refactoring of
GeneNetwork - we would not have considered before. There are projects
in Guix we have not touched that could benefit. I would love to use
true open models and create models deterministically with free software
instead of the industry tooling, so we get back our transparancy and
control. That is where our effort should be going. Our effort should
not go to knee jerk statements about why AI is bad for us. Guix is
sending the wrong message and educating people in the wrong way.
Beyond your wider successes and intuition,
its important to reflect on what you have to say - as your domain has
benefited from machine learning technologies, in ways most people would
struggle to understand.
You have managed to sway Arun regarding professional approaches for
using LLMs in any case:
https://codeberg.org/guix/guix-consensus-documents/pulls/13#issuecomment-15442889
(as a heuristic, I can handle disagreeing with either of you - both
requires more effort!).
We are part of *change* and we should make it abide by our ideas of
controlling our
own destiny. I don't mind a policy, per se, or even a pledge, but I do
mind
being made to look like an idiot. And that from someone who's
ancestors where partial to famous church splits ;).
Forgive me for prejudicing your ancestors but allow me to speculate on
the period you may be referring to.
The Reformation constituted a socio-technological movement initiated by
the shockjock, Luther.
Here is one of his shitposter (enshittification?) memes he commissioned:
https://commons.wikimedia.org/wiki/File:The_Papal_Belvedere.jpg
Its goals centered on:
* autonomy of local practitioners over the centralised appararatus of
large institutions;
* the removal visual and emotive stimuli and distractions;
* the prohibitive costs of largescale centres that were an opportunity
cost of tradespersons and local resources (build a cathedral, lose a
forest);
* undoing exploitation from superstitious practices and its
commercialization of indulgences (tokens);
* disintermediating intermediaries and having full control of the means
of production;
* Thumbing his nose at the Diet of Worms.
To bring things closer/back to economics, its worth considering the
history of design and logistics during times of major disequilibrium.
This was a factor which contributed to Irish collapse during Án Gorta
Mór:
```
The traditional Irish fishing boat, the currach—a lightweight,
hide-covered or canvas boat—was designed for nearshore fishing and short
distances. While ideal for navigating rocky coasts and dangerous surf,
it was unsuitable for offshore fishing or large hauls. Its limited
capacity and vulnerability to Atlantic weather, especially in winter,
meant it could not sustain the fishing operations needed to serve the
inland starving population.
To make matters worse, the onset of the potato crop failure triggered a
collapse of this already fragile coastal economy. With no crops to sell
and no cash reserves, many fishing families were forced to pawn or sell
their nets, lines, and even boats to buy food or pay rent. These tools
were often the only tangible assets they owned. Once sold, they were
gone for good. And without gear, even the meager ability to fish
vanished.
Inland, where rivers and lakes were full of trout, salmon, and eels, the
story was the same: denial of access. Under British law, freshwater
fishing rights belonged to the landlord. These rights were jealously
guarded. Tenants were forbidden to fish without permission.
Getting caught poaching from a landlord’s river could result in
eviction—a death sentence during the famine. So, while Ireland’s rivers
ran full of life, they were often locked behind absentee landlords’
walls. Landlord privilege was protected even as people perished in
sight of sustenance.
```
https://praoh.org/starving-in-a-sea-of-plenty-the-fish-story-that-whitewashes-the-irish-famine/
A community needs to be mindful regarding sustainable behaviours -
assuming a world in flux.
Failure to respond correctly results in bentdouble people building paths
to knowhere:
https://www.frrandp.com/2025/01/famine-roads.html
I would not be surprised if Guix being a refuge against AI has major
upswing opportunities -- Guix should be more bearish than bullish.
I had already alluded to security matters above.
Lisp could easily serve as the goto for people eshewing convenient modes
of code creation - which may result in some interesting outcomes.
Providing an environment whereby somebody can use and develop Guix
without any artificial crutches should be a hard criteria - if it isnt
already covered in GCD008.
Kind regards,
Jonathan
