Inkwell: Authors and Artists

    
Why do you assume that these responses were programmed rather than
extrapolated from things ChatGPT learned from its training data? 
  

    
There is usually deliberate randomness in a large language model's
output to prevent it from going into an infinite loop. (In the
developer version, the amount of randomness is a parameter you can
set called the "temperature.") If we were being scientific about our
poking and prodding, we'd want to run the same query many times and
see what the distribution of responses is. Otherwise you can read
intent into responses that are just a roll of the dice.

Here are a few of my non-scientific attempts, using GTP-3 directly:

> What do you think?

> I think it is important to consider the ethical implications of
any new technology before introducing it into our lives. We must
consider how it may impact our lives and the lives of those around
us. We must also be aware of any potential risks associated with the
technology and weigh them against the potential benefits.
Ultimately, we must make sure that any new technology is designed
and implemented in a way that is socially responsible and ethically
sound.

Pretty generic. Let's add context:

An interview with a Catholic theologian:

> Q. What do you think?

> A. I think that the Catholic Church has many beautiful teachings
and traditions that have been passed down through the centuries. I
believe that the Church has a rich history and a deep appreciation
for the teachings of Jesus Christ, which should be respected and
followed. I also think that the Catholic Church is a great source of
spiritual guidance and support for its believers, and that it is
important to recognize and cherish its traditions.

Or how about:

> From an interview with a Democratic Presidential Candidate:

> Q. What do you think?

> A. I believe that we need bold, progressive policies that will
make a real difference in the lives of everyday Americans. We need
to focus on creating jobs with good wages, access to quality
healthcare, investments in our crumbling infrastructure, and an end
to systemic racism and discrimination. We must also ensure our
economy works for everyone, not just the wealthy few.

It often doesn't understand its role, though:

> From an interview with Vladamir Putin just after the Russian
invasion of Ukraine:

> Q. What do you think about the situation in Ukraine?

> A. The situation in Ukraine is deeply concerning. We are deeply
concerned about the escalation of violence and the suffering of
civilians in the region. We are doing our utmost to contribute to a
peaceful resolution to the crisis, and we remain committed to
finding a diplomatic solution that respects the sovereignty,
territorial integrity and independence of Ukraine.

Other disappointing role-playing failures omitted. Maybe the next
version will do better?
  

    
<karish> fair point - I made an assumption lacking, perhaps, the
exact kind of context <bslesins> is illustrating. 

Not knowing enough AI science - and specifically the details of
ChatGPT’s architecture and training data scope - we are left to
guess what inherent biases might be under the hood, coloring the
decision-making and framing of responses. 

They might not be explicit blue-vs-red so much as whether one valued
epistemology bubbles up higher than another by the fact that the
engineers studied one particular philosophy of code architecture
more than another, or developed code more often in agile shops than
in waterfall-governed organizations. 

Speaking of coding philosophies, here’s some recent thinking on how
quickly and in what direction the tech will be industrialized: 
https://www.linkedin.com/feed/update/activity:7017481818761224192

Should we consider what a capitalized AI might do versus something
that’s distributed open-source?
  

    
Ok, another break from sandbagging.

Brian Selinsky (#43):  Waymo uses geofencing.  = no corner cases. 
They don't pretend to have a general model.

Bruce:  I think a real part of the problem is that neither Zuck nor
anyone else who matters has any sense of history or what that might
offer him.  Like the first time someone tried to put a bunch of
avatars around a conference table (Active Paper, 1995) and how well
that went over (not). After everyone was done putting on VR headsets
and puking their guts out, I'd argue that the best things to come
out of VR1.0 were both artworks, and I bet Pat Lichty remembers: 
Char Davies' "Osmose" and Brenda Laurel's "Placeholder", which was a
true shared space and thus part of a notional metaverse, although
how Zuck would monetize it is not clear.  No matter how I try, VR2.0
still looks to me like a bunch of people repeating mistakes made 30
years ago in attempts to sell into a dead market.  Games, hell yeah.

Bruce:  Detourn Bruno Latour:  We were never normal.  And bring on
the Dr. Brown's celery tonic!

Epiphany engine:  that's as good a description of the ACTLab
framework as I've seen.

Virtual Sea Monkey (#79):  We still have loudness wars.  They've
just gotten more subtle.

Jane Hirschfield (#124):  I've been drilling down into the long
covid figures with some epidemiologists.  If you include a fudge
factor for misdiagnosed cases and the ones that didn't make it into
the database, you wind up with the number of people likely to
require some sort of serious medical care for the rest of their
lives being scarily high -- high enough to siphon off significant
medical resources for decades and impact standards of care for the
entire population.  Or maybe not... it's just extrapolation.

Mackreed (#125):  Timnit Gebru called them stochastic parrots.
  

    
I tend to read from various nooks and crannies of the current
cyberspace, as well as from mainstream sources. Speculations and
philosophies are all over the place. I think about the old Buffalo
Springfield song, "For What It's Worth": "Nobody's right if
everybody's wrong." You might also say "Nobody's right, if
everybody's right." 

We're thinking about chaos right now, because we now have visibility
into chaos, into various streams of contradictory thoughts and
interpretations that were previously remote and unavailable, because
we lacked the means to access them as readily as we can with the
now-mainstream Internet.

I see smart assertions that contradict other smart assertions. Is it
important to settle on a belief? In my case, I meditate (not often
enough lately) and I try to see what's real. It's essentially human
to try to interpret and judge what we build internally from our
limited senses, but there's so much that we don't see and don't
know. 
  

    
A couple of interesting articles/papers here alleging that "science"
has become less "disruptive" and also "innovation" has become less
"useful."

Not just in this stricken decade, mind you, this alleged decline
seems to have been going on for quite a while.

In other words, instead of being a hard-charging,
scientifically-driven, high-tech society, we're basically Jonathan
Swift's Laputa; we're hopeless, otherworldly nerds and geeks who
waste our time trifling with meaningless software and sterile
business models while the world goes to hell in a handbasket.

https://americanaffairsjournal.org/2022/11/web3-the-metaverse-and-the-lack-of-
useful-innovation/

https://www.nature.com/articles/d41586-022-04577-5
  

    
#129 --- char davies! osmoses still never equaled...
  

    
I'm not sure I believe the premises of either one of these articles
-- scientists love to moan while business people love to say
business is bad -- but it seems like typical life for any oligarchy.
Why should you bust your ass "disrupting" and "innorating" when the
end game is to get acquired and get folded into the status quo of
the operating system? 

 Ideally, in circumstances of modernity; you want no findings or
products at all.  Instead, just get ChatGPT to write you a
non-fiungible token proposal, then rocket to the moon, pull the rug
and cash out.  How can you beat that?  It's perfect.

The Nature piece is mostly about science papers (because those are
easy to analyze).   I'm not convinced that "science" has all that
much to do with science papers.  We've got plenty of climate-science
findings that would disruptively set the world on its ear if
inconvenient truths had any political clout.  However, they don't.

Also, and very interestingly, more money for science R&D, and bigger
tech for Big Tech, those don't seem to help matters at all.  In fact
the apparent wealth seems to be a major part of the problem.
  

    
I'm reminded of the business wisdom of Jeff Bezos, who famously said
that at Amazon every day should be day one.  You should never
surrender the gusto of a start-up -- no matter how big or important
the company might become.

This doctrine implies that, as you get richer, you need to take
ambitious risks commensurate with your company's  size and power. 
Not that you deliberately put yourself in more danger (as Elon Musk
seems to like to do), but that you innovate on an ever-larger social
scale.  

For instance, instead of trying to create a new product, like a
cuter talking clock, you try to create an entire new industry
sector, like Amazon Alexa with its ecosystem of app developers
contributing "Alexa skills."

If you've built a failed clock-radio, it's just another defunct
gadget, but if you threw away ten billion for ten years trying to
creat ean entire tech ecosystem, and that doesn't happen, that's a
hole in the landscape like a coal-mine falling in.   Most start-ups
die, so if you always behave like a start-up, it makes it appear
that you and all your peers are always at the brink of collapse.

That doesn't much feel like "innovation," even if it innovates. 
It's like life at the mercy of a feudal overlord whose word is law,
but he might die of pox next Wednesday.
  

    
Alexa is an interesting case study--what changes because a person
has one or doesn't? What has it changed that some number of people
now have smart homes, smart appliances, Ring camera doorbells, and
is it leading to the frog not noticing that we move into an
ever-heating surveillance culture?

Separately, functioning eco-systems have niches up and down the
scale of the visible and charismatic. Much of what is being looked
at in this conversation is mainstream culture, media-covered
culture, capitalism-modulated culture. The presumption that making a
tidy sum of money and/or accumulating a tidy fiefdom of power, or at
the very least countable clicks, is *the* driver and motivator. 

What significant things might be worth noticing if one looked in
other directions?  Wes Jackson's Land Institute breeding perennial
grains comes to mind, or the use of mushrooms or microbes to address
toxic wastes. Historically, the monks of Lindisfarne keeping their
books might be an example. By definition, the non-celebrity
experiments are hidden from broad view. 

I wonder what potential engines of change this group of people may
know about, in these realms. To speak about these as 'disrupter'
forces (perhaps for the good) is to adopt a certain
language/attitude/ set of constructs. But sometimes evolution's long
experiment throws out an interesting alternative that doesn't prove
cancerous. Being rather pessimistic about the current state of the
world, I'm not counting on such unexpected recombinatory
possibilities to emerge, in science, in culture, in social compact.
But hoping that under the roiling and thrashing and performative
planet-cide some tendrils are taking root that may yet grow.
  

    
Punctuated equilibrium is as good a model for
scientific-technological discovery as it is for evolution. There was
a huge burst of disruptive discoveries and accomplishments in the
mid-20th century - atomic energy, DNA, the laser, the transistor,
plate tectonics, space travel, supply chain management, cybernetics
- whose implications are still being worked out and implemented.
That burst of innovative accomplishment was driven by extraordinary
funding and motivation that was provided by the drive to win World
War II.

I don't think we should expect that rate of disruption to be
continuous any more than we should expect a repetition of the
revolution in physics of the first two decades of the 20th century.
Applications of these new technologies have had world-shaping
impact: greatly increased agricultural production, very rapid
advances in medical capabilities, a global panopticon from
space-based observation of the Earth's surface. We look at only a
small fraction of the disruption brought by applied technology when
we focus on the aspects that have made the obvious differences in
our personal lives: cell phones, online publishing and commerce,
social media, ubiquitous compute-heavy solutions.
  

    
> we're hopeless, otherworldly nerds and geeks who
> waste our time trifling with meaningless software and sterile
> business models while the world goes to hell in a handbasket.

Amen. Or, "I resemble that remark."

(And amen to the last two posts, by <karish> and <jh>.

Re Jane's post, I'm trying to think if I know any potential engines
of change that are percolating. There might be something within the
"Game B" framework, though I don't know how much activity there is
within that frame, and how much coherence.
<https://www.gameb.wiki/index.php?title=Game_B>

>>>
Here are some different constructions that point to Game B:

1) Game B is the flag on the hill for an omni-win civilization that
maximizes human flourishing.
2) Game B is the environment that maximizes collective intelligence,
collaboration, and increasing omni-consideration.
3) Game B is building or developing the capacity to navigate
complexity without resorting to complicated systems.
4) Game B is establishing coherence within complex systems.
5) Game B is a meta-protocol for hyper-collaboration.
6) Game B is the infinite game where the purpose is to continue
playing. Game A is the finite game where the purpose is to win.
7) Game B is the theoretically optimal condition for creative
collaboration and, thus, for maximal innovation.
8) Game B must orient its primary innovation capacity towards
cultivating individual and collective sovereignty. It must foster
awareness of how choices show up and are decided, more than it
augments individual and collective power.
9) Game B is a new mode of societal, economic, and political
organization that leverages people's authentic, long-term interests
towards a healthier, more cooperative society and improved
well-being. A Game B system is any cooperative, mutually-beneficial
system that can outcompete exploitative, adversarial systems through
manifest appeal and willful, voluntary participation.
<<<
  

    
re: <129> I don't think it's a fatal flaw that Waymo is geofenced.
It's less geofenced than a train or a bus line (within its service
region) and more geofenced than taxis, but it could do its bit as
long as the number of destinations increases, and they pick up the
pace a bit.

Each mode of transportation has its own (built) habitat. These
habitats get larger or smaller based on what happens to
infrastructure. For example, bikes are probably never going to be
practical everywhere cars and trucks are (considering the distances
in farm country), but they can be useful in more places as the
infrastructure improves.

Though, infrastructure can deteriorate and be repurposed, too. Bike
paths that used to be railways are pretty common.
  

    
<137> Game B seems pretty big on top-down theorizing. Are there
examples somewhere?
  

    <scribbled by bslesins Sun 8 Jan 23 12:37>
  

    <scribbled by bslesins Sun 8 Jan 23 12:37>
  

    
Helpful links in the 'projects' section of the wiki:
<https://www.gameb.wiki/index.php?title=Category:Projects> 

Incidentally I'm not advocating Game B - I thought that framework I
pasted above was interesting, but in practice I think GameB is more
talk than action, and it has issues referenced in this critique:
<https://medium.com/the-abs-tract-organization/game-b-ullshit-3d9294d75e23>

GameB has a political aspect, and potential connection to
intellectual dark web actors. All of this stuff is worth discussing,
because it supposedly offers a path forward when "GameA" collapses.
  

    
Meanwhile:

Bolsonaro supporters breach security barriers, break into Brazilian
Congress and presidential palace
<https://www.cnn.com/2023/01/08/americas/brazil-bolsonaro-supporters-breach-con
gress/index.html>
  

    
But despite the weird politics and failings of GameB, there might be
projects within that framework that could be interesting. OTOH, it
might be smoke and mirrors.
  

    
It is a different model. To even conceive that this kind of thinking
could out-compete Game A (rather than follow its collapse), as the
last bullet point says, would be a precursor to that being possible.

The stories we humans tell ourselves about who we are affect who we
are and what we do. The reward system for pro-social choices would
have to feel more compelling than Game A winning. There are
side-cultures in which this is the case Some traditional indigenous
communities still. And a few opt-in cultures-- monastic communities,
Doctors Without Borders, soup kitchen volunteers, service
communities from bake sale parents to Habitat for Humanity,
volunteer firefighters are a few disparate examples that come to
mind. These are people choosing not to maximize profit or prestige
in how they relate to the world of others. Surely not all of them do
this because they are incapable of obtaining the alternative set of
rewards. 

Thinking about this conversation before it started up this year, I
was interested to hear what would be said/thought about the leading
edge developments--AI, politics, whether the pandemic is leaving
anything changed and how. It's a bit odd to me that there's been
little or no talk of Ukraine. But Ivan Illich's idea of "shadow
work" is pretty entrenched in my practice of noticing, and these
dark-horse realms seem worth looking at also. Particularly when the
foreground landscape is so very dispiriting. And looking with
different vocabularies and dictions and stories and materials seems
to me also a needed and useful augmentation. 
  

    
jane, if you look at post #98, the person who might have brought the
most re-orienting to the war on UKR is in the hospital with c19.

the war on UKR is what i've been most obsessed with in the past yr
--- but it doesnt lend itself to prophecy or whatever futurists like
to name as what they do.
  

    
I think we should talk about Ukraine, for sure, but I have a set of
posts first...
  

    
One of the problems with GameB, I think, is that the participants
don't really have a handle on how deeply GameA has burrowed into
their own heads, and the extent to which they play and depend on
that game. When you're sitting in the comfort of a home with
persistent electrical and water service, it's hard to imagine being
without. We in Texas learned a bit of that lesson in February 2021,
when winter storm Uri knocked out our utilities, and in Austin where
I live, that cascaded into a water system failure that lasted for
days.

We take a lot for granted, and when those folks talk about an end to
"GameA" and replacement with GameB, I wonder if they realize the
level of pain and uncertainty an end to GameA would entail. Same for
the baby fascists in the USA right now, who are intent on tearing
down the government - not realizing how much their lives and
livelihoods, their health and safety, have depended on that same
government system.
  

    
The idea of starting a movement is compelling, but how does it work?
I watched a whole new way of life emerge as the Internet
mainstreamed and everybody moved online, and I've seen a few things
that looked like emerging movements to some extent catalyzed by
computer mediated communication. I've seens some cool things emerge,
like net activism via organizations like EFF and their adherents,
and I've seen some terrible things, like Trumpism and Q Anon. I
still don't know how to intentionally start a movement - those
things seem to form, the way you see rainclouds coming from nothing
on weather radar.

We tried to start something...
  

    
In the mid-2010s RU Sirius, Krist Novoselic and I attempted to start
what we referred to as the Open Source Party - i.e. a political
party based on the principles of the open source software movement. 

Here's a summary I wrote for the Open Source party:

>>>
The Open Source Party is a political movement that derives both
inspiration and methods from Open Source software principles. The
fundamental Open Source principles as they apply to this party,
where laws, policies, and political processes are seen as a body of
code, are:

Transparency: the code, and any changes to the code, are visible and
understandable.
The code is accessible and modifiable.
Inclusion: anyone can access the code and propose modifications,
which may be accepted by democratic consensus, or by executive
decision in a framework decided democratically.
As a matter of scope, we limit our activity to the United States but
encourage development of a global Open Source party that creates
models to work in other national contexts.

Our effort is meant have a democratizing transformative effect that
is fair to all. We're committed to uses of technology to create
platforms that will support our mission.
<<<

<https://web.archive.org/web/20160220050752/http://opensourceparty.net/>