Inkwell: Authors and Artists

    
Via email from Jamais Cascio:

> "i feel optimistic in spite of the many dark things happening,
optimistic that we are at the beginning of a new era.”

I’ve been asked, more often than I can count, whether I’m a
pessimist or an optimist. My normal answer is “Both. I’m a
short-term pessimist and a long-term optimist. We actually know what
we need to do to make the world better, but we’re having a hell of a
time getting to the point of being willing and able to do it.”

But Josh Berger’s comment, via <magdalen> (hi, T! Really glad you’re
a part of this), made me think of previous times in history where
we’ve found ourselves at the beginning of a new era. Were those
causes for optimism, pessimism, or something else? All of the above?
Because I absolutely agree that we are on the cusp of something new,
culturally and politically, driven by/mediated through/changing the
shape of our technologies. Paulina Borsook is absolutely spot-on
that technology doesn’t make people better and kinder, but our
technologies do allow us to express our kindness in new ways, to
different audiences and new families. 

Whatever this new era turns out to be, part of it will almost
certainly be entirely new pathways for expressing that kindness,
even if those pathways are also thoroughfares for malignancy. 

Or maybe I just need more coffee.
  

    
> what we think is reason shifts

So too, perhaps, 'what we think is technology shifts.'
  

    
Via email from Matthew Battles:

Thanks for your generous response to my questions, James—and to
everyone in this thread for the sharing of ideas, questions, and
hopes. 

"[T]o let ones' eyes adjust--and listen to those who've been
outside"... this does seem the right way to meet the new darkness.
It feels like it's been a bit of a breakthrough year for
minoritarian futures, for indigenous perspectives on futurity in
particular. Are there NDN fictions, first-nations speculations,
afro- and other-futurisms that folks are following? 

It's notable that amid so much sublunary confusion, it's also been a
year of looking to the heavens. Over against roadsters in space, I'm
inclined to privilege the visit of 'Oumuamua, the first interstellar
object to be detected and identified as such. Beyond the wonders of
Ultima Thule, Mars, and the gravity-echoes of colliding black holes,
however, the hygiene with which we approach stargazing also seems in
need of a healthy readjustment. The urge to speculative fabulation
in our astronomical imaginary needs to be nourished not only by new
phenomena & discoveries, but new/old perspectives. What do you all
see when you look to the stars? Whose constellations ring true for
you now; whose celestial navigation do you trust?
  

    
a friend just appointed me codirector of her hindsight institute

(https://www.prabapilar.com/hindsightinstitute/ — and yes it's
mostly a joke.


so in that spirit, of everything old is new again, was reminded of
george stewart's 'earth abides' — where a passing observation by the
protag-survivor was that he now avoided dogs: they had gone feral
and grown sleek feasting on the remains in hospitals in urban
centers.

this came to mind in fall  2018i i saw a documentary about dogs gone
feral in tierra del fuego. argentine program of moving ppl from
urban centers to tierra del fuego to work in electronics (yes,
right) meant folks arrived there with little infrastructure nor
knowledge of rural life. so they let their dogs run free or
abandoned them — and how quickly these dogs became feral, destroying
the traditional sheepherding, and messing with the ecosystem
overall.

fireland dogs

https://www.brownpapertickets.com/event/3609429

evidently this is a global problem, feral dogs in the urban/wildland
interface.
  

    
Hey everyone, first post, though I've been reading along all week,
as I also do every year.

The first thing I thought of when I saw we'd have multiple featured
guests this year was how for the first time this year the Kennedy
Center Honors honored a *team* instead of just the usual
individuals.  That was the four co-creators of Hamilton - watch the
documentary about the creation, ("Hamilton's America" PBS), to see
why it's well deserved.

I am really enjoying each of your contributions, as well as how it
seems to change the dynamic of the topic from being a firehose of
early-warning observations and reports from the field to more of a
conversation.
  

    

thanks, keta. hope it's fun to read.

so, you guys, how 'bout that Singularity? that's the moment when
artificial/machine intelligence outstrips that of humanity, and/or when
machine-based superintelligence gets so, like, *super*, that it wreaks
massive changes on human civilization. 

Bruce says it's dead. various smart scientist people have proclaimed that
it isn't possible, because, ya know, humans are soooooo special and
ultrasmart, nothing can outpace us. 

one of my friends in the AI field claims we've already reached singularity.
old news, he says, adding that the USA and silicon valley are pretty
inconsequential compared to Asia and the Middle East in terms of technology
and culture.

<tvacorn> above says: 

>  if the
>  singularity was really dead then I doubt people like Gates would be
>  pushing this macabre vision in which we literally outsource the
>  essence of our humanity ? our emotional life along with compassion
>  and empathy  ? to highly sophisticated machines

which Sherry Turkle at MIT has been warning us about for years.

maybe the issue is how to define or frame the singularity. have we created
outboard wetware that precisely mimics the human brain? nope. have we
created a giant machine 'intelligence'-of-sorts that we now serve, rather
than the other way 'round? i'd say we have. and perhaps that's what the
singularity really is: the moment we start serving the robots.
  

    
"have we created a giant machine 'intelligence'-of-sorts that we now
serve, rather than the other way 'round? i'd say we have. and
perhaps that's what the singularity really is: the moment we start
serving the robots."

How exactly do you see humans serving machines? I'm missing it.

The technological singularity assumes that an artificial
intelligence would become a learning system that would run through
cycles of improvement sufficient to make it a superintelligence that
would surpass all human intelligence. 

The difficulty I have with this speculative concept is that nobody's
provided a definition of "intelligence" - in fact it's defined many
ways for many contexts. And, if only for that reason, there's no
clear way to compare machine intelligence to human intelligence.
While machines do simulate human intelligence, it's questionable
whether a machine thinks like a human thinks. I don't think anyone
quite knows how the latter works anyway. 

I'm not afraid that Colossus is going to take all power, that humans
will be subservient to computers.

My concern would be that we delegate responsibilities to computers
that they ultimately don't handle well.  Jake had a speculative
design concept that involved a future where AIs had replaced mayors
in various cities, with results to disastrous that the practice was
abandoned - I think  outlawed - though AIs could still serve as
consultants. Maybe he'll have more to say... 
  

    
To have a Singularity, you're supposed to have searingly fast
advances in processing power, memory and bandwidth that go on
indefinitely.  It's not going to happen.

On the plus side, I was watching AlphaZero play chess an entire year
ago during the previous State of the World. Now I'm still doing
that, there's better analysis of what the thing is up to, I'm
enjoying it more than I did and I'm even starting to appreciate its
beauty.
  

    

Thanks for revisiting this topic <magdalen>. Lest anyone think that
the Singularity and transhumanist "dream" is passe, a quick read of
a November article in the "New Yorker" indicates otherwise. The
article talks about DARPA's ongoing attempts at brain/machine
interfacing via experimentation on human subjects. It describes how
a paraplegic woman who was the subject of one such "successful"
experiment eventually ended up supporting R&D efforts to develop
thought-controlled military aircraft. 

The article – a bit of a puff piece -- did remarkably little to
question the ethics or wisdom of such experiments but did go on to
briefly describe earlier research in using embedded brain tech to
control human emotions and social behavior: "After the race riots of
the late nineteen-sixties, two Harvard neurosurgeons proposed that
neural electrodes could be used to quell social violence. In 1972, a
Tulane psychiatrist used them to try to create “heterosexual
arousal” in a gay man." 
  

    
Good piece to bring up-- but I read it as sounding a VERY clear
warning about the militarization of the research. Presented as
augmenting the capacities of the injured, the article stressed
rather strongly that it could by the way be used to augment the
capacities of battlefield robots and/or soldiers. That it was a kind
of conceptual bait and switch that allowed the ethics to go
unquestioned, I took that as no small part of the point. 

But the piece wasn't particularly about letting the machines make
the decisions, and then the decisions about the decisions.

 
  

    
>My concern would be that we delegate responsibilities to computers
that they ultimately don't handle well. 

Which gets us right back to Facebook and social media algorithms...

Another example would be electronic trading bringing down global
financial markets.


I think your whole post <82> is fascinating Jon, because as you
query the idea of a technological singularity, what comes up for me
is how you also touch on the really interesting parts of the Gaia
hypothesis.

>The technological singularity assumes that an artificial
intelligence would become a learning system that would run through
cycles of improvement sufficient to make it a superintelligence that
would surpass all human intelligence. 

If you just replace the word "technological" with "biological" and
the word "artificial" with "distributed" you get a sentence that
could be Lovelock.  Yes, the entire living system became
superintelligent and self-regulating, surpassing humans long ago.

And as far as nobody providing a definition of intelligence, what
has been fascinating me this year has been places where intelligence
has been pointed out to me that I had completely missed precisely
because I had too rigid a definition of intelligence.  Trees think. 
Forests think.  Slime mold thinks (and can design the Tokyo subway
system given a map of the city).  Perhaps places think.  Very likely
it all thinks.

It's very true that nobody knows how a machine thinks, (and thus we
can't be sure the simulation is the reality), but I'm starting to
wonder if the intellectually faulty (or at least treacherous)
problem of defining intelligence "many ways for many contexts" is
also somehow  a key to something important and not always just a
fixable kind of sloppiness.

Someone asked above "What keeps you up at night?"  Well, one thing
for me is playing out the implications of the Gaia hypothesis - or
more warmly stated, the reality of "Mother Nature".  Because when
you're talking about machines and singularity, certainly the thing
to fear is becoming subservient to them.  But in the case of
perceiving other kinds of intelligence, receptivity is of primary
importance, and in the case of Mother Nature, service is exactly
what you want, and the whole game is releasing blocks to being of
service, in service, all that.  

This brings me to what Matthew Battles commented in <78>, that:

>It feels like it's been a bit of a breakthrough year for
minoritarian futures, for indigenous perspectives on futurity in
particular. Are there NDN fictions, first-nations speculations,
afro- and other-futurisms that folks are following? 

And I would say, yes, that is exactly where my interest has been
drawn this year.  Ilarion Merculieff of the Global Center for
Indigenous Leadership is particularly articulate.  The fascinating
way that Standing Rock transformed the language of protest - we are
not "protesters" we are "protectors".  More on all that in later
posts.
  

    

I watched rather a lot of chess videos last year, and not because I
like chess.  I watch the games because I’m seeking out aesthetic
principles for generative art systems.

People sometimes call chess “The Drosophila of Artificial
Intelligence.” However, it’s not the long computer-science history
of AI and chess that interests me.  I don’t play chess, and I’ll
never be good at it.   Mostly, I want to understand why AlphaZero,
the neural-net super-chess player — currently undefeated by any
person or any engine — plays in such a pretty way.

To start, I have to make an aesthetic argument that AlphaZero indeed
plays “beautifully.”  I don’t want to just declare that, de gustibus
non disputandum — I’ll try to shore that up with some evidence.   

So, well, at least the thing’s famous: it’s a hit.  AlphaZero is
genuinely popular among chess fans.  The more conventional chess AIs
—  the “engines,”  in fan-speak— can certainly kick the ass of any
human player.  However, human chess fans never clamor to see their
games.  The engines bore people; their games lack grace.  

AlphaZero, though, has some major admirers.   Chess masters even
like it.  They praise its play in public.
  

    

Also, you might try comparing AlphaZero’s games to the recent human
world chess championship between Magnus Carlsen and Fabiano Caruana.


These two guys, Carlsen and Caruana, are excellent modern chess
players and very erudite professionals, but they play more like
conventional chess engines than AlphaZero does.  In their recent
head-to-head match, Carlsen and Caruana ground it out in endless
draws.  They never blundered: they were inhumanly consistent.

So, they had to decide their world championship in sudden-death
matches, where there’s very little time to devote any human
intelligence to the state of the board.  Then Carlsen then won
promptly because he’s got a well-known knack for playing rapidly.

None of those Carlsen-Caruana games are beautiful “immortal” games.
Those game are world-class, but they won’t be fondly remembered
decades from now.
  

    

I’m no chess pro, so a lot of the stamp-collector subtleties are
bound to pass me by.   However, I can tell if the events on the
board would make a live audience gasp and applaud.  Chess does have
an aesthetics; people get a frisson from it; they’re thrilled, even.


Chess has been pretty for a long time.  I’d even argue that there
are  aesthetically “prettier” games that date from the old-fashioned
1800s, at least in the kinetic-art sense of the way that pieces
elegantly swoop and move around.   

Years ago, chess people had very little theory. There weren’t a lot
of book openings or book endings to memorize, so grandmasters were
bashing it out like sword-fighters.   So in these much older,
historic  games there’s lots of crowd-pleasers, the too-bold
strikes, the head-exploding unnecessary complications, deliberate
tricks, traps, cunning swindles — psychological operations even,
where one grandmaster knows he’s got the other on the ropes, so he
moves in for the kill with dazzling stunt moves that he knows will
upset the opponent.

AlphaZero doesn’t do any of that.  It doesn’t even do what its
best-known public opponent, the engine “Stockfish,” does.   It never
plays like a human, but also,  it doesn’t play like  any standard
computer code “engine”  with its motor-like sets of specialized
component subroutines, and its value-weighting system.

It’s hard to describe what it does, in the arcane computer-science
realities of neural-net back-propagation, but, well, it’s still just
software, all right.  AlphaZero is the winningest chess engine ever,
but it’s not an AI god, it's not getting insanely better.  It’s not
an amazing, mystic Singularity that is accelerating off the charts;
its game-skill seems to be leveling off.  

Also, AlphaZero’s play, once you get used to watching it, is a bit
same-y.  AlphaZero wins all the time, but it is not brilliantly
inventive.  It doesn’t toss off an endless bravura series of
super-complicated, brain-scrambling gambits.  It’s just dazzlingly
unorthodox by the previous standards of humans or non-neural AI
engines.
  

    

Some people say that AlphaZero's play is “beautiful” in a
mathematical way, in the way that math is “beautiful” when it’s
simple and true.   But I doubt that AlphaZero has actually “solved”
chess, in the sense that its neural net method of play is the one
“true” method, and the only method that henceforth will ever win an
AI chess game.

I believe that, oddly, because I’ve got more faith in AI than that. 
I’m a big AI skeptic, but the defeated Stockfish has got heaps more
AI in it than AlphaZero does.   “AI” has not conquered chess,
because AlphaZero is just a modestly-sized neural net.  As an
“engine,” AlphaZero’s not all that big, not powerful in processors,
not ultra-fast  in code execution and it doesn’t have heaps of Big
Data about chess-game databases.   

I’m inclined to surmise that gangs of very specialized AIs with
databases of all of AlphaZero’s games would be able to gang up on it
and defeat it.  Also, new neural-net architectures are waiting.
  

    

I may be wrong about that last part; maybe the phase-space of chess
has been exhausted by AlphaZero, and it’ll always be the greatest at
chess, forever.  But would it be the greatest at some variant that
wasn’t chess, a super-complex game that was a million squares
across, instead of eight?  Probably not.  That means that we can put
aside a hushed, amazed respect for AlphaZero.  It’s okay to pick at
it critically.  Like you might analyze other beautiful, nonhuman
phenomena. 

    Snowflakes, maybe.  Human artists can’t draw a billion beautiful
snowflakes in ten minutes, but it’s pretty common to have them fall
out of the sky.  Sometimes a snowfall is very pretty, even
enchanted-looking, other times snow is damp, gloomy, ugly and
oppressive, and it’s all right of us to look out the window and
assess that — “pretty snow.”  That’s not presumptuous of us, it’s
more like a pleasant statement of our willingness to live in the
world we're in.
  

    


    AlphaZero is not an intelligent human, though it would do well
at a chess Turing test.   Chess masters usually know when they’re
blind-matched against engines.  AlphaZero, though, plays with such
dazzling efficiency that it would probably be world-famous if nobody
knew what it was.  It would slaughter all other human players with
its naive brilliance, but it might be taken for a child-prodigy.

     I appreciate the public commentary that chess analysts offer
about AlphaZero, but they always attribute intentionality to it,
which I dislike.  “AlphaZero wants to do this, it plans to do that,
it wants to avoid doing this…”. 

      That certainly happens in human chess, and even engine chess
has some kind of value-driven routine where it’s “trying” to
“achieve” some state of the board, in some sense of “trying.” 
Watching AlphaZero play is more like watching frost forming on the
the window.  “The frost has plenty of moisture over here but it
‘wants’ to invade the part of the window over here where it’s not as
damp,” that remark seems illegitimate to me.  It gets in the way of
properly appreciating the beauty of what’s going on.  It’s a
tooth-fairy explanation.
  

    


     I do have some ideas about why AlphaZero’s play looks so
pretty.  If it was an artist asking me for advice as a critic, this
is the helpful feedback I would offer to young Alpha.  “You’ve
clearly got a gift for that part — you should do more of that!”

1.  Since AlphaZero taught itself to play, its play is free of the
human installed base, so it’s liberated to do things that surprise
humans.  It has a kind of charming wunderkind naivety.  That element
of surprise is beautiful — it has a gosh-wow factor — but I don’t
think it’ll last.  Ten years from now, people will be talking quite
knowingly about neural-net play.  The sense of wonder has a short
shelf life.

2.  It doesn’t value the pieces.  It’s not emotionally or
algorithmically attached to them; to “throw away” a valuable rook
for a cheap knight is no problem.  This looks very daring and even
ethereal by human standards.  Even engines won’t do that because
they tend to hoard everything, rather like Apple mopping up profits.

3.  AlphaZero always plays the whole board — it never gets tangled
up in the hot corner where all the action seems to be.   So it’s
continually “retreating” to distant areas of the board where the
bishops, rooks and queen are just as powerful, but it’s hard for a
human eye and brain to stay focussed.  

     After you see AlphaZero do this twenty or thirty times, though,
you’re like: “Huh.  It’s doing that bishop-way-over-yonder thing
again.”  I think this is the new aspect of AlphaZero’s native play
that human masters will pick up pretty quickly.

4.  It’s very tidy.  It makes small, rather feline movements to
adjust its positional structure, rather than bulldozing in to crush
all resistance.  It commonly does these little grace-notes at the
exact time that a human player would be getting really excited and
bashing-in for the kill.  There’s something really pretty about
these small, neat moves — they’re like blue-notes in jazz.  Or,
rather, it’s just got its own swinging tempo; jazz defeating
chamber-music.

There’s something oddly gracious and feminine about this machine’s
apparent attention-to-detail.  These apparently small and irrelevant
housekeeping moves often turn out to be keys to victory.  So you
can’t dismiss that as “tidiness” - it’s like complaining that your
neat-freak Mom does the laundry too often, while next door they’re
all dying of typhus.

 AlphaZero’s refined, delicate chess play feels like an advance in
civilization, rather than just an advance in the game.  It’s like
it’s figured out table manners and washing your hands — it makes the
previous era of chess look like feudal life in a barracks.
  

    


So, well, I enjoyed that.  An engrossing game, chess.  Been around
quite a while, and in 2019, particularly interesting.  

 However — what does chess imply for the aesthetics of other,
seemingly very different forms of algorithmic expression?  Is there
some kind of Artificial Artistic Intelligence lurking here, which
might have broader applications for other things that get diligently
wound-up and seem to run on their own, such as code art, motion
graphics, device art, installations, techno music?

I’m still working on that issue, in a Drosophila lab-style, but I do
think about it differently than I did just a year ago.   And when
this year 2019 is over, well, maybe I’ll get somewhere.  We have to
hope, even if AlphaZero can’t and doesn’t.

As special brain-candy for hardcore chess fans, here’s Napoleon
Bonaparte getting his ass kicked by the “Mechanical Turk.”
Allegedly, anyway.  I admit it: I wasn’t there.

http://www.chessgames.com/perl/chessgame?gid=1250610
  

    
Fascinating.  I brought up Ilarion Merculieff, the Aleut elder
earlier.

He describes himself as being of the last generation that had a
traditional education, and says that it primarily consisted of
watching and wondering.  When he was five or six he says he
remembers watching a cliff full of seabirds for days and weeks and
months on end, watching the birds do things like fly up in a flock
and then wheel and dip and turn as one, wondering, "How do they do
that?"  

He says the traditional way of education was simply to support the
naive child in its curiosity.  His grandfather was the one primarily
in charge of his education, and when the young Ilarion would ask a
question, instead of answering, his grandfather would just point him
back to observing - the answer was there, in the preserved and
developing naivete.  We think of traditional knowledge as being
handed down, but maybe it is protecting the great plastic capacity,
keeping alive the possibility of seeing with new eyes that is the
secret. "Aiglatson," as it were.

Now I'm interested in AlphaZero too...
  

    
Bruce I’d be very interested in any thoughts you might have about
posts 84 and 85. 

My take on the transhumanist agenda is that it’s about perfecting
and enhancing human capability. There are plenty of pop culture
examples around to illustrate how deeply this meme is embedded in
our cultural mapping. Transformers, the Terminator movie, the
Nietzschean notion of “Übermensch”, often roughly translated as
“Superman.” So it seems fair to ask if the AI/ BMI quest driving to
the singularity (I see them as existentially equivalent) connects at
some level with these value-laden memes and the overall but likely
misguided notion of human perfectibility.

>86 and earlier
My concern would be that we delegate responsibilities to computers
that they ultimately don't handle well.

Yes of course and this brings up the issue of ethical and even
spiritual considerations. Designers and programmers will likely
never be able to create a true ethical sensibility and the more
spiritual aspects of human experience in AI and no degree of
developed intelligence can be considered worthy without this
checkpoint. 

The logical conclusion is that such creations will be ontologically
flawed from day one, even as they may make important contributions
to various other human pursuits and endeavors. Some will argue that
ethical “capability” can be encapsulated in an algorithm, despite
the fact that ethics debates have raged for centuries. Implicit is
the notion that programmers themselves will act as the final
arbiters of whatever ethical profiles are developed. But here’s the
Catch 22: the programmer can only create something that already
exists within the programmer’s mind.

>86 again. The Gaia hypothesis stipulated a “living” system not just
a complex adaptive system in my understanding
  

    
I haven't watched as many Chess games as Bruce has, but I've been
reading a lot of conversation around AlphaZero and AlphaGo/Master
etc... it was fascinating and disconcerting to watch this "alien
intelligence" (a Go master's word) land in a world of human
players... 

The part I keep returning to is the Advanced Chess model, which I
trot out a lot because I think it has a very strong and interesting
message. When Kasparov was defeated by Deep Blue, he was very very
pissed off, but he came back the next year with Advanced Chess (aka
Centaur Chess) which was the same game played by human/computer
teams (as opposed to human vs computer). And this immediately became
wildly interesting, with wholly new approaches to play being
generated very quickly.

One of the most startling findings of Advanced Chess is that while
even a modest chess computer can now thrash any human player, a
human and a modest computer working together can beat a much more
powerful computer playing alone. There's a transformative
combinatorial effect at work that magnifies the strengths of both
ways of thinking (and, to my mind, emphasises their differences in
interesting ways).

Google has talked a bit about what it calls the Optometrist
Algorithm, which it developed with Tri Alpha Energy, a fusion
startup. This is essentially Advanced Chess applied to scientific
experiment: runs of TA's experimental fusion reactor are assessed by
a human and a machine learning algo, with the algo presenting a
human overseer with a selection of possible tweaks to the
parameters, and the human choosing which set to follow. One result
is one that Bruce notes in the chess play: the whole board gets
played. Humans have a natural tendency to ignore the edges and
concentrate on a few variables (e.g. piece value in chess). The algo
doesn't do this, so everything is in play, but the human plays
hunches too.

Centaur approaches (which Matt Jones, now design director at Google
AI, has been talking about for some time) are a lot more hopeful
than singularity fears - hoping, of course, that we get to be the
front end of the horse.
  

    
I found keta's post (#86) quite fascinating--in its entirely, but
especially in raising the comparison of the Singularity to the Gaia
Hypothesis, and our differing emotional responses to them. But of
course Gaia isn't something new, potentially transcending our human
capacities, Gaia produced them. Our cultural distrust of 'the
machines' taking over is rooted perhaps in a right humility, the
sense that anything we created, even if it then begins to enact its
own evolution-equivalent, cannot be as wise as the sum of the
unfathomably numerous experiments run by the biosphere through time.


But then if we include what we have made as part of Gaia's
making--what then? Might I/we have a different, warmer, emotional
response? The way that one of my geneticist friends (unlike most of
my geneticist friends) is wildly enthusiastic about human beings
taking control of our own future evolution, full steam ahead? (Don't
worry, friend is now retired, and was studying what is, not creating
what is not yet.)

Not being a chess player, I cannot appreciate AlphaZero's beauty
except by the picture Bruce draws of its novel ways of playing. To
do the effective thing in ways not previously known is surely one of
our human definitions of beauty. This perhaps makes it qualitatively
different from other forms of computer-generated "art." Chess is a
game with a goal, knowable, recognizable; one party or the other
wins. Art, in the more classical sense of that word, leads toward
some not-before-seen destination, which nonetheless leaves the
viewer/reader with a feeling that something has been effected. 

In my own field--words--computer-generated works are thus far rather
dispiriting to engage with. There may be passing beauties, but you
could arrive at the same kind of inventions by any randomizing
process. The works as a whole, though, feel vapid.  

James brings in another view yet with the Centaur model. Expansion.
A useful and lovely metaphor--though I wonder, is it the Singularity
when the classical Greek Centaur image trades its halves? When the
humans become the embodying hooves and the AI the head? 
  

    
(slippage)

> There's a transformative combinatorial effect...

This reminds me of the scenario offered by William Calvin in his book,
"THE THROWING MADONNA: FROM NERVOUS CELLS TO HOMINID BRAIN" 

> Calvin, a distinguished Univ. of Washington neurophysicist, starts 
> off this collection of essays with a wild surmise. We are right-handed
> (most of us) because women discovered early on that babies calm down 
> when pressed where they can hear the heartbeat; that means cradling 
> the baby with your left hand. Some wise lady tool-users also 
> discovered that they could hold a child and lob a rock at a passing 
> rabbit, thereby increasing the larder at little energy expenditure. 
> Since one side of the brain is usually better at programming rapid 
> sequencing (like throwing a ball), mothers with left-brain sequencers 
> might then be more successful at hunting--and mothering--and thus 
> increase their genes in the pool. Hence, the Throwing Madonna. Though 
> Calvin milks this theme in the first few essays (in terms of bigger 
> brains, handwriting, etc.), he admits it's conjectural, and then 
> moves on. 

taken from:
<https://www.kirkusreviews.com/book-reviews/william-h-calvin-2/the-throwing-mad
onna-from-nervous-cells-to-homi/>
  

    
> wildly enthusiastic about human beings taking control of our own
> future evolution

I've been nursing the notion that, as generations have passed, we have
drifted away from the original understanding of 'Cybernetics' as
the science of "control and communication in the animal and machine."

James Lovelock initially defined Gaia as:

  "a complex entity involving the Earth's biosphere, atmosphere, oceans,
   and soil; the totality constituting a feedback or cybernetic system
   which seeks an optimal physical and chemical environment for life on
   this planet."

Perhaps because the 'natural' and 'constructed' worlds came to be seen
as opposites, it tends to be difficult to keep the ideas of 'animal' and
'machine' in mind long enough to discern the true cybernetic aspects
of each.