Dear Olia,
thank you so much for your pointed observations. I would like to extend
on both Sophia and Jarvis. Both seem to be driven by control and
automation phantasies on the one hand and anthropmorphization on the
other. "I’ve built a simple AI" already uses this "a AI", ascribing
agency (with the implication it would be human-like agency) to it,
obscuring the human labor behind it and the machinic limitations it
necessarily has.
The Team behind Sofia, for instance, consists of the CEO, the animation
and interaction designers, the software enginieers, the content
developers, robotic face architect, robotics engineers, a production
supervisor, a language processing A"I" engineer, IT-Manager, office
staff, events coordinator and freelance project workers according to
their website. A team of at least 23 humans to get an elaborated
deception, a version of a mechanical turk, running.
Interestingly Sofia is this embodied, female slave, while Zuckerbergs
male voiced Jarvis is disembodied and omnipresent in his house.
I also liked your point behind marketing A"I" as a _product_ as in the
Jarvis case because I didn't see it from this angle. I rewatched this
demo video which (https://www.youtube.com/watch?v=vvimBPJ3XGQ) pretty
much addresses almost all phantasies a master could expect of his slave
AI, waking him up gently, providing a schedule of the day, preparing a
toast, surveilling and teaching the child, surveilling and providing
access at the door (for Zuckerbergs parents). What is excluded in this
narration is of course the master's violence against the slave, the
brutal availability of sexual services, but yes we are in advertisement,
I get it.
The video has a humorous moment built in, where Mark asks Jarvis to play
"the best Nickelback songs" and Jarvis answers passive agressively "I'm
sorry Mark, I'm afraid I can't do that, there are no good Nickelback
songs". Intented as a funny moment, it anticipates the "sorry, I'm
afraid" we read in every redacted or censored GPT answer today.
But it also has this moment of disobedience, which Mark diffuses by
answering (while his wife sits silently on the sofa besides him,
swinging gently the baby, failing even the simplest Bechdel test):
"Okay, how about just play some songs our whole family likes". To which
a children song follows. It is intriguing, how Mark dominates Jarvis, as
well as Priscilla and their first daughter Maxima just by his masters voice.
But this short moment of non-function or disobedience also ties in with
some of our (humans) largest fears when it comes to automation: being
dominated by robots as the other side of dominating them. This fear of
being dominated that may turn into the pleasure of being dominated as in
a proper SM relationship. In this sense the Jarvis product demo contains
all tropes of countless science fiction books and movies.
Having had Covid twice in winter, I re-watched lot of the star treck
series, a guilty pleasure. I apologize. And it stuck with me, to what
large extend this series (one example of many) deals with the
human-machine domination subdomination theme: Data, Seven of Nine, the
Emergency Medical Hologram, the caretakrer, Vic Fontaine are constantly
involved in these tropes.
I agree with you that A"I" has become an object to be marketed. We are
dealing with a double dimension here. In the imaginary space A"I" is
also a phantasmatic object of domination and submission, tied to the
dimensions of sadism and masochism and the master-slave logic.
I would love to see tiktok edits crossing the Jarvis product demo with
Star Treck tropes, addressing this double dimension.
Warm regards, Francis
Happy new year, dear nettimers!
In the mid of AI summer, I often think back about last spring, and one
product, or rather it's announcement, that brought us to where we are
now.
/https://pad.profolia.org/s/I_ve_built_a_simple_AI -- illustrated
version with links /
AI’s 75 year history is a chain of periods filled with excitement and
anticipation of the singularity, taking turns with times of ignorance
known as AI winters. I on purpose count from right after WWII – 1946,
when Mechanical (later Machine) Translation effort has started, and
not from 1956 when the term AI was coined, because the term is less
important when the idea that machines can perform understanding (of
languages), and also because it was exactly the failure of Machine
Translation that designated beginning of the first known AI winter in
the beginning of 1960’s.
On the eve of 2024 we, no doubt, are enjoying AI midsummer. It’s so
brutally, unnaturally hot, that some start to suspect that this time
it won’t be just a winter that will compensate for the heat, but
rather something usually depicted as an ice age or a nuclear winter is
in front of us. It will be very cold and dark for a very long time…
The last winter, which was caused by the failure of the 5th generation
computer and logic based AI systems in the late 80’s was rather short.
It was over in the mid 90’s. From that time research in machine
learning grew steadily. But science and industry seasons do not
necessarily correlate to the one’s in mass media. Public attention at
that time was glued to the Internet, WWW, Social Networks and all of
the above becoming magically available on mobile devices in the end of
the noughties.
In the beginning of the tens mass media started to report about Big
Data and the Internet of Things. Let’s, from today’s perspective, call
it a thaw. Real spring, or mass media AI spring started in 2016. I see
3 events that made a break in popular consciousness:
– Hanson’s humanoid robot Sophia,
– Google’s neural network algorithm Deep Dream and
– Mark Zuckerberg’s Jarvis.
One still remembers Sophia’s glorious appearances at the United
Nations and on the covers of lifestyle magazines. Sophia’s existence
also started a wave of discussion about ethics in AI and social
robotics, that are still actual and vital. Psychedelic images of Deep
Dream are also hard to forget. And the impact of the algorithm is
still here since style transfer and surreal visuals became AI’s
flagship businesses. But Jarvis? Why do I even mention this rather
modest personal project of Zuckerberg to control his house facilities
through his phone? The app itself was not remarkable, but the words
that announced it were.There were two posts. On January 3, 2016,
Zuckerberg published his New Year resolution: “My personal challenge
for 2016 is to build a simple AI to run my home and help me with my
work.”
And on December 19, 2016 he reported “I’ve built a simple AI”. These
posts quoted many times during 2016 and 2017 started a trend talking
about AI as if it is an app, a product, something at hand. “I’ve built
a simple AI” played a big role in bringing AI to where it is now – a
whirlwind of crap, when not only every further automation in software
products, but simply every new function or feature is marketed as
Artificial Intelligence.
That AI became a synonym to a chatbot is somewhat historically
justifiable. But there is hardly any app today that is not marketed as
AI. A filter is an AI, a file is referred to as an AI output. Ai (only
A is upper case, as if a given name) in the slang of Humane and name
of their Ai Pin is the sign of the same process.
AI is an object, let’s call it Ai. In 2016 “I’ve built a simple AI”
suggested that AI was a thing. Could be simple, could be complex, but
a thing, ready to be packed. In my opinion, it was the event that
marked the beginning of objectification of the Artificial Intelligence
concept, its oversimplification, and devaluation.
--
Francis Hunger, Ph.D.
Artistic Practice http://www.irmielin.org
Mastodon https://dair-community.social/@databasecultures
--
# distributed via <nettime>: no commercial use without permission
# <nettime> is a moderated mailing list for net criticism,
# collaborative text filtering and cultural politics of the nets
# more info: https://www.nettime.org
# contact: [email protected]
