eMusings
Your eyes and ears on the worlds of art, culture, technology, philosophy - whatever stimulates the mind and excites the imagination. We remind you that 20 years of
back issues of eMusings can be found on our archives page.
The media is rife with mis- and dis- information about AI. Let's look at some of the more cogent discussions:
We begin with a fascinating and important article. Stanford University presents an important article called "The Turing Trap:
The Promise and Peril of Human-like Artificial Intelligence". The lure of designing machines that imitate human responses has obsessed researchers for decades. Humanoid robots have been imagined for
centuries but this is the first time that industry has surpassed fantasy. What are the negative impacts of this thinking?
Stanford suggests an alternative approach called
Augmentation rather than Imitation. The new approach implies that humans and machines can work together, preventing a concentration of power and unimagined wealth in the hands of a few with the
concurrent disempowerment of those without power. (Thanks to BZ for this).
Scientists at the University of Texas Southwestern Medical Center
have come up with an algorithm called "Deep Distilling" that apparently is able to explain its reasoning for the first time. Standard AI relies on already known data, but has difficulty dealing
with unexpected events, causing a breakdown. The problem is magnified in medicine and pharmacology as well as traffic controls and automatic driving, where mistakes can be fatal. Compared to the
reasoning of a child, the new process relies less on accumulating massive amounts of information and instead uses building blocks designed step by step. This way the learning process itself can
be studied and better understood.
Mark
Zuckerberg claims that we will soon be able to use brain signals to control our AI glasses. The neural interface he is describing is embedded in a wristband which would allow the user to simply think about how
they want to move their hand and it would happen in the virtual world without having to physically move their arm. I won't begin to dive into the negative propensities of this one.
By now you have been reading about Sora, the Open AI algorithm that turns text into a short video. Although it is not currently available to the
general public, other companies like Google, Meta and Runway have broadcast their own efforts at the same technology. According to the publicity, Sora will be able to understand nuances as well as actions, and will create the
videos at once instead of frame by frame. Under study still are issues of safety, hateful content, bias, and clear labeling of the video as AI. According to
CBS, Sora and its cousins are especially dangerous prior to the upcoming elections. Companies like banks are particularly under the gun to offer tools that can protect their customers from malicious messages.
Another study from Stanford University reveals a new AI algorithm that can tell whether a
brain scan comes from a woman or a man with 90% accuracy. The implication is that gender-related differences can be discerned within the brain, a finding which was formerly not accepted. At issue are the treatments of learning
disorders, differences in social abilities, and the treatment of neurological and psychiatric disorders.
Efforts to place ChatGPT brains inside of robot bodies are raising major ethical questions.
Ishika Singh, a Ph.D. computer scientist at the University of Southern California, questions whether a robot can make dinner - ie, enter a kitchen, go through drawers and refrigerators, take and cook ingredients, and set the
table. The problems arise when the robot meets a problem that it did not anticipate. In other words, scientists are trying to teach robots about the world, but that world is continuously changing. Is it possible for robots to supercede
their programmed knowledge? Some robots have been known to "hallucinate" or "make things up". Are built-in safeguards like "output toxic language" enough? Or should LLM's (large language models) not be installed into robots at all?
Questions have surfaced about ChatGPT's
memory. Can and should it remember your interests and preferences in future chats? OpenAI is experimenting with an algorithm that will remember what you say across all chats. The company is suggesting a "temporary chat mode" to erase what you
have previously said, referred to by some as "targeted brain surgery". Privacy concerns arise here. Could your interests turn up in someone else's chat memory? What about the sometimes bizarre comments that ChatGPT comes up with? Will they
also be retained in your future chats?
Researchers at Dalhousie University in Canada are using AI to decode the language of chickens, which they describe
as "having a universal translator for chicken speech". It seems that chickens make sounds that vary in context, pitch and tone. Nonverbal communications of the animals are also being studied, like eye blinks and facial temperatures. The
research has not yet been peer-reviewed, but it is hoped to influence better farming, animal health, and ethical treatment of farm animals.
The proliferation of designer proteins has come under scrutiny as the availability of these new
biocomponents has rapidly spread. They are being applied, for example, to greener biofuels, drugs that last longer, protein-based vaccines, and plastic-destroying proteins. Once again ethics and responsibility have sounded an alarm. Some
scientists feel that the new proteins should have barcodes embedded into their code so that they cannot trigger a pandemic, for example. The rapid speed at which these designer proteins are created and spread are of particular concern,
since proteins are the building blocks of life. Several approaches in their design are currently being used. One, called structure-based AI, trains algorithms to ignore noisy data. Another teaches AI to find connections between protein "words"
and put these connections into what is called "biological data". Several measures aimed at biosecurity are being considered by nations around the world, hoping to provide safeguards without inhibiting research.
Now on to other March treats:
Judithe Hernandez has the first major retrospective of her artwork after a 53-year career. Her exhibition
is titled "Beyond Myself, Somewhere, I Wait for My Arrival", delving into the experiences of Mexican migrant women. The exhibition is on view at the Cheech Marin Center for Chicano Art & Culture of the Riverside Art Museum. You can
see more of her work in the Juarez series from the National Museum of Mexican Art.
"Salvage 2.0" is the title of the works of Jay Sae Oh's art, where she mixes found objects in jute packages and cords. The trash is hidden
from view yet still identifiable, both erased and memorialized. In essence she seems to be saying "Rubbish can be gorgeous". In-depth views of her work are available at her own site.
Kandy G. Lopez works with yarn and fabric the way other artists work with canvas and oils. She says she learns from her materials and learns from the subjects of
her woven portraits. Textures and details work together to produce stunning portraits, with fibers often found in thrift shops and donated bags of yarn. She shows people of color with confidence and dignity, with shoes, shirts and pants
woven in exquisite detail.
Christian Rex van Minnen casts his sardonic eye somewhere between the beautiful and the sardonic. Titles like "Peaceable Kingdom Slowed and Throwed" reflect
an inventive and curious mind. Van Minnen is fond of using drawing, solid objects, distortions and strange juxtapositions to achieve macabre effects, startling the mind and awakening the eye.
If you are a fan of small motorized vehicles, you might want to take a look at these electric mobility scooters. Named Tectus, these
3-wheeled single passenger scooters are made in Canada. They feature 2 electric motors, one for the front and one for the rear, which together offer 2,000 watts going at a maximum speed of 20 mph (32 km/h). The lithium batttery pack
is said to run for up to 99 miles (160 km). Instead of a door, the Tectus features a motorized canopy that can be opened remotely, and a lidded box in the rear for packages. Two models are available, one called Deluxe for
$7,000. USD, and the other, Ultimate, for $9,000. USD. The latter gives you a backup camera, running lights, a stereo sound system, GPS tracking, and heating and air conditioning.
Treat yourself to the watercolor illustrations of Jean Mallard. Imaginative fantasies about dark and dusk appear like glimpses into a child's mind with a
sense of enchantment and wonder.
Cephalopods in stunning tonal variations were drawn by a French pharmacist-turned-naturalist named Jean Baptiste Verany in
1851. Verany showed the lively creatures with amazing accuracy for the time - a contemporary review mentioned the "suppleness of the flesh, the grace of the contours, the flexibility of the membranes, the transparency, and the coloring".
Verany's work strongly influenced the glass artists Leopold and Rudolph Blaschka.
Kim Keever brings his background as an engineer and work for NASA to his fascinating compositions. Keever specializes in underwater images, meticulously constructed. He pours paint into
a 200 gallon tank of water in his studio, then uses a large-format digital camera to photograph the results. Sometimes he will also drop plaster objects into the tank as well and then drop paint pigments into the mix. The results are
quite amazing and well worth your time to view.
Finally, we invite you to enjoy the exquisite watercolors of Lisa Hunt, along with a couple of videos of the artist at work.
c. Corinne Whitaker 2024
want to know more about the art?
front page , new
paintings, new
blobs, new
sculpture, painting
archives, blob
archives, sculpture
archives,
photography archives, Archiblob archives, image
of the month, blob
of the month,
art headlines,
technology news, electronic
quill, electronic
quill archives, art smart
quiz, world art
news,
eMusings,
eMusings archive, readers
feast, whitaker
on the web,
nations one,
meet the giraffe,
studio map, just
desserts, Site
of the Month, young
at art,
about the artist?
copyright 2024 Corinne Whitaker