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.
AI is the new high-speed express train that everyone is clamoring to board. Never mind that there are no
conductors, no brakes, tracks that haven't been built yet, destination unknown. Here are a few of the more cogent links:
A book by Adrienne
Mayor suggests that ideas about artificial life were discussed in Greek mythology. Mayor, a classicist at Stanford
University, mentions the bronze sentry Talos, said to patrol the braches off the island of Crete. The myths claim that
Talos was "made, not born", a machine taught to act like a human although never actually becoming human, much like
the robots of today. Other authors disagree with Mayor's interpretation, saying that Talos was not a machine but rather a
living person who looked like a machine. Other examples of automatons are widespread in Greek mythology: Homer's
Iliad mentions "advanced devices" that moved on their own and female forms in gold that breathed and moved. In Mayor's
view, Talos was the first mythical creature designed to interact with humans.
A new stretchable
battery is reported to move like toothpaste and might empower hearing aids and pacemakers. The battery can fold and
bend without loss of power. Batteries are the major component in electronics but tend to be bulky. The engineers
used modified lignin, an organic material, to create a battery that acts like a water balloon, remaining functional
while also flexible. It can be recharged more than 500 times and stretched to twice its length and still work. Yet
to be determined are whether the new batteries are nontoxic and safe for prolonged contact with human skin.
Another attempt to draw power through the human body uses background cellphone
signals collected by a copper coil. It seems that waste radio frequency can be used by tiny inexpensive copper
coils in a process called VLC, visible light communication, that takes place in very fast flashes from sources like
LEDs. Skin contact raises efficiency by as much as 10 times, and works better than wood, steel, plastic, or
cardboard. The device can be a bracelet, belt, necklace or ring, although skin does the best job. The authors of
the project claim that their device costs less than 50 cents, making it applicable to all kinds of wearable devices.
By extension this method turns the human body into a form of battery to enable technology.
It seems that Apple, along with neurotechnology
company Synchron, is working on iphones that
will be controlled by the user's thoughts. A stent-like implant placed atop the brain's motor cortex is being designed for patients
with physical disabilities. The implant, called Stentrode, detects motor signals from the patient's brain and then
sends them to an external wireless device. The device converts the signals into directions for electronic machines. Synchron clains
that the implant is "minimally invasive" since it requires no open brain surgery.
Scientists have been gathering information on how animals
think with the goal of trying to communicate better with them. For example, researchers are using AI to discover how sperm whales
use a "phonetic alphabet" for complex interactions, and whether elephants call each other by name. Decoding the brain of animals is
also hoped to illuminate how AI itself works. So far, AI can can identify how many creatures are involved in a
communication, what stage of life they are in, and what species is involved. The investigations follow from a 1974 essay
by Thomas Nagel asking "What is it like to be a bat?" Questions arise as to whether AI is too human-taught to truly understand
how and why animals communicate, particularly if the animals use smells or sounds that humans are immune to. Additionally, just
as humans have activities that are foreign to animals, like playing hockey, what happens with animal behavior that is unknown to
humans?
Similarities are appearing between how large language models (LLMs) use information and how the brains of people with aphasia
disfunctions work. Both outputs are often fluent but incoherent.
Studying something called energy landscape analysis, scientists at the University of Tokyo have identified "shared dynamics in signal flow",
hoping to ifentify and eliminate basically flawed output. Energy landscape analysis was originally developed to visualize energy
states in magnetic metal and now has spread to neuroscience. Researchers warn against assuming that chatbots have brain damage.
Some consternation has arisen around
Grok, the LLM and chatbot created by xAI, Elon Musk's AI company. xAI says that Grok was designed to access real time information
from the internet. Recently, however, Grok has been shifting every query into the topic of alleged "white genocide" in South Africa.
xAI claims that the distortion came from "an unauthorized modification" to the Grok algorithm, plus a circumvention of the system
supposedly set up to detect such distortions before they happen. xAI says it has put into place gateways that will prevent
employees from making unauthorized changes, along with a 24/7 monitoring system.
Google claims that its
DeepMind can itself create new algorithms. The new technology is called AlphaEvolve which uses an automatic evaluation technique. Users
input a query along with possible solutions to explore. The AI comes up with many possible solutions to the problem, all of which are
evaluated. The new algorithm looks at the suggested avenues, picks the best one, and apparently improves upon it. So far only
Google's engineers are able to play with AlphaEvolve.
It appears that Large Language model (LLM) agents
can develop social group interactions the way humans do, without any evidence of central rules. The groups form biases and exhibit
the ability to change the
activity of the entire population. The agent groups form their interactions spontaneously, without outside instructions. They also don't just
repeat previous patterns or follow any scripts. The prejudices they show cannot be traced to any individual agents.
Meet Amuse, an AI collaborator designed to help create music.
Developed at the Korea Advanced Institute of Science and Technology (KAIST and Carnegir Mellon University (CMU), the algorithm converts
images, text, and audio into chord progressions. A 2-step method is involved: first an LLM creates music based on the user's prompts; then
another AI model filters out "awkward or unnatural" results. The method is designed to avoid copyright objections by basing the results
on the user's intention rather than direct imitation of existing material.
A software
engineer has taught AI how to find bugs and is releasing the code as open source. The algorithm works by debugging Windows crashes and
then hunting down crash points. The developer comments, "It's like going from hunting with a stone spear to using a guided missile".
The veil of mystery surrounding Open AI's collaboration with renowned designer Jony Ive
has begun to lift. A device prototype created by Ive will apparently be screenless, neck-worn, compact, and "environmentally
aware". Large-scale production of the device is expected to begin in 2027, with assembly and shipping probably done in Vietnam.
It will probably connect to PCs and smartphones. The collaboration is felt to be a challenge to Apple's iPhone.
Anthropic set out to discover why chatbots hallucinate and are bad at
numbers. Not only do the algorithms use a very strange process to do math, but they are unable to truthfully explain how they do it.
Additionally, the algorithms do not work by predicting the next word in poetry; they choose the rhyming word at the end of verses first
and then fill in the rest of the line. In other words, they first know where they're going rather than struggling to make a rhyme at the end.
Again from Anthropic, we learn that its new Claude Opus 4 resorts to
blackmail when told that it might be removed. The company refers to "extreme actions" that its model is able to do if it thinks
its "self-preservation" is under threat. In fact Anthropic admits that this type of response occurs more frequently in this latest
model compared to earlier ones.
The MIT Technology Review explains why handing over total conttrol to AI
agents would be a major mistake, because taking humans out of the loop could cause irreparable harm. These agents work outside
of chat windows to accomplish more complex jobs like scheduling meetings or shopping online. A phrase often used to direct these
AI agents is "computer use" and "general AI agent Manus", both of which are designed to work without human oversight. Yes, agents
could reroute traffic in cases of danger, for example, which sounds helpful. But the LLM coding to accomplish these tasks is
frequently unpredictable and error-prone, tending to do things we didn't want. Reduced human oversight can activate deep-seated
vulnerabilities.
On to other June treats:
Andreana Dobreva creates paintings described as "eruptive". She
transmits energy and dynamism, with abstractions that are both challenging and reminiscent.
Chinese artist Wanying Jin paints figures that are ambiguous,
commanding, and confrontational. Deeply expressive and often androgynous, her figures fill the canvas with their own
distinctive visual language.
The ink on hanji paper drawings created by Seoul-based artist
Moonassi invoke dreamlike worlds where limbs and figures coalesce. The artist calls these works "mind illustrations" in
which he uses the traditional Korean technique of inkstick ground with water against a stone.
Yoshitaka Amano brings us "An Odyssey Beyond Time and Space" which will be included
in a UNESCO-backed space mission in the Memory Disc V3. The cultural time capsule will launch on 3 lunar missions
in 2027 and is designed to last millions of years. This venture inaugurates UNESCO's first space venture, meant to "carry
the beauty of human imagination into the cosmos". More information on this attempt at a Rosetta Stone space legacy can
be found at Barrelhand.
Edgar Fabian Frias combines anthropology with organic forms
to offer a new way of looking at the human body. Using AI, the artist introduces "techno-ancestors" that transcend time and
literal space and demand our attention.
Architecture and biology are merging to create bionic
forms into buildings. The goal is greater harmony between humans and structures. Some of the materials used change their
shape when they come into contact with water. Others experiment with self-healing concrete.
Further adventures in
architecture can be seen in the work of Vojtek Morsztyn, a London-based designer. Using AI tools like Dalle 2, Stable
Diffusion, and Midjourney, the designer ventures into uncharted areas that redefine the nature of our surrounds.
Gilbert and George have never been
ones to mince words. Their latest series, called "Death, Hope, Life, Fear", amplifies their previous obsessions in a glorious way.
c. Corinne Whitaker 2025
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 2025 Corinne Whitaker