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 now part of the fabric of our lives. Here are a few of the more cogent observations:
LGTV will begin using more personalized
ads with AI technology that analyzes the viewer's emotions and beliefs. Under scrutiny will be each viewer's
personality traits, lifestyle choices, and personal interests. In partnership with Zenapse, LGTV will divide web users
into highly specific segments of the market. Zenapse claims it can identify the emotions of people watching tv based
on their consumption patterns.
Axios reports that over 50% of the top AI companies
based in the U.S. have at least one immigtant founder or co-founder.
The list is led by India (9 founders), China (8 founders) and France (3 founders). AI companies with 2 immigtant founders
include Australia, the U.K., Canada, Chile, and Romania. Specifically, Open AI has 2 immigrant co-founders - Elon Musk who
was born in South Africa and Ilya Sutskever born in Russia.
C. elegans is one of the most analyzed animals in science. Extensive efforts, so far in vain, have been made to make a computer
simulation of it. Since 2011, an open source software called Open Worm has been trying to make
a digital twin, called the "holy grail" of systems biology. Currently it takes roughly 10 hours of computer time to generate 5
seconds of results,
meaning understanding how a worm squiggles forward in a flat terrain of "goo". When asked why so much effort is going into this project, the
researchers sometimes quote Richard Feynman, "What I cannot create, I do not understand". The nematode is not quite as long as the
width of a human hair. It is not a parasite: it can eat, it can reproduce. One scientist stated, "It’s born and it develops, and it ages
and it dies—all in a millimeter." To date, 4 Nobel Prizes have been given to work on it. Interestingly, one of the most difficult
challenges involves reproducing how it moves backwards, or up and down, or why.
In a similar quest, engineers at Georgia Tech have developed a
5 inch soft robot that can jump 10 feet high although it doesn't have legs. Imitating the movements of a small parasitic
worm, the device is made of a silicone rod with a carbon-fiber spine. A researcher at University of California, Berkeley, used
high-speed cameras to watch nematodes twist their bodies into different shapes depending on where they wanted to go. He discovered
that the tiny worms change their center of mass to control which way they jump. Ishant Tiwari explains, "Kinks are typically dealbreakers.
Kinked blood vessels can lead to strokes...But a kinked nematode stores energy that is used to propel itself in the air".
At the National University of
Singapore, researchers have developed a method to make personalized gum tissue grafts by combining AI with 3D bioprinting.
The new process does not require the traditional grafting of tissue from the patient's mouth, a complex and painful procedure,
although it still has complicating issues like nozzle size, bio-ink viscosity, print speed, and extrusion pressure.
Researchers at
Anthropic have found a method to figure out how AI algorithms process data and make decisions. The AI models turn out to
be more complex than expected. When composing poetry, for example, they appear to plan ahead. They function the same regardless
of language. They sometimes work backwards rather than building forward from facts. The studies illuminate issues that have stumped
engineers up until now: the scientists really don't know how the algorithms come up with specific responses. The breakthrough came
from treating models like biological systems. In the poetry process, the algorithms worked first with rhyming words for the
end of each line before writing. In terms of language, the AI models do not make separate frameworks for each language. Instead
they deal with abstract concepts before composing. Sometimes, however, the models' claims of what they do simply do not
match what they actually are doing.
AI Emergence, founded by former IBM researchers, is offering an AI agent creation system. Humans use text prompts to
ask for the tasks they want to do. AI algorithms then create agents that can accomplish the job. The process is called a
no code natural language AI-powered multi-agent builder that works in real time. The founders claim that it works without
human obstacles but under human control. The method first checks for pre-existing appropriate agents before creating a new one.
The new one
learns to self-play and then can create additional agents, named Multiple Enterprise AI agents, to undertake related tasks. Involved
are skills like long-term memory, planning and verification, and searching surrounding task spaces for potential amplification.
In the company's words, "we're marrying LLMs’ code generation capabilities with autonomous agent technology". Other enterprises
can also bring in their own models and 3rd party agents to work within the new framework.
AI boosted cameras are being used to help the blind navigate
around obstacles. The cameras are mounted on a pair of glasses which record live images of the surrounding area and then
use audio alerts and vibration to feed information to the wearer. The basic live-feed data is augmented by machine-learning
algorithms that detect things like doors, walls, furniture, and other people. Being tested as well are flexible artificial skin patches
worn on the wrists and fingers to give further information on how close an object is that needs to be grasped.
Test results were roughly 25% better than using a cane.
Open AI's new o3 and o4 mini modules
tend to hallucinate more than earlier models. These so-called reasoning models are defying the researchers' attempts to find out why.
An outside non-profit AI research lab has confirmed the troubling findings.
At the same time that AI reasoning is being upgraded, it is also showing an increase in hallucinatory responses. Sam
Altman, the CEO, is now claiming that we are in an era of Renaissance rather than Revolution. The PR blast comes on the
heels of Open AI first occupying a nonprofit niche in the company's corporate structure and then converting it to for-profit status.
Their ChatGPT recently introduced native image-generation, which Open AI claims had more than 400 million weekly active users
since February.
Salvador Dali experimented with movies, but his most complex film was never
made, until now. Titled "Giraffes on Horseback Salad", it was meant to star Harpo Marx and include flaming giraffes. Google;s
DeeepMind Veo 2 generative video model is now tackling the project, attracted by Dali's lack of coherence.
Anthropic has just announced new problems
with its Claude generative AI. Apparently, even with safety checking, Claude is being used to create malwware not only by sophisticated coders but by
people with no coding experience. Politically motivated commands were given, involving several countries and languages, with the goal of
exerting long-term influence rather than just flooding social media.
On to other May treats:
A striking sculpture at Rockefeller
Center in Manhattan is attracting considerable attention. The piece was created by Jaume
Plensa, an artist based in Barcelona, Spain. If you look closely at his site, you will see that he is no stranger to
eye-catching public installations,
"Elaborate Textile Paintings" is the phrase
being used
to describe the works of Anne von Freyburg. The artist appears to have thrown large dabs of paint onto fabrics and textiles. She claims to have been
inspired by 17th century Dutch Masters' flower still lifes. You can see more of her pieces at the Saatchi
Gallery.
Saatchi is also showing colorful geometric abstractions by Dominic
Beattie, based in London and Spain.
For those of you with curious, inventive minds, the Online Encyclopedia of Integer Sequences (OEIS)
may pique your interest. You can check your favorite sequences, or look into Puzzle sequences, Hot sequences, or Classic
sequences. For instance, my birthday came back "well-defined solution for recurrence".
Shiori Tono has developed an unusual
technique for creating her paintings. She first makes a structured grid with strips of masking tape, leaving some squares
exposed. This process reflects her understanding of personal memories, with some exposed and others hidden. At the end,
she removes the masking tape and often finds unexpected results, just as memories tend to be imperfect.
Large-scale imposing installations are the trademark of Avram Finkelstein,
whose latest project is titled "Something Terrible Has Happened". The artist and writer lives in Brooklyn, New York, and has been an
active participant in political art. You can see more of his imposing
installations here.
The large crushed metal sculptures of Kennedy
Yanko span several galleries in New York City. Yanko collects old metal scraps from junk yards and then creates skins
of paint for them, a method that she calls "paintings without a canvas". The combination of badly scarred metal fragments
with paint seems to combine despair and hope. Her works have been described as holding a "haunting gut-punch physicality".
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