…a new neural network based on GPT-4 finds errors in its work and fixes them.
CriticGPT, a model based on GPT-4, writes critiques of ChatGPT responses to help human trainers spot mistakes during RLHF
…a new neural network based on GPT-4 finds errors in its work and fixes them.
CriticGPT, a model based on GPT-4, writes critiques of ChatGPT responses to help human trainers spot mistakes during RLHF
Could we identify an alien terraformed planet through the detection of greenhouse gases? This is what a recent study published in The Astrophysical Journal hopes to address as a team of international researchers investigated whether artificial greenhouse gases could be detected from an exoplanet whose alien inhabitants could be attempting to terraform that world, either from trying to control its climate or terraforming an uninhabitable planet into a habitable one. This study holds the potential to help scientists better understand the criteria and methods for identifying an extraterrestrial civilization, especially with the number of confirmed exoplanets increasing almost weekly.
“For us, these gases are bad because we don’t want to increase warming” said Dr. Edward Schwieterman, who is an Assistant Professor of Astrobiology at the University of California Riverside and lead author of the study. “But they’d be good for a civilization that perhaps wanted to forestall an impending ice age or terraform an otherwise-uninhabitable planet in their system, as humans have proposed for Mars.”
A group of physicists specialized in solid-state physics from the University of Cologne and international collaborators have examined crystals made from the material BaCO2V2O8 in the Cologne laboratory.
They discovered that the magnetic elementary excitations in the crystal are held together not only by attraction, but also by repulsive interactions. However, this results in a lower stability, making the observation of such repulsively bound states all the more surprising.
The results of the study, “Experimental observation of repulsively bound magnons,” are published in Nature.
Philosopher Wilfrid Sellars had a term for the world as it appears, the “manifest image.” This is the world as we perceive it. In it, an apple is an apple, something red or green with a certain shape, a range of sizes, a thing that we can eat, or throw.
The manifest image can be contrasted with the scientific image of the world. Where the manifest image has colors, the scientific one has electromagnetic radiation of certain wavelengths. Where the manifest image has solid objects, like apples, the scientific image has mostly empty space, with clusters of elementary particles, held together in configurations due to a small number of fundamental interactions.
The scientific image is often radically different from the manifest image, although how different it is depends on what level of organization is being examined. For many purposes, including scientific ones, the manifest image, which is itself a predictive theory of the world at a certain level or organization, works just fine. For example, an ethologist, someone who studies animal behavior, can generally do so without having to concern themselves about quantum fields and their interactions.
Posted in space
In one of the final chapters of his book: Reality+: Virtual Worlds and the Problems of Philosophy, David Chalmers asks, have we fallen from the Garden of Eden? “Eden” in this case is a metaphor for living in a world where everything is as it seems, matching our pre-theoretical view of reality.
In Eden, everything exists in a three dimensional Euclidean space. And time flows from one moment to the next with an absolute now across all of space. In Eden, color is an intrinsic property of objects, so the apple really is red. And objects like rocks are truly solid. In Eden, we have free will in the classic contra-causal sense of that term.
Once we lived in Eden. But then there was a fall. We ate of the Tree of Science and were cast out.
Go to https://ground.news/sabine to get 40% Off the Vantage plan and see through sensationalized reporting. Stay fully informed on events around the world with Ground News.
Geoffrey Hinton recently ignited a heated debate with an interview in which he says he is very worried that we will soon lose control over superintelligent AI. Meta’s AI chief Yann LeCun disagrees. I think they’re both wrong. Let’s have a look.
🤓 Check out my new quiz app ➜ http://quizwithit.com/
💌 Support me on Donorbox ➜ https://donorbox.org/swtg.
📝 Transcripts and written news on Substack ➜ https://sciencewtg.substack.com/
👉 Transcript with links to references on Patreon ➜ / sabine.
📩 Free weekly science newsletter ➜ https://sabinehossenfelder.com/newsle…
👂 Audio only podcast ➜ https://open.spotify.com/show/0MkNfXl…
🔗 Join this channel to get access to perks ➜
/ @sabinehossenfelder.
🖼️ On instagram ➜ / sciencewtg.
#science #sciencenews #artificialintelligence #ai #technews #tech #technology
Year 2016 face_with_colon_three
Clothing tears could be a thing of the past if a new material capable of “healing” itself after being ripped proves to be commercially viable.
Researchers at Pennsylvania State University created a fabric-coating technology derived from squid ring teeth that allows conventional textiles to self-repair.
“Fashion designers use natural fibers made of proteins like wool or silk that are expensive and they are not self-healing,” said Melik Demirel, a professor of engineering science and mechanics at Penn State. “We were looking for a way to make fabrics self-healing using conventional textiles. So we came up with this coating technology.”
Note that this does not involve Planck mass fermionic black holes!
A population of massive black holes whose origin is one of the biggest mysteries in modern astronomy has been detected by the LIGO and Virgo gravitational wave detectors.
According to one hypothesis, these objects may have formed in the very early Universe and may compose dark matter, a mysterious substance filling the Universe. A team of scientists has announced the results of nearly 20-year-long observations indicating that such massive black holes may comprise at most a few percent of dark matter. Therefore, another explanation is needed for gravitational wave sources.
The results of the study were published in two articles, in Nature and the Astrophysical Journal Supplement Series. The research was conducted by scientists from the OGLE (Optical Gravitational Lensing Experiment) survey from the Astronomical Observatory of the University of Warsaw.
Researchers develop versatile paper-based electronic devices demonstrating both neuromorphic computing capabilities and physically unclonable functions for security applications.