Science

13821 readers
12 users here now

Subscribe to see new publications and popular science coverage of current research on your homepage


founded 5 years ago
MODERATORS
1
2
3
4
5
 
 

Expanded edition of the paper on AI in national research ecosystems focuses on the Global South

@science

https://council.science/news/ai-national-scence-second/?utm%5C_source=rss&utm%5C_medium=rss&utm%5C_campaign=ai-national-scence-second

6
 
 

Call for proposals to host in-person meeting of the Liaison Committee of the ISC Regional Focal Point for Latin America and the Caribbean | deadline: 10 March

@science

https://council.science/news/call-proposals-in-person-meeting-rfp-lac-lc-2025/?utm%5C_source=rss&utm%5C_medium=rss&utm%5C_campaign=call-proposals-in-person-meeting-rfp-lac-lc-2025

7
 
 

Andisearch Writeup

The Greenberger-Horne-Zeilinger (GHZ) paradox demonstrates a stark contradiction between quantum mechanics and local realism through an "all-versus-nothing" test of quantum nonlocality[^1]. First proposed in 1989 by Daniel Greenberger, Michael Horne, and Anton Zeilinger for four particles, the paradox was refined to three particles in 1990 with input from Abner Shimony[^1].

The paradox centers on the GHZ state, a highly entangled quantum state of three or more qubits, typically written as:

|GHZ⟩ = (|000⟩ + |111⟩)/√2

For three photons, this represents a superposition where all photons are either horizontally polarized (HHH) or vertically polarized (VVV)[^1].

The key elements that create the paradox are:

  1. Perfect correlations between measurements on separated particles
  2. EPR's definition of "elements of reality"
  3. The assumption of local realism

The contradiction emerges through a set of four quantum mechanical predictions[^1]:

Y₁Y₂X₃|GHZ⟩ = +|GHZ⟩
Y₁X₂Y₃|GHZ⟩ = +|GHZ⟩
X₁Y₂Y₃|GHZ⟩ = +|GHZ⟩
X₁X₂X₃|GHZ⟩ = -|GHZ⟩

These predictions are incompatible with any local hidden variable theory, where measurements must have definite classical values. The mathematical contradiction appears because in quantum mechanics:

Y₁Y₂X₃ · Y₁X₂Y₃ · X₁Y₂Y₃ · X₁X₂X₃ = -1

While in local hidden variable theories, this product must equal +1[^1].

Recent developments have expanded the GHZ paradox beyond its original formulation:

  • Multi-setting versions allow observers to measure more than two observables[^2]
  • Extensions to higher-dimensional quantum systems (qudits) have been developed[^9]
  • Applications in quantum communication, cryptography, and secret sharing protocols have emerged[^1]

The first experimental observation of GHZ correlations was achieved by Anton Zeilinger's group in 1998, work that contributed to his share of the 2022 Nobel Prize in Physics[^1].


[^1]: Greenberger–Horne–Zeilinger state - Wikipedia

[^2]: Multisetting Greenberger-Horne-Zeilinger paradoxes

[^9]: Greenberger-Horne-Zeilinger paradoxes from qudit graph states

(😐 I think that the percentage of quantum physicists who require high doses of medication is especially high.)

8
 
 

=o

9
 
 

I went through my bookmarks and found an old hacker news discussion thread where people are going in circles with some quite sincerely insisting that crows are more intelligent or every bit as intelligent as humans and that it's a kind of specieism and arrogance to suggest humans are more intelligent.

I felt like I was losing my mind reading that thread, which I think is why I bookmarked it.

I get appreciating the remarkable intelligence of animals and understanding their capabilities and the application of different forms of intelligence in different contexts. And the importance of having humility when it comes to understanding human intelligence and how a lot of our productive capacity comes from standing on the shoulder of giants. But take all of those caveats and add them all together and none of them I think at the end of the day amount to the idea that we should be uncertain about whether humans are more intelligent than crows.

I think there's a trap here of vortex of excessive humility that seems like a virtuous principle, but ends up missing the forest for the trees and putting people in the preposterous position of insisting that there's nothing special about humans building jumbo jets or being able to run hospitals compared to crows who apparently in the right circumstances could if they wanted to.

So I'm not crazy, right? Can reasonable people agree that humans are more intelligent than crows? And if that question sounds like a crazy question to ask in the first place, I'm glad you agree. But check out the Hacker News thread and try not to lose your mind.

https://news.ycombinator.com/item?id=24583981

10
 
 

For anyone interested in the open access research paper associated with this write up, here you go: https://www.nature.com/articles/s41551-025-01350-7

11
12
13
 
 

A recent study reveals that age plays a significant role in the outcomes of intermittent fasting. Researchers discovered that chronic intermittent fasting disrupted the development of insulin-producing beta cells in young mice. The findings raise concerns about potential risks for humans, especially teenagers.

14
15
 
 

Andisearch Writeup

A new study published in Science reveals that humpback whale songs share statistical structures with human language, specifically Zipfian distribution and Zipf's law of brevity. Researchers analyzed eight years of humpback whale song recordings from New Caledonia using methods typically used to evaluate infant speech. Key findings: * Whale songs exhibit statistically coherent subsequences conforming to Zipfian distribution, similar to human language. * The lengths of these subsequences adhere to Zipf's law of brevity. * This suggests that cultural transmission and learning play a significant role in shaping communication systems across species. The study highlights the importance of cultural transmission in learning complex communication systems and challenges the idea that such structural properties are exclusive to human language. Researchers suggest that humpback whales may learn their songs by tracking transitional probabilities between sound elements, similar to how human babies learn language. The research team was led by Inbal Arnon, Emma Carroll, and Jenny Allen.

16
 
 

cross-posted from: https://lemmy.ml/post/25907293

Constructing artificial intelligence that aligns with human values is a crucial challenge, with political values playing a distinctive role among various human value systems. In this study, we adapted the Political Compass Test and combined it with rigorous bootstrapping techniques to create a standardized method for testing political values in AI. This approach was applied to multiple versions of ChatGPT, utilizing a dataset of over 3000 tests to ensure robustness. Our findings reveal that while newer versions of ChatGPT consistently maintain values within the libertarian-left quadrant, there is a statistically significant rightward shift in political values over time, a phenomenon we term a ‘value shift’ in large language models. This shift is particularly noteworthy given the widespread use of LLMs and their potential influence on societal values. Importantly, our study controlled for factors such as user interaction and language, and the observed shifts were not directly linked to changes in training datasets. While this research provides valuable insights into the dynamic nature of value alignment in AI, it also underscores limitations, including the challenge of isolating all external variables that may contribute to these shifts. These findings suggest a need for continuous monitoring of AI systems to ensure ethical value alignment, particularly as they increasingly integrate into human decision-making and knowledge systems.

17
 
 

N=133,000 over 40 year time period

Observational study so limitations may apply but the findings do hold after adjusting for a large list of factors I will quote here:

educational attainments, family history of dementia, menopausal status with hormone use status, total energy intake, regular antidepressant drug use, history of depression, BMI, moderate-to-vigorous physical activity, neighborhood SES [socioeconomic status], marital status, living arrangement, smoking status, histories of hypertension, diabetes, and hypercholesterolemia and intakes of sugar-sweetened beverages, fruits, vegetables, whole grains, poultry, fish, eggs, nuts and legumes, low-fat dairy products, high-fat dairy products, and alcohol

18
19
14
submitted 2 weeks ago* (last edited 2 weeks ago) by shreddy_scientist@lemmy.ml to c/science@lemmy.ml
 
 

Researchers have unveiled the precise brain mechanisms that enable animals to overcome instinctive fears. The study in mice could have implications for developing therapeutics for fear-related disorders such as phobias, anxiety and post-traumatic stress disorder (PTSD).

20
21
22
23
24
25
view more: next ›