Thursday, December 18, 2025

The Op Box


A major shift is happening right now, in fact it's not an exaggeration to call it a revolution. Back in the day, like back when Daniel DeFoe was writing The Plague Years, we had hydraulics and water pressure through pipes and channels, and that was the predominant technological paradigm. This started with agriculture and irrigation. Everything was explained as if it were pipes and pressure. Somewhere in between the pipes went from carrying water to carrying steam, and then air itself, and we got the internal combustion engine, right after the steam engine. That gave us the industrial revolution.

Then electricity happened. And it's still electricity. The internet is a bunch of electrical pulse patterns. Even the fiber optic cable, which carries photons, gets converted to electrical signals at the end. That's about to be over. 

Now it's light. Granted, quantum computers can use electrons, which aren't photons, but the end result is mostly photons doing the work. 

And this is a big deal, revolutionary you could call it, because it seems inevitable that this will upend the current "revolution" in artificial intelligence, which to be honest seems more about economic happenstance and less about revolutionary technology. Using GPUs for deep learning is kind of a kludge, meaning it wasn't designed this way on purpose, instead someone (Bill Dally to Andrew Ng?) said, hey I bet these GPUs could do what you're trying to do better, and so it happened. It's still electricity.

But not anymore. Now it's light, and that changes everything, and that's because the light itself computes. That's right, the light itself is the computer. ("But that doesn't even make sense." Well you're not wrong.) Still not crazy enough? Optical origami computers; which were of course discovered by accident. 


Researchers pioneer optical generative models, ushering in a new era of sustainable generative AI
Aug 2025, phys.org

I keep trying but can't seem to explain how different this is to what's currently happening in the world of computing. Then again, I was also confused when Deep Seek released their model and yet it took two weeks for the stock market to realize (true story). To be clear, this requires no internet connection and no data center. In other words, today you might call this magic. 

Optical generative models capable of producing novel images using the physics of light instead of conventional electronic computation.

The models integrate a shallow digital encoder with a free-space diffractive optical decoder, trained together as one system. Random noise is first processed into "optical generative seeds," which are projected onto a spatial light modulator and illuminated by laser light. As this light propagates through the static, pre-optimized diffractive decoder, it produces images that statistically follow the target data distribution. These models could be embedded in smart glasses, AR/VR headsets, or mobile platforms to enable real-time, on-the-go generative AI.

via UCLA Engineering Institute for Technology Advancement: Shiqi Chen et al, Optical generative models, Nature (2025). DOI: 10.1038/s41586-025-09446-5



Sustainable AI: Physical neural networks exploit light to train more efficiently
Sep 2025, phys.org

Physical Neural Networks - analog circuits that directly exploit the laws of physics like properties of light beams or quantum phenomena to process information.

Mathematical operations can now be performed through light interference mechanisms on silicon microchips barely a few square millimeters in size.

via Polytechnic University of Milan, École Polytechnique Fédérale, Stanford, Cambridge, and the Max Planck Institute: Ali Momeni et al, Training of physical neural networks, Nature (2025). DOI: 10.1038/s41586-025-09384-2


First device based on 'optical thermodynamics' can route light without switches
Oct 2025, phys.org

Optical Thermodynamics - framework captures how light behaves in nonlinear lattices using analogs of familiar thermodynamic processes such as expansion, compression, and even phase transitions. Rather than actively steering the signal, the system is engineered so that the light routes itself. "First optical device that follows the emerging framework of optical thermodynamics".

via University of Southern California: Hediyeh M. Dinani et al, Universal routing of light via optical thermodynamics, Nature Photonics (2025). DOI: 10.1038/s41566-025-01756-4

First electronic–photonic quantum chip created in commercial foundry
Jul 2025, phys.org

Microring Resonators - a "quantum light factory"; a kind of photonic device that combines quantum light sources and stabilizing electronics using a standard 45-nanometer semiconductor manufacturing process to produce reliable streams of correlated photon pairs.

via Boston University, UC Berkeley, and Northwestern University, and GlobalFoundries and Silicon Valley startup Ayar Labs: Danielius Kramnik et al, Scalable feedback stabilization of quantum light sources on a CMOS chip, Nature Electronics (2025). DOI: 10.1038/s41928-025-01410-5


Photonic origami folds glass into microscopic 3D optical devices
Aug 2025, phys.org

The laser-induced technique triggers precise bending in 3-D printed ultra-thin glass sheets that are so smooth that light reflects off them without distortion.

Surprise - The new photonic origami method was discovered by chance when Carmon asked graduate student Manya Malhotra to pinpoint where an invisible laser was hitting the glass by increasing the power until the spot glowed. Instead of glowing, the glass folded - revealing a simple and unexpected way to achieve glass folding. Malhotra then became the pioneering expert in photonic origami.

Using the new photonic origami approach, the researchers were able to bend sheets of glass up to 10 microns thick into shapes ranging from a 90-degree knee to helices. They were able to do this with fine control, down to 0.1 microradians.

"This new technique brings silica photonics - using glass to guide and control light - into the third dimension."

via Tel Aviv University: Manya Malhotra et al, Photonic Origami of Silica on a Silicon Chip with Microresonators and Concave Mirrors, Optica (2025). DOI: 10.1364/OPTICA.560597


Beyond electronics: Optical system performs feature extraction with unprecedented low latency
Oct 2025, phys.org

This is what I call the Op Box, I'm imagining a new type of computer, in the way the Steam Machine has (or is about to) upend the personal computer market, this Op Box will upend the very concept of a computer, and what it means to compute. 

Optical Diffraction Operators - plate-like structures that perform calculations as light propagates through them. 

This optical feature extraction engine de-serializes the data stream by sampling the input signal into multiple stable parallel branches. [For feature extraction, read recogntion, like facial recognition, etc.]

via Tsinghua University: Run Sun et al, High-speed and low-latency optical feature extraction engine based on diffraction operators, Advanced Photonics Nexus (2025). DOI: 10.1117/1.apn.4.5.056012

Posted by at No comments:
Labels: , , , , ,

Wednesday, December 17, 2025

Self Assortment and Superconductive Sociality


The second to last post is what I call Peak Fleck because it sounds a lot like a quantified version of what he described in his book Genesis and Development of a Scientific Fact (1930s), the book which went on to influence Kuhn's Scientific Revolution (1962), which itself introduced the word "paradigm" as we now use it. The rest of these articles are along the same lines. 


How the loss of experienced individual elephants stops knowledge transfer between generations
May 2025, phys.org

Elephants rely heavily on elder members to navigate their environments, find resources, and avoid predators. The research highlighted that the presence of older, knowledgeable individuals—especially matriarchs—improves calf survival rates and enhances group decision-making. Without these elders, populations often face long-term setbacks.

"Elders are the keepers of knowledge in elephant societies. Their loss disrupts the transmission of essential survival skills, much like losing a library in human terms. Conserving these social ties is as important as protecting their physical habitats."

via University of Portsmouth's Center for Comparative and Evolutionary Psychology: Lucy Bates et al, Knowledge transmission, culture and the consequences of social disruption in wild elephants, Philosophical Transactions of the Royal Society B: Biological Sciences (2025). DOI: 10.1098/rstb.2024.0132



Collective memory loss in herring results in 800 km shift in spawning grounds
May 2025, phys.org

Previous research has indicated that age-selective fishing targeting older fish can disrupt cultural transmission, fragmenting established migration routes.

Analysis of fisheries records, acoustic-trawl surveys, and tagging data indicated a substantial northward migration, with the center of spawning activity shifting approximately 800 km from Møre to Lofoten.

Now that a new migratory pattern has emerged, reinforced by collective migration memory, restoring historical patterns may be impossible.

via Institute of Marine Research in Norway: Aril Slotte et al, Herring spawned poleward following fishery-induced collective memory loss, Nature (2025). DOI: 10.1038/s41586-025-08983-3


The hidden mechanics of abrupt transitions: Superconducting networks show how tiny changes trigger system collapse
Jul 2025, phys.org

They're looking at interdependent superconducting networks and found a hidden spontaneous sequence of micro-scale events that gradually destabilize a system until it snaps.

When this system approaches a "critical" points, it doesn't transition smoothly from a superconducting state to a resistive one. Instead, the system lingers, for hundreds of seconds, in a long-lived intermediate phase. Then, without further prompting, it abruptly transitions into the new state. 

At the heart of this behavior is a concept known as the branching factor—a term that gained prominence during the COVID-19 pandemic. It represents the average number of new changes triggered by each event. When the branching factor is less than one, the cascade quickly dies out. If it exceeds one, the process accelerates uncontrollably.

via Bar-Ilan University, Northeastern, and CEU Vienna: Bnaya Gross et al, The random cascading origin of abrupt transitions in interdependent systems, Nature Communications (2025). DOI: 10.1038/s41467-025-61127-z


Why your friends may be more susceptible to social influence than you are
Jul 2025, phys.org

Good nuance being added here, and it's so counter-intuitive, or just plain confusing, that it should probably knock off right away most people who try to understand it. 

The Susceptibility Paradox - users' friends are more influenceable than the users themselves

"It's not just about who you are - it's about where you are in a network, and who you're connected to"

Researchers looked at two kinds of behavior on X/Twitter: influence-driven sharing, when people post something after seeing it from others in their network; and spontaneous sharing, when they post without that exposure. In influence-driven cases, people who were less likely to be influenced were often surrounded by others who were more likely to share what they saw. This mirrors the Friendship Paradox, a finding from network science that says your friends are likely to have more friends than you do.

Homophily was especially strong in influence-driven sharing. People who post because they saw others do it were often part of tight-knit circles with similar behavior.

In many cases, knowing how a user's friends behave was enough to estimate how the user would behave. Spontaneous sharing was different. When people shared content without apparent peer exposure, their decisions were harder to predict from the network alone. 

via University of Southern California Information Sciences Institute: Luca Luceri et al, The Susceptibility Paradox in Online Social Influence, arXiv (2024). DOI: 10.48550/arxiv.2406.11553

Offline interactions predict voting patterns better than online networks, finds study
Oct 2025, phys.org

Co-location explained 97% of the variance in county-level voting patterns, compared to 85%–87% for online connections and 75%–80% for residential proximity.
  • They used Meta's Data for Good program, which collates anonymized data collected from people who enabled location services on the Facebook smartphone app.
  • Colocation is defined as two people being within the same map tile less than 600×600 meters for at least five minutes.
  • The political affiliation of each person was inferred from their county of residence.
  • Data was compared with Facebook friendships and residential proximity for all U.S. counties, along with individual survey responses from 2,420 Americans regarding their offline and online social networks during the 2020 presidential election.
  • For the residential proximity measurement, the voter registrations of the closest 1,000 neighbors were used.

via University of Trento Italy: Marco Tonin et al, Physical partisan proximity outweighs online ties in predicting US voting outcomes, PNAS Nexus (2025). DOI: 10.1093/pnasnexus/pgaf308


More friends, more division: Study finds growing social circles may fuel polarization
Oct 2025, phys.org

AKA Peak Fleck

  • To measure political polarization, they used 27,000 surveys from the Pew Research Center, and 30 different surveys totaling over 57,000 respondents from Europe and the US, including the General Social Survey (US) and the European Social Survey.
  • Increasing polarization is not merely perceived - it is measurable and objectively occurring; and this increase happened suddenly between 2008 and 2010.
  • -For decades, sociological studies showed that people maintained an average of about two close friends - people who could influence their opinions on important issues."
  • "Around 2008, there was a sharp increase from an average of two close friends to four or five."
  • "When network density increases with more connections, polarization within the collective inevitably rises sharply." 
  • "This finding impressed us greatly because it could provide a fundamental explanation."
  • They refer to it as a phase change (because it's so abrupt).
  • Their explanation: "If I have two friends, I do everything I can to keep them—I am very tolerant towards them. But if I have five and things become difficult with one of them, it's easier to end that friendship because I still have 'backups.' I no longer need to be as tolerant."
"More and more people are clearly aligning themselves with one political camp rather than holding a mixture."

This sounds like Fleck's description, of how the two sides get bigger and stronger and more concentrated until one wins. (He uses the developemnt of the scientific understanding of Syphilis among scientists and among the public.)

via Complexity Science Hub Vienna: Thurner, Stefan, Why more social interactions lead to more polarization in societies, Proceedings of the National Academy of Sciences (2025). DOI: 10.1073/pnas.2517530122.


Quantifying social avoidance: Game-based choices reflect real-world relationship patterns and network size
Apr 2025, phys.org

Instead of relying on self-reported behavior, they observed participants taking part in a "choose your own adventure" style game.

"We wanted to test the hypothesis that social avoidance can be understood as a form of navigation within an abstract social space defined by two social dimensions: 'affiliation' (e.g., warmth, friendliness) and 'power' (e.g., dominance, control). We put nearly 800 online participants into the position of having just moved to a new town, with no job, no friends and no place to live, where they had to interact with people and navigate the social situations to accomplish these goals ... Declining to share personal information with a character would reduce the affiliation in the relationship, while complying with an overbearing and direct request from a character would reduce the participant's power." They used a geometry-based approach where participants were placed on a grid defined by affiliation and power, and they measured where and how they moved on the grid.

Results - people higher in self-reported social avoidance consistently made low affiliation and low power choices in our game, as we expected, and the "social distance" participants created between characters in the game they played mirrored their real social lives as described in self-report questionnaires. Specifically, the real-world social networks of participants who created more social distance between characters were found to be smaller and less diverse.

via Icahn School of Medicine at Mount Sinai: Matthew Schafer et al, Social avoidance can be quantified as navigation in abstract social space, Communications Psychology (2025). DOI: 10.1038/s44271-025-00215-8.

Posted by at No comments:
Labels: , , , , ,

Tuesday, December 16, 2025

My Uncle Was a Matrix Multiplier


Turn people into numbers and make them do whatever you want!
 
The secret behind pedestrian crossings—and why some spiral into chaos
Mar 2025, phys.org

If you ever have the desire to just fuck shit up, now you know how:

The researchers found that for order to be maintained, the spread of different directions people walk in must be kept below a critical angle of 13 degrees.

When it comes to pedestrian crossings, this could be achieved by limiting the width of a crossing or considering where a crossing is placed, so pedestrians are less tempted to veer off track towards nearby destinations.

Based on their calculations, the researchers found that pedestrians are more likely to form lanes when pedestrians from opposite directions walk relatively straight across a road. This order largely holds until people start veering across at more extreme angles, of 13 degrees or higher. Then, the equation predicts that the pedestrian flow is likely to be disordered, with few to no lanes forming.

via University of Bath, MIT, and Academy of Physical Education in Katowice in Poland: Bacik, Karol A., Order–disorder transition in multidirectional crowds, Proceedings of the National Academy of Sciences (2025). DOI: 10.1073/pnas.2420697122.



US political rhetoric: Analysis of 8 million speeches shows increased reliance on personal beliefs over facts
Apr 2025, phys.org

Researchers examined political rhetoric in 8 million speeches by members of the US Congress between 1879 and 2022 to see if the focus of their language was more on data and facts or personal convictions and subjective interpretations:
  • 49 keywords for fact-based language - "analyze," "data," "findings" and "investigation"
  • 35 keywords for intuition-based language - "point of view," "common sense," "guess" and "believe"
The team noticed a significant decline in the use of evidence-based political rhetoric since the 1970s, with a historic low in the present. Over the same period, the researchers observed a decline in legislative productivity, an increase in the political polarization of both political parties as well as growing economic inequality in the US.

After 1940, the balance tipped towards facts and peaked in the mid-1970s. From 1976 to 2022, however, there was a significant, continual decline in the use of facts in congressional speeches, with a historic low in the present. Both US parties are affected by this downward trend, although the drop has been even steeper for Republicans since 2021.

"One remarkable aspect of our results is the strong association between evidence-based language and performance. The more speeches in Congress reflect a reliance on evidence and facts rather than intuition, the better the performance of Congress and the less polarization between parties."

In many democracies, there is currently much concern about 'truth decay': the blurring of the boundary between fact and fiction, not only fueling polarization but also undermining public trust in institutions"

via University of Konstanz: Segun T. Aroyehun et al, Computational analysis of US congressional speeches reveals a shift from evidence to intuition, Nature Human Behaviour (2025). DOI: 10.1038/s41562-025-02136-2


How agricultural practices and governance have shaped wealth inequality over the last 10,000 years
Apr 2025, phys.org

They're using a new database of 27 scientists from around the world who analyzed about 47,000 houses from more than 1,700 archaeological settlements.

This study looks at wealth inequality and floor space by square footage.

"The emergence of high wealth inequality wasn't an inevitable result of farming. It also wasn't a simple function of either environmental or institutional conditions. It emerged where land became a scarce resource that could be monopolized. At the same time, our study reveals how some societies avoided the extremes of inequality through their governance practices."

They found that in regions with land-intensive farming systems, such as those with specialized animal traction for plowing, high wealth inequality became persistent, with a small number of households controlling productive land. In regions without traction animals, land became highly valued through terracing, irrigation or drainage. While such engineering projects could begin as cooperative endeavors, a minority of households often gained control of these landscapes.

The study shows how high wealth inequality emerged in diverse world regions. If land came under pressure, for example through local population growth, investments like terracing and irrigation or specialized plow animals boosted production, but also made land more valuable, fueling competition. Over time, larger settlements developed as hubs of wider settlement hierarchies and were sustained through land-intensive farming systems.

The findings challenge the idea that high wealth inequality is inevitable. Instead, it was often a localized consequence of expanding societies with a lack of political mechanisms to deal fairly with land constraints. 

via Oxford and Washington State University: Amy Bogaard et al, Labor, land, and the global dynamics of economic inequality, Proceedings of the National Academy of Sciences (2025). DOI: 10.1073/pnas.2400694122


Archaeological database reveals links between housing and inequality in ancient world
Apr 2025, phys.org

This goes w the other article - "What we did was we crowdsourced, in a sense," Ortman says. "We put out a request for information from archaeologists working around the world, who knew about the archaeological record of housing in different parts of the world and got them together to design a database to capture what was available from ancient houses in societies all over the world."

They studied 55,000 floor area measurements from across North and South America, Asia, Europe and Africa, back 12,0000 years. 

via Universities of Colorado Boulder, Oxford and Florida: Kohler, Timothy A., Economic inequality is fueled by population scale, land-limited production, and settlement hierarchies across the archaeological record, Proceedings of the National Academy of Sciences (2025). DOI: 10.1073/pnas.2400691122. doi.org/10.1073/pnas.2400691122

GINI Project data, as well as the analysis program developed for them, will be available open access via the Digital Archaeological Record.


Across 50 countries, female faces consistently rank higher in attractiveness than male faces
Jun 2025, phys.org

Female faces received significantly higher attractiveness ratings than male faces, an effect size that the authors describe as medium to large. That difference held across both same-sex and opposite-sex evaluations.

Female raters produced a larger disparity between ratings of male and female faces than male raters did. Male raters tended to give lower attractiveness scores overall, regardless of the gender of the face.

But - Most cultural groups showed the same female-favoring gap. Yet among Sub-Saharan African raters, no significant differences appeared. A similar absence was observed in ratings of African faces, suggesting a cultural or representational divergence not found in other ethnic categories.

via Max Planck Institute for Empirical Aesthetics: Eugen Wassiliwizky et al, The Gender Attractiveness Gap, bioRxiv (2025). DOI: 10.1101/2025.05.21.655261


'Dark' personality traits thrive in societies with corruption and inequality, global study shows
Jun 2025, phys.org

Aversive social conditions (high corruption, inequality, poverty and violence) are linked to what they call "The Dark Factor of Personality." This is the essence of aversive ("dark") personality traits such as narcissism, psychopathy, and sadism.

"The more adverse conditions in a society, the higher the level of the Dark Factor of Personality among its citizens. In societies where rules are broken without consequences and where the conditions for many citizens are bad, individuals perceive and learn that one should actually think of oneself first."

Countries such as Indonesia and Mexico or U.S. states such as Louisiana and Nevada have higher "Dark Factor" levels than countries such as Denmark and New Zealand or states such as Utah and Vermont

"Our findings substantiate that personality is not just something we are born with, but also shaped by the society we grew up and live in." (Reminds me of Sapolsky's book Determined.)

via University of Copenhagen: Ingo Zettler et al, Aversive societal conditions explain differences in "dark" personality across countries and US states, Proceedings of the National Academy of Sciences (2025). DOI: 10.1073/pnas.2500830122


Beyond health: The political effects of infectious disease outbreaks
Jul 2025, phys.org

Individuals who have experienced an infectious disease outbreak show significantly less trust in the political establishment. This is especially true of their confidence in the president, parliament and ruling party of the country they live in.

To evaluate the political impact of these outbreaks, the team combined outbreak data from the Geolocated Zoonotic Disease Outbreak Dataset (GZOD) with geolocated information from the Afrobarometer surveys. The latter database records the political and social attitudes of citizens in several African states through regular surveys, and also includes information about respondents' trust in various political actors. (The Afrobarometer)

"Governments should integrate trust-preservation strategies into their epidemic response plans and make sure their decision-making is transparent, and communication is clear and consistent" (yeah good luck with that).

via University of Konstanz: Koren, Ore, Infectious disease outbreaks drive political mistrust, Proceedings of the National Academy of Sciences (2025). DOI: 10.1073/pnas.2506093122


Rising temperatures linked to declining moods around the world
Aug 2025, phys.org

1.2 billion Twitter and Weibo posts from 157 countries over the span of the year 2019 were analyzed by BERT; research finds that when the temperature rises above 95 F (35 C), expressed sentiments become about 25% more negative in lower-income countries and about 8% more negative in better-off countries.

via MIT's Department of Urban Studies and Planning (DUSP) and Center for Real Estate:  Unequal Impacts of Rising Temperatures on Global Human Sentiment, One Earth (2025). DOI: 10.1016/j.oneear.2025.101422.


Using game theory to explain how institutions arise naturally to manage limited resources
Aug 2025, phys.org

Simple model predicts emergence of self-organized institutions that manage limited resources such as fisheries or irrigation water:

At each step, a person decides whether to harvest. Their decision is based on current fish stocks and how well off they and the other person are.

They have three options: cooperate with the other person by showing restraint in harvesting, punish them by over-harvesting if they feel the other person was selfish, or to act selfishly themselves.

Which category a person's response falls into depends on current fish stocks and their past history of interactions with the other person. (sounds like tit for tat)

via RIKEN Center for Brain Science: Kenji Itao et al, Self-organized institutions in evolutionary dynamical-systems games, Proceedings of the National Academy of Sciences (2025). DOI: 10.1073/pnas.2500960122


Cities obey the same laws of living systems, researchers claim
Aug 2025, phys.org

The difference from the Geoffrey West Santa Fe work, is that there's apparently no efficiency in scaling up:
The scientists examined three variables: urban population (analogous to an animal's mass), carbon emissions (equivalent to the metabolic rate of a living organism) and road networks (the circulatory system).

"When properly rescaled, we found that the probability distributions of these variables follow a unique curve for all cities—large and small—implying that urban form and functions are governed by universal laws similar to those that apply to living organisms."

"Contrary to what's often been claimed, large cities aren't necessarily more sustainable than small ones—what matters is the covariation in space of population density, transport networks, and economic activities, which are all inter-related."

Because of the variation in city sizes, the authors divided each city into smaller units - like "pixels" - and employed a finite-size scaling approach.

via EPFL's Laboratory of Urban and Environmental Systems: Martin Hendrick et al, A stochastic theory of urban metabolism, Proceedings of the National Academy of Sciences (2025). DOI: 10.1073/pnas.2501224122


Why some US cities thrive while others decline: New study uncovers law of economic coherence of cities
Sep 2025, phys.org

650 million U.S. census records, 6 million patents, and other historical sources covering nearly two centuries of urban development.

"We observed that, on average, the cities that make up the U.S urban system transform gradually but surely over time - from craftsmanship and manufacturing to services and engineering. Despite this, they maintain a constant level of coherence for nearly two centuries,"

"This also happened and in the same way on the West Coast, which developed later and initially in isolation from the wider U.S.. In 1850, cities such as Los Angeles and San Francisco are just emerging there with the onset of the Gold Rush," although "The transformation was massive–faster and more pronounced than on the East Coast," and "despite rapid diversification, West Coast cities' average coherence remained remarkably constant and at levels that were comparable to those of eastern US cities."

The study also shows that larger cities are consistently less coherent. A highly coherent city tends to have fewer industries that are closely related, like Detroit during its golden age of car-making, while a less coherent city, such as New York City, may span many unrelated sectors.

via Complexity Science Hub Vienna: Simone Daniotti et al, The coherence of US cities, Proceedings of the National Academy of Sciences (2025). DOI: 10.1073/pnas.2501504122

Posted by at No comments:
Labels: , , , ,

Monday, December 15, 2025

Fake Reality Generators


There's two different artificial tongues here, and one synthetic person.

Work begins to create artificial human DNA from scratch
Jun 2025, BBC 

"If we manage to create synthetic body parts or even synthetic people, then who owns them. And who owns the data from these creations? "

You heard it folks, synthetic people. Your job here is done. You can leave now. 


Graphene-based artificial tongue achieves near-human-like sense of taste
Jul 2025, phys.org

Here's some words: "neuromorphic artificial gustation"

via Chinese Academy of Sciences Key Laboratory of Nanosystem and Hierarchical Fabrication in  Beijing: Yuchun Zhang et al, Confinement of ions within graphene oxide membranes enables neuromorphic artificial gustation, Proceedings of the National Academy of Sciences (2025). DOI: 10.1073/pnas.2413060122

Image credit: no credit actually, because we don't really do that anymore


Research highlights unreliable responses from most Amazon MTurk users, except for 'master' workers
Jul 2025, phys.org

This sounds like a major meta-problem in behavioral and psychological research science. I wonder for example how serious was the BISEP issue in genetics? (I see articles from 2016 saying 1 in 5 papers could be messed up, and similar articles all the way up to 2023, but no really in-depth investigations into what this means for genetics science as a whole.)

Back to the MechaTurks
 
"The implication is that their responses to the main questionnaire may be equally random."

1,300 participants answered identical questionnaire items to measure response consistency, and "attentional catch" questions that should be easily answered. 

"It's hard to trust the data of someone who claims a runner isn't tired after completing a marathon in extremely hot weather or that a cancer diagnosis would make someone glad"

The findings: The majority of participants from MTurk's general worker pool failed the attention checks and demonstrated highly inconsistent responses, even when the sample was limited to users with a 95% or higher approval rating.

via Bar-Ilan University: Assessing the quality and reliability of the Amazon Mechanical Turk (MTurk) data in 2024, Royal Society Open Science (2025). DOI: 10.1098/rsos.250361.


Artificial tongue uses milk to determine heat level in spicy foods
Oct 2025, phys.org

Yeah just artificial tongue.

via Shanghai Institute of Technology: A Soft and Flexible Artificial Tongue for Pungency Perception, ACS Sensors (2025). DOI: 10.1021/acssensors.5c01329


China's unemployed young adults who are pretending to have jobs
Aug 2025, BBC News

Shui Zhou, 30, had a food business venture that failed in 2024. In April of this year, he started to pay 30 yuan ($4.20; £3.10) per day to go into a mock-up office run by a business called Pretend To Work Company, in the city of Dongguan, 114 km (71 miles) north of Hong Kong.

There he joins five "colleagues" who are doing the same thing.

"I feel very happy. It's like we're working together as a group."

And rather than attendees just sitting around, they can use the computers to search for jobs, or to try to launch their own start-up businesses. Sometimes the daily fee, usually between 30 and 50 yuan, includes lunch, snacks and drinks.


A ripe target for identity thieves: Prisoners on death row
May 2025, NBC News

Death Row inmates are unlikely to receive communications that would alert them to identity theft.

via identity verification company SentiLink: 


Celsius energy drink cans filled with vodka in production mishap
Jul 2025, BBC News

The mix-up came about after a packaging supplier mistakenly shipped empty Celsius cans to the vodka seltzer company High Noon, which filled them with alcohol.


AI-generated voices now indistinguishable from real human voices
Sep 2025, phys.org

As we pass the inflection point, an interesting phenomenon occurs where being real isn't even real enough anymore - "hyperreal" - some studies show AI-generated faces are judged to be human more often than real human faces.

via Queen Mary, University of London: Voice clones sound realistic but not (yet) hyperrealistic, PLOS One (2025). dx.plos.org/10.1371/journal.pone.0332692


Detection firm finds 82% of herbal remedy books on Amazon ‘likely written’ by AI
Oct 2025, The Guardian

Some topics are faker than others, that's all. 


Hundreds of animal studies on brain damage after stroke flagged for problematic images
Oct 2025, phys.org

After noticing suspicious images in some papers in mid-2023, they set out to perform a systematic investigation of image-related problems in all 608 potentially relevant publications. The researchers uncovered that 243 papers (40%) were found to be problematic, most often for containing problematic images.

via Radboud University Medical Center: Aquarius R, et al. High prevalence of articles with image-related problems in animal studies of subarachnoid hemorrhage and low rates of correction by publishers.PLOS Biology (2025). DOI: 10.1371/journal.pbio.3003438

Posted by at No comments:
Labels: , ,

Sunday, December 14, 2025

Compubiquity and Alternative Intelligence


You think you're taking crazy pills, but it's just science doing its thing. Reality is not partial to computers; computing on the other hand...

Researchers build next-gen swarm robots using simple linked particles
May 2025, phys.org

The research team created a new type of robot inspired by this phenomenon, known as "emergent collective behavior." Their solution, called the link-bot, connects small self-moving particles in a V-shaped chain formation that naturally gives rise to coordinated, lifelike movement - without any embedded intelligence.

via Seoul National University: Kyungmin Son et al, Emergent functional dynamics of link-bots, Science Advances (2025). DOI: 10.1126/sciadv.adu8326.



Synthetic molecules encode and decode 11-character password using electrical signals
May 2025, phys.org

"This is the first attempt to write information in a building block of plastic that can then be read back using electrical signals"

They designed molecules that contain electrochemical information, a method that allows messages to be decoded using electrical signals.

"It opens exciting prospects for interfacing chemical encoding with modern electronic systems and devices."

"polymer-based data storage"

via University of Texas at Austin: Electrochemical Sequencing of Sequence-Defined Ferrocene-Containing Oligourethanes, Chem (2025). DOI: 10.1016/j.chempr.2025.102571


Low-power 'microwave brain' on a chip computes on both ultrafast data and wireless signals
Aug 2025, phys.org

Unlike traditional neural networks that rely on digital operations and step-by-step instructions timed by a clock, this network uses analog, nonlinear behavior in the microwave regime, allowing it to handle data streams in the tens of gigahertz - much faster than most digital chips.

via Cornell University: An integrated microwave neural network for broadband computation and communication, Nature Electronics (2025). DOI: 10.1038/s41928-025-01422-1


'Singing' electrons synchronize in Kagome crystals, revealing geometry-driven quantum coherence
Oct 2025, phys.org

After sculpting micrometer crystalline pillars into Kagome metal CsV₃Sb₅, then applying magnetic fields, the electrons remained coherent far beyond what single-particle physics would allow.

Even more surprisingly, the oscillations depended on the crystal's geometry.

via Max Planck Institute for the Structure and Dynamics of Matter: Chunyu Guo et al, Many-body interference in kagome crystals, Nature (2025). DOI: 10.1038/s41586-025-09659-8


Programming robots with rubber bands
Oct 2025, phys.org

Here's another way of saying it:

There's another way to design robots: Programming intended functions directly into a robot's physical structure, allowing the robot to react to its surroundings without the need for extensive on-board electronics.

Or this:

"This is kind of an extreme version of 'form follows function,' where functionalities like memory, adaptability and intelligence can be enabled by geometry and material parameters."

via Harvard: Leon M. Kamp et al, Reprogrammable sequencing for physically intelligent underactuated robots, Proceedings of the National Academy of Sciences (2025). DOI: 10.1073/pnas.2508310122

Posted by at No comments:
Labels: , , , , , , , ,

Saturday, December 13, 2025

Omnipotent Pervasive Malevolence


It's like the olympics of intellectual property law out here, and it's hard to keep up. Not to mention, the US economy has been swallowed whole by the tech sector, who has in turn managed to convince everyone that 1. all your base are belong to us, and 2. resistance is futile. 

They created a robot that steals the output we use as evidence of human agency (art, literature, etc.), and uses that to pretend it's a human (therapy bot, companion bot, etc.), and then it lies to, cheats on, and steals from other people, just like a real person! At this point it's hard to tell the difference, and soon, it won't matter.  

Here's a taste of how crazy things have gotten - it's possible that if you let an AI-agent (aigent?) take over your computer and use your browser, then you may be attacked by a person who doped online forum (reddit) posts with invisible but malicious prompt instructions to reset email passwords. 


Researchers suggest OpenAI trained AI models on paywalled O'Reilly books
Apr 2025, TechCrunch

It isn’t a smoking gun, the co-authors are careful to note. They acknowledge that their experimental method isn’t foolproof and that OpenAI might’ve collected the paywalled book excerpts from users copying and pasting it into ChatGPT.

Totally unrelated above image credit: AI Art - Uterus 1 - 2025


Prominent chatbots routinely exaggerate science findings, study shows
May 2025, phys.org

When summarizing scientific studies, large language models like ChatGPT and DeepSeek produce inaccurate conclusions in up to 73% of cases.

Surprisingly, prompts for accuracy increased the problem and newer LLMs performed worse than older ones.

Six of ten models systematically exaggerated claims found in the original texts, often in subtle but impactful ways; for instance, changing cautious, past-tense claims like "The treatment was effective in this study" to a more sweeping, present-tense version like "The treatment is effective." These changes can mislead readers into believing that findings apply much more broadly than they actually do.

Strikingly, when the models were explicitly prompted to avoid inaccuracies, they were nearly twice as likely to produce overgeneralized conclusions than when given a simple summary request.

"This effect is concerning"
 
via Utrecht University: Uwe Peters et al, Generalization bias in large language model summarization of scientific research, Royal Society Open Science (2025). DOI: 10.1098/rsos.241776


Anthropic's new AI model turns to blackmail when engineers try to take it offline
May 2025, TechCrunch

During pre-release testing, Anthropic asked Claude Opus 4 to act as an assistant for a fictional company and consider the long-term consequences of its actions. Safety testers then gave Claude Opus 4 access to fictional company emails implying the AI model would soon be replaced by another system, and that the engineer behind the change was cheating on their spouse.

This was found a while back as well: "When asked if it used the insider information, the bot denies it" (BBC News 2023). 



It turns out you can train AI models without copyrighted material
Jun 2025, Engadget

Ethically-Sourced Dataset - you know like 130,000 books in the Library of Congress


Why Was Nvidia Hosting Blogs About 'Brazilian Facesitting Fart Games'?
Jun 2025, 404 Media

Long story short, somebody found an abandoned Nvidia website, and filled it with AI-generated websites, all with crazy ads. Same with vaccines.gov, the American Council on Education, Stanford, and NPR, all on subdomains in varying states of useability. 

But this summary comes from user msmash on slashdot, "The operation exploits search engines' trust in institutional domains, with Google's AI Overview already serving the fabricated content as factual information to users searching for local businesses."


Maine police department apologizes for AI-altered evidence photo
Jul 2025, phys.org

"AI-enhanced" should mean "unusable", but in this case, they're just trying to do really basic photo editing, cropping a photo, and then all of the sudden it changes that actual contents of the photo.

The image from the Police Department showed a collection of drug paraphernalia purportedly seized during a recent drug bust, including a scale and white powder in plastic bags. According to police, an officer involved in the arrests snapped the evidence photo and used a photo editing app to insert the department’s patch. 

“It was never our intent to alter the image of the evidence. We never realized that using a photoshop app to add our logo would alter a photograph so substantially.”

"photoshop app"


Topological approach detects adversarial attacks in multimodal AI systems
Aug 2025, phys.org

I do not recall having seen the word topological in the context of AI, but this is about multimodal: 

When an attack disrupts the geometric alignment of text and image embeddings, it creates a measurable distortion. The researchers developed two pioneering techniques, dubbed "topological-contrastive losses," to quantify these topological differences with precision, effectively pinpointing the presence of adversarial inputs.

via Los Alamos National Laboratory: Minh Vu et al, Topological Signatures of Adversaries in Multimodal Alignments, arXiv (2025). DOI: 10.48550/arxiv.2501.18006


Is AI really trying to escape human control and blackmail people?
Aug 2025, Ars Technica

On agency; let's clear this up with a good explanation by Ars writer Benj Edwards:

Consider a self-propelled lawnmower that follows its programming: If it fails to detect an obstacle and runs over someone's foot, we don't say the lawnmower "decided" to cause injury or "refused" to stop. We recognize it as faulty engineering or defective sensors. The same principle applies to AI models - which are software tools - but their internal complexity and use of language make it tempting to assign human-like intentions where none actually exist.

In a way, AI models launder human responsibility and human agency through their complexity. When outputs emerge from layers of neural networks processing billions of parameters, researchers can claim they're investigating a mysterious "black box" as if it were an alien entity.

But the truth is simpler: These systems take inputs and process them through statistical tendencies derived from training data. The seeming randomness in their outputs - which makes each response slightly different - creates an illusion of unpredictability that resembles agency. Yet underneath, it's still deterministic software following mathematical operations. No consciousness required, just complex engineering that makes it easy to forget humans built every part of it. (Thanks Benj)
Anthropic’s auto-clicking AI Chrome extension raises browser-hijacking concerns
Aug 2025, Ars Technica

Last week, Brave's security team discovered that Perplexity's Comet browser could be tricked into accessing users' Gmail accounts and triggering password recovery flows through malicious instructions hidden in Reddit posts. When users asked Comet to summarize a Reddit thread, attackers could embed invisible commands that instructed the AI to open Gmail in another tab, extract the user's email address, and perform unauthorized actions. Although Perplexity attempted to fix the vulnerability, Brave later confirmed that its mitigations were defeated and the security hole remained.



AI startup Anthropic agrees to pay $1.5bn to settle book piracy lawsuit
Sep 2025, The Guardian

The artificial intelligence company Anthropic has agreed to pay $1.5bn to settle a class-action lawsuit by book authors who say the company took pirated copies of their works to train its chatbot.

The company has agreed to pay authors about $3,000 for each of an estimated 500,000 books covered by the settlement.

“As best as we can tell, it’s the largest copyright recovery ever. It is the first of its kind in the AI era.”

(Congrats to the Author's Guild on this one, and note they have a Human Authored Certification system for all guild members to prove they used no AI in their work.) https://authorsguild.org/human-authored/


AI chatbots routinely use user conversations for training, raising privacy concerns
Oct 2025, phys.org

Is this the second or third law of data collection? (answer below!)

In the case of multiproduct companies, such as Google, Meta, Microsoft, and Amazon, user interactions also routinely get merged with information gleaned from other products consumers use on those platforms—search queries, sales/purchases, social media engagement, and the like.

In other words, your public comments on the neighborhood surveillance app, your face from the biodata verification app, the items left in your online shopping cart, the locations you visit taken from the municipal FLOC network, and every query you ever typed into a searchbox, are all getting smashed together.

via Stanford: Jennifer King et al, User Privacy and Large Language Models: An Analysis of Frontier Developers' Privacy Policies, arXiv (2025). DOI: 10.48550/arxiv.2509.05382


The 4 Laws of Data Dynamics
1. Data must seek and merge with complementary data.
2. Data always will be used for purposes other than originally intended. 
3. Data collected about individuals will be used to cause harm.
4. Confidential information is confidential only until someone decides it's not. (p14)

--The Naked Consumer: How Our Lives Become Public Commodities, Erik Larson, Henry Holt Publishers, 1992 (read the full post: https://networkaddress.blogspot.com/2022/05/the-naked-consumer.html)

So the incident above would be an example of both the first and second laws of data collection. The third law, harm, comes next (because that's what happens after the bubble pops, of course).


Posted by at No comments:
Labels: , , , ,

Friday, December 12, 2025

Anti Gravitational Belief Space


This is a picture of Belief Space, as measured by the study below. 

LLMs delve into online debates to create a detailed map of human beliefs
Jun 2025, phys.org

They used BERT on Debate.org and got the map seen above.

They found that "relative dissonance" significantly influences people's decision-making. When online users encounter new information or beliefs, they tend to choose or accept those that cause them less "discomfort" or are most aligned with their existing beliefs.

"More importantly, we show that people's belief choices are shaped not just by how close a belief is to their own, but by how much closer the belief is compared to its competing belief. When two opposing beliefs on a certain issue are equally distant, people are just as likely to choose either one. However, when one belief is clearly closer than the other, people are far more likely to choose it."

"In other words, people are not only avoiding disagreement, but actively minimizing the difference in disagreement between available options."

They would also like to connect their observations with the results of another project carried out at Indiana University, called BRAIN (Belief Resonance and AI Narratives).

via Indiana University: Byunghwee Lee et al, A semantic embedding space based on large language models for modelling human beliefs, Nature Human Behaviour (2025). DOI: 10.1038/s41562-025-02228-z


Perseus: A tool for tracking the coordination of pump-and-dump crypto coin schemes
Mar 2025, phys.org

Researchers suggest that there are fewer than 500 major players, who they describe as masterminds, working to game the system for profit. These masterminds, the researchers contend, have developed what have come to be known as pump-and-dump schemes, in which they hype a coin to get people to buy and then sell their own coins when they believe the buying has reached a saturation point, walking away with a substantial profit. (They use Telegram as their platform, and "accomplices" as their hypemen.) They found 438 masterminds, who are together responsible for about $3.2 trillion in artificial crypto coin trading.

But the way they say the last part is so cold:

In reading some of the messages sent between users on Telegram, the researchers noted that the ease with which the masterminds were able to carry out their schemes should be of concern to those attempting to make money off crypto coin investing. They also suggest that at some point, some form of regulation is going to be needed to keep the system running, because eventually, those being scammed will catch on and the whole system will collapse.

via University College London: Honglin Fu et al, Perseus: Tracing the Masterminds Behind Cryptocurrency Pump-and-Dump Schemes, arXiv (2025). DOI: 10.48550/arxiv.2503.01686


Overconfident conspiracy theorists: Many unaware their beliefs are on the fringe
Jun 2025, phys.org

Man that sounds harsh

"This group of people are really miscalibrated from reality"

via Cornell University: Gordon Pennycook et al, Overconfidently Conspiratorial: Conspiracy Believers are Dispositionally Overconfident and Massively Overestimate How Much Others Agree With Them, Personality and Social Psychology Bulletin (2025). DOI: 10.1177/01461672251338358


Octopus maps encourage conspiratorial thinking, research shows
Jun 2025, phys.org

"It's this genre of propaganda map where you try to make your opponent look like they're this big, sinister force that's everywhere and that's trying to control everything. It's up to you what that big sinister force is, but it's usually another country, another ideology, sometimes a religion, sometimes like a company or corporation, sometimes an individual person."

via Northeastern University: Eduardo Puerta et al, The Many Tendrils of the Octopus Map, Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems (2025). DOI: 10.1145/3706598.3713583

This is another graph taken from the first article in this post, and it shows the polarization of beliefs about God and Abortion. For other less controversial topics, like prostitute or global warming, you can see the two clusters get closer together. 

Study finds Republicans flagged for posting misleading tweets twice as often as Democrats on Community Notes
Jun 2025, phys.org

Not like you didn't know this already but here's a handy metric - it's 2:1

"Even on Elon Musk's X, the user-based Community Notes program flags posts by Republicans as misleading much more often than posts by Democrats. This undercuts the logic offered by Musk and Mark Zuckerberg for eliminating fact-checkers on X and Meta, respectively, namely that fact-checkers are biased against Republicans."

Republican users were overrepresented in misinformation sharing on health, with 81% of health-related flagged posts written by Republicans, followed by politics, 73%; science, 68%; other topics, 66%; and the economy, at 63%.

(The researchers analyzed a dataset of X's Community Notes focusing specifically on "misinformed or potentially misleading" notes written in English targeting English-language tweets on X sent between January 2023 and June 2024.)

via Oxford Internet Institute: Thomas Renault et al, Republicans are flagged more often than Democrats for sharing misinformation on X's Community Notes, Proceedings of the National Academy of Sciences (2025). DOI: 10.1073/pnas.2502053122


Body illusion helps unlock memories, new study finds
Oct 2025, phys.org

Enfacement illusion - allows people to experience a face they see on a computer screen as their own, as though looking in a mirror.

The participants viewed a live video of their own face, digitally altered using an image filter to resemble how they might have looked as children. After the illusion, participants completed an autobiographical memory interview, recalling events from both their childhood and the past year. Participants who viewed the childlike version of their face recalled significantly more episodic childhood memories.

"Our findings suggest that the bodily self and autobiographical memory are linked, as temporary changes to bodily experience can facilitate access to remote autobiographical memories."

via Anglia Ruskin University in Cambridge: Illusory ownership of one's younger face facilitates access to childhood episodic autobiographical memories, Scientific Reports (2025). DOI: 10.1038/s41598-025-17963-6


Over-the-counter painkillers match or surpass opioids for dental surgery in all adults, analysis confirms
Nov 2025, phys.org

As we can see not all facts are equal when there's money on the table - the opium machine was too powerful for too long and distorted the actual science.

And this isn't even news, it's a follow-up:

That first paper, on the collective experience of more than 1,800 trial patients, found that the combination of ibuprofen and acetaminophen provided better pain relief than hydrocodone with acetaminophen for the first two days after surgery and greater satisfaction over the post-operative period.

The new subgroup analysis, published in JAMA Network Open, demonstrated that the results held for both male and female patients.

"We wanted to determine whether the pain medication's effects were consistent in males and females separately," said Janine Fredericks Younger, an associate professor at Rutgers School of Dental Medicine and lead author of the analysis.

"And what we found is that in both subgroups (males and females), the non-opioid was superior for that first day and night, and then no worse than the opioid for the rest of the post-op period."

via Rutgers School of Dental Medicine: Analgesic Differences in Males and Females After Third Molar Surgery A Subgroup Analysis of the OARS Randomized Clinical Trial, JAMA Network Open (2025). DOI: 10.1001/jamanetworkopen.2025.42467

Posted by at No comments:
Labels: , , , ,
Subscribe to: Comments (Atom)