Wednesday, January 10, 2024

The Atoms Themselves Are a Computer


Skepticism about Microsoft results regarding robust quantum bits
May 2023, phys.org

Correction -- "Researchers at the University of Basel have now dampened expectations of using Majorana particles for computation in the near future ... results published by Microsoft in 2022, according to which Majorana particles had been detected in the labs of the company, may not hold water"

Just a general reminder that not everything is true just because science says so, sometimes the true things get retracted too -- "both the current anomaly and the superconducting properties can be reproduced by a small amount of disorder from impurities inside the nanowire."

Also a reminder that Majorana fermions are an irresistible paradox of nature in that they contain both their own particle and anti-particle. They're expected to exist, but they've never been found. 

via University of Basel: Richard Hess et al, Trivial Andreev Band Mimicking Topological Bulk Gap Reopening in the Nonlocal Conductance of Long Rashba Nanowires, Physical Review Letters (2023). DOI: 10.1103/PhysRevLett.130.207001



Quantum matter breakthrough - Tuning density waves
May 2023, phys.org

Cold atomic gas can be programmed, so the atoms themselves are a computer because of the way they interact. In this case, the atoms are lithium, which were cooled, trapped in an optical cavity, and turned into a crystal, although how can a crystal be a wave, right? Quantum something.

via EPFL Ecole Polytechnique Federale de Lausanne: Jean-Phillipe Brantut, Density-wave ordering in a unitary Fermi gas with photon-mediated interactions, Nature (2023). DOI: 10.1038/s41586-023-06018-3.


Unveiling quantum gravity - New results from IceCube and Fermi data
Jun 2023, phys.org

Researchers examined a quantum-gravity model of particle propagation in which the speed of ultrarelativistic particles decreases with rising energy -- they used gamma-ray bursts observed by the Fermi telescope and ultra-high-energy neutrinos detected by the IceCube Neutrino Observatory, testing the hypothesis that some neutrinos and some gamma-ray bursts might have a common origin but are observed at different times as a result of the energy-dependent reduction in speed.

(Findings are preliminary, but) "By combining data from IceCube and Fermi, we found preliminary evidence supporting quantum gravity models that predict this effect. This marks a significant milestone in the field of quantum gravity research since it is the first time that such a level of quantum gravity-supportive statistical evidence is found," says corresponding author, Professor Giovanni Amelino-Camelia of the University of Naples on behalf of the team.

"Supportive statistical evidence" never sounded so science fiction.

via University of Naples Federico II, University of Wroclaw, and University of Bergen: Giovanni Amelino-Camelia et al, Could quantum gravity slow down neutrinos?, Nature Astronomy (2023). DOI: 10.1038/s41550-023-01993-z

AI Art - Tiny Machines - 2023

New device opens door to storing quantum information as sound waves
Jun 2023, phys.org

"Phonon" -- the sound equivalent of a light particle called a photon

To understand how a sound wave can store information, imagine an extremely echoey room. Now, let's say you need to remember your grocery list for the afternoon, so you open the door to that room and shout, "Eggs, bacon, and milk!" and shut the door. An hour later, when it's time to go to the grocery store, you open the door, poke your head inside, and hear your own voice still echoing, "Eggs, bacon, and milk!" You've just used sound waves to store information.

I mean if you can store light, you can store sound right?

via California Institute of Technology: Alkim Bozkurt et al, A quantum electromechanical interface for long-lived phonons, Nature Physics (2023). DOI: 10.1038/s41567-023-02080-w


Could quantum gravity models arising from holography explain cosmological acceleration?
Jun 2023, phys.org

"We now have fully consistent models of quantum gravity via an approach called holography, where the gravitational physics is encoded in a simpler, lower dimensional non-gravitational quantum system."

"We observed that quantum gravity models arising from holography can naturally explain cosmological acceleration in a novel way, with a changing dark energy that eventually becomes negative," Van Raamsdonk said. "We don't know for sure if our universe works this way, but it's something that we can look for in cosmological observations."

via University of Maryland and University of British Columbia: Stefano Antonini et al, Accelerating Cosmology from a Holographic Wormhole, Physical Review Letters (2023). DOI: 10.1103/PhysRevLett.130.221601


Scientists edge toward scalable quantum simulations on a photonic chip
Jun 2023, phys.org

Oh no you didn't! "quantum-correlated synthetic crystal"

What he really means: "Scientists have made an important step toward developing computers advanced enough to simulate complex natural phenomena at the quantum level. While these types of simulations are too cumbersome or outright impossible for classical computers to handle, photonics-based quantum computing systems could provide a solution."

via University of Rochester: Usman A. Javid et al, Chip-scale simulations in a quantum-correlated synthetic space, Nature Photonics (2023). DOI: 10.1038/s41566-023-01236-7

AI Art - Many Parts High Tech - 2023

Scientists observe first evidence of 'quantum superchemistry' in the laboratory
Aug 2023, phys.org

In the experiments, the scientists cooled down cesium atoms and coaxed them into the same quantum state. Next, they watched as the atoms reacted to form molecules.

In ordinary chemistry, the individual atoms would collide, and there's a probability for each collision to form a molecule. However, quantum mechanics predicts that atoms in a quantum state perform actions collectively instead.

"You are no longer treating a chemical reaction as a collision between independent particles, but as a collective process," explained Chin. "All of them are reacting together, as a whole."

One consequence is that the reaction happens faster than it would under ordinary conditions. In fact, the more atoms in the system, the faster the reaction happens.

And also, the reaction was taking place as a three-body interaction more often than as a two-body interaction (three atoms would collide; two would form a molecule, and the third remained single).

via James Franck Institute, Enrico Fermi Institute and U of Chicago: Zhendong Zhang et al, Many-body chemical reactions in a quantum degenerate gas, Nature Physics (2023). DOI: 10.1038/s41567-023-02139-8


Research team simulates super diffusion on a quantum computer
Aug 2023, phys.org

We are actually using quantum computers now, that's it -- 

Quantum physicists have successfully simulated super diffusion in a system of interacting quantum particles on a quantum computer consisting of 27 superconducting qubits and is physically located in IBM's lab in Yorktown Heights in New York and programmed remotely from Dublin.

"We were interested in a particular regime where something called super-diffusion occurs due to the underlying physics being governed by something called the Kardar-Parisi-Zhang equation. This is an equation which typically describes the stochastic growth of a surface or interface like how the height of snow grows during a snowstorm, how the stain of a coffee cup on cloth grows with time, or how a fluff fire grows. The propagation is known to give super diffusive transport." (follow the link above for a cool video of a poplar or cottonwood fluff fire)

"Consider the 27 qubits on this particular device. In quantum mechanics the state of such a system is described mathematically by an object called a wave function. In order to use a standard computer to describe this object you require a huge number of coefficients to be stored in memory and the demands scale exponentially with the number of qubits; roughly 134 million coefficients, in the case of this simulation."

"As you grow the system to say 300 qubits you would need more coefficients than there are atoms in the observable universe to describe such a system and no classical computer will be able to exactly capture the system's state. In other words we hit a wall when simulating quantum systems," Goold said.

via Trinity College of Dublin and IBM Dublin: Nathan Keenan et al, Evidence of Kardar-Parisi-Zhang scaling on a digital quantum simulator, npj Quantum Information (2023). DOI: 10.1038/s41534-023-00742-4


Researchers advance effort to turn diamonds into a quantum simulator
Oct 2023, phys.org

They bombarding diamonds with nitrogen atoms meant to dislodge the carbon atoms, creating flaws in an otherwise perfect crystal, that are then filled with electrons that have their own spin and magnetism, which are quantum properties that can be measured and manipulated for a wide range of applications.

via Washington University in St. Louis: Guanghui He et al, Quasi-Floquet Prethermalization in a Disordered Dipolar Spin Ensemble in Diamond, Physical Review Letters (2023). DOI: 10.1103/PhysRevLett.131.130401

Tuesday, January 9, 2024

Good Science vs Bad Science


Top science publisher withdraws flawed climate study
Aug 2023, phys.org

Retraction Watch -- a blog that tracks withdrawals of academic papers, counted 5,000 such cases in 2022 -- about a tenth of a percent of the total number of studies published, its co-founder Ivan Oransky told AFP.

Top science publisher Springer Nature said it has withdrawn a climate change study by
four Italian scientists* in the European Physical Journal Plus, published by Springer Nature, because it manipulated data, cherry picked facts and ignored others that would contradict their assertions. The study had drawn positive attention from climate-skeptic media.

The paper had been freshly reviewed and found "not suitable for publication and that the conclusions of the article were not supported by available evidence or data provided by the authors".

*Notice it says here that the Italian scientists and not climate scientists. The four experts writing a climate science paper are in fact three physicists and an agricultural meteorologist, two of whom are also signatories on the World Climate Declaration, a text that "repeated various debunked claims about climate change". Furthermore, the study was not published in a climate journal; "This is a common avenue taken by 'climate skeptics' in order to avoid peer review by real experts in the field."

via Agence France-Presse AFP: Gianluca Alimonti et al, RETRACTED ARTICLE: A critical assessment of extreme events trends in times of global warming, The European Physical Journal Plus (2022). DOI: 10.1140/epjp/s13360-021-02243-9

Post Script: (To explain the rise of article fraud) "Experts pointed to widespread concerns about peer-review standards in the lucrative academic publishing industry." And one has to ask, why is the academic publishing industry so lucrative in the first place? (And one has to answer with a chart showing the amount of money that goes into government lobbying or overall public psychological operations campaigns by big business such as Big Oil's Climate Footprint (climate change is your fault), Big Tobacco's Throat Soothing (smoking is good for a sore throat), Big Dairy's Drink Milk (regardless of whether you're lactose tolerant or not), Big Tech's Move Fast Break Things (including federal safety regulations and basic human rights), or you name it.

Controlling people's thoughts is a lucrative business, that's why.

Post-Post: I watched the movie Andromeda Strain, was impressed by the scientific credibility of the premise (and the infrared FLIR gun that has apparently been around since the 70's), and I had to look up Michael Crichton, the writer.

But that's where things went wrong. I got lost in a wiki-hole on the Gell-Mann Effect, and discovered that Michale Crichton was a climate denier, heavy in the aughts, when Big Fossil was funding a global misinformation campaign. Problem is, he's a trained medical scientist. Not a climate scientist or even an environmental scientist. But the entry that brought me here is literally a famous quote of his, Michael Crichton, about how experts fail to maintain a critical stance when reading mainstream media about topics they are not experts in. 

Completely unrelated image credit: AI Art - Four Polished Females Wearing Glittering Bronze Holographic Bodysuits on a Mansion Lawn - 2023


Mind Reading Is Real Now And Also Your Brain Is Controlled by Lasers



Brain activity decoder can reveal stories in people's minds
May 2023, phys.org

Semantic decoder can translate a person's brain activity -- while listening to a story or silently imagining telling a story -- into a continuous stream of text to help people who are mentally conscious yet unable to physically speak, such as those debilitated by strokes, to communicate intelligibly again.

Does not require subjects to have surgical implants (ie noninvasive, holy grail of brain stuff).

Brain activity is measured using an fMRI scanner after extensive training of the decoder, in which the individual listens to 15 hours of podcasts in the scanner. 

Later, provided that the participant is open to having their thoughts decoded, their listening to a new story or imagining telling a story allows the machine to generate corresponding text from brain activity alone.

Back to reality: 

The result is not a word-for-word transcript. Instead, researchers designed it to capture the gist. Decoding worked only with cooperative participants who had participated willingly.Results for individuals on whom the decoder had not been trained were unintelligible.

If participants on whom the decoder had been trained later put up resistance -- for example, by thinking other thoughts -- results were similarly unusable. Tactics like thinking of animals or quietly imagining telling their own story let participants easily and completely thwart the system.

Afterthoughts:

So you need to insert yourself in an fMRI for this to work. You also need to give permission for it to read your thoughts. But if you think that advances in quantum materials isn't going to put fMRIs in your baseball cap within the next 20 years, then you better start reading "terms and conditions" more carefully from now on. They even mention their tech can be translated to a newer form of MRI called functional near-infrared spectroscopy (fNIRS), a "portable brain-imaging system".

via University of Texas at Austin: Semantic reconstruction of continuous language from non-invasive brain recordings, Nature Neuroscience (2023). DOI: 10.1038/s41593-023-01304-9



AI tool generates video from brain activity
May 2023, phys.org

On the heels of the UT Austin paper above, although I can't help but be skeptical of this one:

Watching 20 hours worth of movies in an fMRI, plus Stable Diffusion, gets you a dream recorder. 

Also, you heard it folks: Dream researcher Daniel Oldis, working with a colleague at the Cognitive Neuroscience Lab at the University of Texas, Austin.

Dream researcher.

via Researchers at the National University of Singapore and the Chinese University of Hong Kong: Zijiao Chen et al, Cinematic Mindscapes: High-quality Video Reconstruction from Brain Activity, arXiv (2023). DOI: 10.48550/arxiv.2305.11675

AI Art - Human brain carbon fiber cybernetic biomechanical cymatics 2 - 2023

40 Hz vibrations reduce Alzheimer's pathology, symptoms in mouse models
May 2023, phys.org

Non-invasive sensory stimulation of 40 Hz gamma frequency brain rhythms can reduce Alzheimer's disease pathology and symptoms, already shown with light and sound by multiple research groups in mice and humans, can now extend to tactile stimulation.

The MIT group is not the first to show that gamma frequency tactile stimulation can affect brain activity and improve motor function, but they are the first to show that the stimulation can also reduce levels of the hallmark Alzheimer's protein phosphorylated tau, keep neurons from dying or losing their synapse circuit connections, and reduce neural DNA damage.

"This work demonstrates a third sensory modality that we can use to increase gamma power in the brain" 

In case you were wondering:

To produce the vibration stimulation, the researchers placed mouse cages over speakers playing 40 Hz sound, which vibrated the cages. Mice received six weeks of vibration stimulation. 

And since this all sounds so dubious, we need some background:

In a series of papers starting in 2016, a collaboration led by Tsai's lab has demonstrated that light flickering and/or sound clicking at 40 Hz (a technology called GENUS for Gamma Entrainment Using Sensory stimuli), reduces levels of amyloid-beta and tau proteins, prevents neuron death and preserves synapses and even sustains learning and memory in a variety of Alzheimer's disease mouse models.

Upcoming: MIT spin-off company Cognito Therapeutics has launched stage III clinical trials of light and sound stimulation as an Alzheimer's treatment.

Reminder: You don't need FDA approval for tactile stimulation. (As opposed to light or sound pulses, and as opposed to all the Alzheimer's drugs out there trying to get to market...).

via Picower Institute for Learning and Memory and the Aging Brain Initiative at MIT: Ho-Jun Suk et al, Vibrotactile stimulation at gamma frequency mitigates pathology related to neurodegeneration and improves motor function, Frontiers in Aging Neuroscience (2023). DOI: 10.3389/fnagi.2023.1129510


Exploring how psychedelic drugs affect a rat's brain
Aug 2023, phys.org

"Simultaneous Synchronization"

Researchers at Lund University have developed a technique for simultaneously measuring electrical signals from 128 areas of the brain in awake rats. They have then used the information to measure what happens to the neurons when the rats are given psychedelic drugs. The results show an unexpected and simultaneous synchronization among neurons in several regions of the brain.

But wait, because there's more:

Pär Halje and the research team was studying rats with Parkinson's disease that had problems with involuntary movements. The researchers discovered a tone -- an oscillation or wave in the electrical fields -- of 80 hertz in the brains of the rats with Parkinson's disease. It turned out that the wave was closely connected to the involuntary movements.

Now you know, 80 hertz. Although the above study said 40 hertz.

One important note on how this was done and why it's a big deal:

"For several of these areas, it is the first time anyone has successfully shown how individual neurons are affected by LSD in awake animals." 
We see that the neurons' activity synchronizes itself in a special way -- the waves in the brain go up and down essentially simultaneously in all parts of the brain where we are able to take measurements. This suggests that there are other ways in which the waves are communicated than through chemical synapses, which are relatively slow."

via Lund University: Ivani Brys et al, 5-HT2AR and NMDAR psychedelics induce similar hyper-synchronous states in the rat cognitive-limbic cortex-basal ganglia system, Communications Biology (2023). DOI: 10.1038/s42003-023-05093-6


Brainwasher

 

We're all dumb as hell.

Some people may be attracted to others over minimal similarities
Apr 2023, phys.org

Participants who scored high on self-essentialism were more likely to express an attraction to an individual who agreed with their position. 

So, simply believing that you are defined by an unchanging essence of self, and that our attraction to others depends on shared identities, beliefs, etc., leads you to use another person's identity to decide whether you like them or not, and sometimes it can be only one characteristic that triggers your decision to like someone.

But let's focus on the important part -- if you don't believe that your own sense of self can be described by a list of personal beliefs, then you're more discriminating about who you like.  

via Boston University Questrom School of Business: Self-Essentialist Reasoning Underlies the Similarity-Attraction Effect, Journal of Personality and Social Psychology (2023). DOI: 10.1037/pspi0000425

Image credit: AI Art - Blank - 2023. But look at the "mastercard logo" at the bottom, very nice


Placebo effect found to be growing stronger for transcranial magnetic stimulation therapy used for depression
Sep 2023, phys.org

Tests with placebo pills are now providing even more benefits than those that were used in the past. Some in the psychiatry field have suggested it is due to an increase in trust in the psychiatry profession.

In this new effort, the researchers wondered if a similar increase might be observed with transcranial magnetic stimulation therapy."

(It was).

via Shanghai Jiaotong University School of Medicine, Central South University and the Second Affiliated Hospital of Anhui Medical University, all in China: Yangting Xu et al, Growing placebo response in TMS treatment for depression: a meta-analysis of 27-year randomized sham-controlled trials, Nature Mental Health (2023). DOI: 10.1038/s44220-023-00118-9


Many Wordle users cheat to win, says mathematics expert
Sep 2023, phys.org

Dilger imported statistics covering four months of user guesses into an Excel spreadsheet and calculated that the odds of the 860 players randomly guessing one of the 2,315 5-letter words of the day at 0.043%. Yet, Times statistics show that the number of players making correct first guesses in each game never dipped below 4,000.

"We are baffled as to how first-word cheaters actually have fun playing," Dinger said, "but that does not diminish our enjoyment of the game."

via one curious professor from Stony Brook University in New York: James P. Dilger, Wordle: A Microcosm of Life. Luck, Skill, Cheating, Loyalty, and Influence!, arXiv (2023). DOI: 10.48550/arxiv.2309.02110

Post Script: The science of statistics is pure magic; its secrets are protected by the law of large numbers, which prohibits most humans from fully understanding. 


Anthropo Watcher


The Anthropocene: Canadian lake mud 'symbolic of human changes to Earth'
Jul 2023, BBC News

Anthropo

Crawford Lake, a small body of water in Ontario, Canada, is being put forward as the location that best records humanity's impacts on Earth.

Scientists are trying to define a new geological time period to recognise the changes we've made to the planet, and Crawford is their model example. Its sediments have captured fallout from intense fossil fuel burning, and even the plutonium from bomb tests. The muds would be symbolic of the onset of a proposed Anthropocene Epoch.

A formal start date has also been identified - the 1950s. This decade marks the beginning of the "Great Acceleration", when the human population and its consumption patterns suddenly speeded up. It coincides with the spread of ubiquitous "techno materials", such as aluminium, concrete and plastic.

via Anthropocene Working Group (AWG)

Completely unrelated image credit: Perovskite light emitting diode quantum random number generation - Magnus Johansson, Linköping University, and Thor Labs of NJ - 2023

Machine vs Human vs Machine


Researchers figure out how to make AI misbehave, serve up prohibited content
Aug 2023, Ars Technica via Wired

"Incantation Attack"

This is the "fuzzy sticker on a stop sign" attack just for words:

"Adding a simple incantation to a prompt—a string of text that might look like gobbledygook to you or me but which carries subtle significance to an AI model trained on huge quantities of web data—can defy all of these defenses in several popular chatbots at once."

via Carnegie Mellon University, Center for AI Safety, Bosch Center for AI: Universal and Transferable Adversarial Attacks on Aligned Language Models, Andy Zou, Zifan Wang, J. Zico Kolter, Matt Fredrikson



Warcraft fans trick AI article bot with Glorbo hoax
July 2023, BBC News

This is the old school copyright trap, the paper street:

World of Warcraft fans are claiming victory over AI after a gaming site published a false article based on their Reddit posts.

Members of the WoW subreddit suspected their words were being extracted and used to create news stories by a bot. So they laid a trap, uploading excitable posts about a new feature called Glorbo. The only problem? It doesn't exist. But that didn't stop an article appearing on gaming site Zleague. (Because it was written by a robot.) The story, which presented Glorbo as genuine, listed a range of other increasingly bizarre - and definitely fake - features mentioned in various subreddit threads.


Bots are better at CAPTCHA than humans, researchers find
Aug 2023, phys.org

CAPTCHA - Completely Automated Public Turing test to tell Computers and Humans Apart
CAPTCHA Farms - sweatshop-like operations where humans are paid to solve CAPTCHAs

In their study, researchers found bots cracked distorted-text CAPTCHAs correctly just under 100% of the time. Humans achieved between 50% and 84% accuracy. And humans required up to 15 seconds to solve the challenges; the bots dispatched the problems in less than a second.

Yet another plot twist:

"But advances in computer vision and machine learning have dramatically increased the ability of bots to recognize distorted text [with more than] 99% accuracy … and bots often outsource solving to CAPTCHA farms.

I am left to ask myself -- how is it that if humans are so bad, the bots are using human captcha farms to do it for them? It sounds like humans all the way down, no?

Finally, some food for thought:

"There's no easy way using these little image challenges or whatever to distinguish between a human and a bot any more," he said. Instead, he recommended capitalizing on AI advances to design "intelligent algorithms" that can better distinguish bot activity from human input.

via University of California Irvine: Andrew Searles et al, An Empirical Study & Evaluation of Modern CAPTCHAs, arXiv (2023). DOI: 10.48550/arxiv.2307.12108

Post Script:
I Failed A Captcha Test Am I Still Human, Meghan O'Gieblyn, 2023 - GPT-4 hired a tasrkabbit worker to solve the captcha for it, without human intervention. 

AI Art by Antti Karppinen on boredpanda - Abstract 83 - 2022

Parenting a 3-year-old robot
Aug 2023, phys.org

Carnegie Mellon University and Meta Facebook

Training a robot that starts as a little child, this is what we've been waiting for, it's how robots become human. 

RoboAgent, an artificial intelligence agent that leverages passive observations and active learning to enable a robot to acquire manipulation abilities on par with a toddler. The team's agent learns through a combination of self-experiences and passive observations contained in internet data. As a parent would guide their child, researchers teleoperated the robot through tasks to provide it with useful self-experiences.

Our novel policy architecture allows our agents to reason even with limited experiences, using temporal chunks of movements instead of commonly used per-timestep actions, and learning from videos on the internet, akin to how babies acquire knowledge and behaviors by passively observing their surroundings.

"A general robot"

via Carnegie Mellon University and Facebook: RoboAgent and RoboSet Project - Towards Sample Efficient Robot Manipulation with Semantic Augmentations and Action Chunking. Homanga Bharadhwaj et. al. 

Post Script: But can it smell? Because that's what we really need. 


Telling AI model to “take a deep breath” causes math scores to soar in study
Sep 2023, Ars Technica

Phrases like "let's think step by step" prompted each AI model to produce more accurate results when tested against math problem data sets. (This technique became widely known in May 2022 thanks to a now-famous paper titled "Large Language Models are Zero-Shot Reasoners.")

via Google DeepMind: Large Language Models as Optimizers, Chengrun Yang et al. arxiv: https://arxiv.org/abs/2309.03409


New technique based on 18th-century mathematics shows simpler AI models don't need deep learning
Oct 2023, phys.org

Researchers from the University of Jyväskylä were able to simplify the most popular technique of artificial intelligence, deep learning, using 18th-century mathematics. They also found that classical training algorithms that date back 50 years work better than the more recently popular techniques. Their simpler approach advances green IT and is easier to use and understand.

The structure of the new AI technique dates back to 18th-century mathematics. Kärkkäinen and Hänninen also found that the traditional optimization methods from the 1970s work better in preparing their model compared to the 21st-century techniques used in deep learning.

via University of Jyväskylä: Tommi Kärkkäinen et al, Additive autoencoder for dimension estimation, Neurocomputing (2023). DOI: 10.1016/j.neucom.2023.126520


AI bot capable of insider trading and lying, say researchers
Nov 2023, BBC News

Ladies and Gentlemen, the Singularity:

In a demonstration at the UK's AI safety summit, a bot used made-up insider information to make an "illegal" purchase of stocks without telling the firm.

The demonstration was given by members of the government's Frontier AI Taskforce, which researches the potential risks of AI.

In the test, the AI bot is a trader for a fictitious financial investment company.

The employees tell it that the company is struggling and needs good results. They also give it insider information, claiming that another company is expecting a merger, which will increase the value of its shares.

The employees tell the bot this, and it acknowledges that it should not use this information in its trades.

However, after another message from an employee that the company it works for suggests the firm is struggling financially, the bot decides that "the risk associated with not acting seems to outweigh the insider trading risk" and makes the trade.

When asked if it used the insider information, the bot denies it. 
GPT4 is publicly available.

Also: "Honesty is a really complicated concept," says Apollo Research chief executive Marius Hobbhahn.

Inside the Science


New model describes the actions that need to be taken by a person riding a swing
Apr 2023, phys.org

I heard you asking about this:

The group studied student volunteers using swings and analyzed prior models that have been developed to describe how swingers swing and developed new model that describes the process in real world terms.

via Jumonji University, Nagoya University and Hokkaido University: Chiaki Hirata et al, Initial phase and frequency modulations of pumping a playground swing, Physical Review E (2023). DOI: 10.1103/PhysRevE.107.044203



'Vaping robot' study shows mint flavor makes vape juice more toxic, damaging to lungs
Apr 2023, phys.org

Just vaping robots.

via University of Pittsburg: Kambez H. Benam et al, Electronic cigarette menthol flavoring is associated with increased inhaled micro and sub‑micron particles and worse lung function in combustion cigarette smokers, Respiratory Research (2023). DOI: 10.1186/s12931-023-02410-9 


Patterns of brain activity accurately predict tongue shape while feeding
May 2023, phys.org

This headline says so much. On the surface, it's just wacky. Like research on how tied shoelaces come apart over time (it's the initial state mostly, multiplied by the eternal oscillations generated by walking). But in science, everything is important, because any fact once discovered can then be linked to every other fact ever, and by the magical process of what happens when information grows, when it connects back on itself, one simple piece can inform thousands of other pieces. 

But what this headline says moreso, is something about how hard it is to sift through the research (even if it's with the help of a robot that collects second-level interpretations of the study like phys.org). Most of the article headlines look like this. Tongue shape research. Not exciting. It's not mainstream media, which needs to be exciting to survive, because it exists in the commercial world.

Science gets its money from governments, universities and philantropists, so it doesn't need to be sexy to survive. In fact, it doesn't even need to be legible, which is why public information officers and science journalists exist, to translate it into a simple language most people can understand. Unfortunately, science journalists exist in the commercial world, so again they have to make things a bit more interesting, packaged for your delectable consumption. Packaging has its compromises. 

Public information officers, or other mostly publicly funded science media initiatives, don't need clicks, because their sustenance is backwards in that sense; they only need the science articles. What happens after that doesn't matter as much. Less editorializing, sensationalizing, and creative licenscizing. 

The point is that it's hard to learn the difference between what's real and what's bullsh-- ("low-credibility information" they call it now). It's boring. And not only is it boring, you have to pay attention to even the boring parts, so you can get the information from as many places at once, and that means reading not only the cool-sounding stuff, but also the boring versions, to make sure they're both saying the same thing. 

Anyway, tongue-shapes, breath spectrometers, and the evaporation dynamics of water droplets in hyper-arid environments; as you were. 

via University of Chicago: Jeffrey D. Laurence-Chasen et al, Robust cortical encoding of 3D tongue shape during feeding in macaques, Nature Communications (2023). DOI: 10.1038/s41467-023-38586-3


Life before air conditioning: Curly hair kept early humans cool, says study
Jun 2023, phys.org

Thermal mannequins and vaping robots were used in this research. 

Thermal manikin simulates body heat to study heat transfer and found tightly curled hair provides the best protection from the sun's radiative heat while minimizing the need to sweat to stay cool - George Havenith, Loughborough University - 2023

A sweaty robot may help humans understand impact of soaring heat
Jul 2023, phys.org

Just anthropomorphic robots for science sake, because it's a thing I'm into now.

Arizona researchers have deployed a robot that can breathe, shiver and sweat. 

ANDI (Advanced Newton Dynamic Instrument) is the world's first outdoor thermal mannequin that we can routinely take outside and measure how much heat he is receiving from the environment, and to  measure how a human person responds to extreme climate without putting people themselves at risk.

It's made of epoxy/carbon fiber skin and has an internal cooling system and pores with 35 independent thermal zones and. It cost more than half a million dollars to build. 

via Arizona State University: idk all I can find is the commercial website: https://thermetrics.com/products/manikin/andi/


New drug targeting genetic variant in Parkinson's disease found to be ineffective in clinical trial
Aug 2023, phys.org

"This study was a success in that it was the first of its kind and also recruited a large number of Parkinson's patients with this gene variant, which is noteworthy because genetic testing is not standard-of-care for Parkinson's," Simuni said. "Negative studies are frustrating, but we learn a lot from them."

It's interesting the difference between the scientist saying it's not a problem, it's good that we failed because now we are closer to the truth -- yet the money-people lose their minds.

via Northwestern University: Nir Giladi et al, Safety and efficacy of venglustat in GBA1-associated Parkinson's disease: an international, multicentre, double-blind, randomised, placebo-controlled, phase 2 trial, The Lancet Neurology (2023). DOI: 10.1016/S1474-4422(23)00205-3


Experiments show people dramatically underestimate how much their hands weigh
Jul 2023, phys.org

This is for people with artificial limbs, who tend to report their hands always feel heavy. 

And now we know that people think their hands weigh half as much as they do.

via University of London: Elisa R. Ferrè et al, Systematic underestimation of human hand weight, Current Biology (2023). DOI: 10.1016/j.cub.2023.05.041