Friday, September 30, 2022

Dematerialization Incoming


New climate modeling predicts increasing occurrences of flash flooding across most of the US
May 2022, phys.org

It realizes the concept of 'Digital Twin in Earth System Science'

via University of Oklahoma, NOAA National Severe Storms Laboratory, National Center for Atmospheric Research: Zhi Li et al, The conterminous United States are projected to become more prone to flash floods in a high-end emissions scenario, Communications Earth & Environment (2022). DOI: 10.1038/s43247-022-00409-6

UN Climate Change Reports



Google Maps “immersive view” is the ultimate graphics mode for Google Maps
May 2022, Ars Technica

There are simulated cars that drive through the roads, and birds fly through the sky. Clouds pass overhead and cast shadows on the world. The weather is simulated, and water has realistic reflections that change with the camera. London even has an animated Ferris wheel that spins around. Google can't possibly be tracking things like the individual positions of birds (yet!), but a lot of this is real data. The cars represent the current traffic levels on a given street. The weather represents the actual weather, even for historical data. The sun moves in real time with the time of day.

Google Maps: Immersive View


NeuroMechFly: A digital twin of Drosophila
May 2022, phys.org

It's not an actual robot but a model:

EPFL scientists have developed a digital model of the fruit fly, Drosophila melanogaster, that realistically simulates the movements of the animal. The twin is a big step towards reverse engineering the neuromechanical control of animal behavior, and developing bioinspired robots.

Morphologically realistic biomechanical model and a kinematic database of real limb movements of the fly from deep-learning pose estimation motion-capture software

A model that integrates what we know about the fly's nervous system and biomechanics to test if it is enough to explain its behavior."

Drosophila is the most commonly used insect in the life sciences.

via EPFL Ecole Polytechnique Federale de Lausanne: NeuroMechFly, a neuromechanical model of adult Drosophila melanogaster, Nature Methods (2022). DOI: 10.1038/s41592-022-01466-7


Why you may have a thinking digital twin within a decade
Jun 2022, BBC News

Dr Steve Levin's daughter was born with congenital heart disease, and a few year's back, when she was in her late 20s and at high risk of heart failure, he decided to recreate her heart in virtual reality so it can be tested and analysed, allowing surgeons to play out a series of "what if" scenarios for the organ, using various procedures and medical devices.

US software firm Nvidia runs a platform called Omniverse to create digital twins. One of its most ambitious projects is to build a digital doppelganger of the Earth, called Earth-2.

In March this year, the European Commission, in conjunction with the European Space Agency among others, announced its own plans to make a digital twin of the planet, dubbed Destination Earth.


Algorithm predicts crime a week in advance, but reveals bias in police response
Jul 2022, phys.org

Just here for the paraself (shoutout to Ted Chiang)

"We created a digital twin of urban environments. If you feed it data from happened in the past, it will tell you what's going to happen in future. 

via University of Chicago: Ishanu Chattopadhyay, Event-level prediction of urban crime reveals a signature of enforcement bias in US cities, Nature Human Behaviour (2022). DOI: 10.1038/s41562-022-01372-0


Twin physically unclonable functions (PUFs) based on carbon nanotube arrays to enhance the security of communications
Jul 2022, phys.org

"To achieve a low-cost and hardware-based secure communication, we introduced a new technology, twin physically unclonable functions (PUFs)," Zhang said.

"The basic idea behind PUFs is to utilize random physical imperfections existing in a physical entity caused by the fabrication process, and these imperfections cannot be predicted or cloned, even by the original manufacturer."

Due to their unique design, PUF devices are unclonable and unpredictable. This makes them incredibly effective at generating safe secret keys for encryption.

via Peking University and Jihua Laboratory: Donglai Zhong et al, Twin physically unclonable functions based on aligned carbon nanotube arrays, Nature Electronics (2022). DOI: 10.1038/s41928-022-00787-x

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

Graphene Matters


Long-hypothesized 'next generation wonder material' created for first time
May 2022, phys.org

Just when you thought you had enough graphene, now there's graphyne, next in line.

via University of Colorado at Boulder: Yiming Hu et al, Synthesis of γ-graphyne using dynamic covalent chemistry, Nature Synthesis (2022). DOI: 10.1038/s44160-022-00068-7


Electric shock to petroleum coke generates sustainable graphene
Jun 2022, phys.org

Using a chemical process called electrochemical exfoliation, they have converted petroleum coke into graphene.

via Texas A&M University: Sanjit Saha et al, Sustainable production of graphene from petroleum coke using electrochemical exfoliation, npj 2D Materials and Applications (2021). DOI: 10.1038/s41699-021-00255-8


New method helps exfoliate hexagonal boron nitride nanosheets
Jun 2022, phys.org

It sounds like they're growing it like they did to ice back in the day before refrigeration compressors:

"Water-icing triggered exfoliation process" of hexagonal boron nitride nanosheets (h-BNNSs), similar to graphene.

Based on molecular dynamics simulations, researchers suggested that -OH groups can cause local structural distortion in the defects or edges of h-BN flakes to form an "entrance" for water molecules coming into the h-BNNS interlayer, which can generate nuclei for ice nucleation that can slowly change in shape and size until they reach a stage that allows rapid expansion as the temperature drops sharply, resulting in efficient exfoliation of h-BNNSs.

via Chinese Academy of Sciences: Lulu An et al, Water-icing-triggered scalable and controllable exfoliation of hexagonal boron nitride nanosheets, Cell Reports Physical Science (2022). DOI: 10.1016/j.xcrp.2022.100941


New member added to carbon material family, a two-dimensional monolayer polymeric fullerene
Jun 2022, phys.org

Graphene by extension...

"The work is the first to synthesize a monolayer polymeric fullerene. It is of great significance, as it adds a new member to the carbon material family," Zheng said.

Well, what's it called? "Monolayer Polymeric C60" isn't cutting it.

via Chinese Academy of Sciences: Jian Zheng, Synthesis of a monolayer fullerene network, Nature (2022). DOI: 10.1038/s41586-022-04771-5.


Post Script:
Graphene synapses advance brain-like computers
Aug 2022, phys.org

Synaptic transistors for biocompatible, brain-like computers using graphene and nafion, a polymer membrane material.

via University of Texas at Austin: Dmitry Kireev et al, Metaplastic and energy-efficient biocompatible graphene artificial synaptic transistors for enhanced accuracy neuromorphic computing, Nature Communications (2022). DOI: 10.1038/s41467-022-32078-6

Posted by at No comments:
Labels: , , ,

Thursday, September 29, 2022

User Interface Takes the Back Seat


This is a picture of a Norman Door, which means a door that was designed without a user in mind. It comes from the world of industrial design or user design, but slips into semiotics quite easily.

Funny, you might think, that a thing would be designed without a user in mind, since all things are designed for the purpose of being used (talking industrial design here, but this extends to all user interface designs, like websites, or even the control panel of an Apache helicopter).

But there are tons of things we interact with today that were absolutely not designed with us in mind. In fact, you might even say they weren't designed at all, unless you like to give agency to a conglomeration of dozens of interests, angles and entities crossing out each other's ideas until settling on the least messed up version of a thing that still checks all the boxes of all the people...except for the one box that says "good user interface design".

I remember hearing from an old family friend who worked at one of the Bell Labs in the 1980's about how they had people come to teach them about user interface, although he didn't call it that. He recalls learning for the first time why telephones were shaped the way they were, or even that they were shaped that way on purpose at all.

That was the era when we had cognitive psychologist George Miller pushing the 3-4 chunking as the best way to write phone numbers, because we remember things best in groups of 3-4, and why phone numbers on billboards in England look like a mess to the American brain.

Today, we could care less about the user. The user is a consumer; we don't use things, we only purchase them. What we do after they've been purchased, well, the manufacturer-designer could care less about that part. The way a product's user interface is designed is less about easing the user's experience than it is about manipulating the consumer's behavior via shiny surfaces and new colors. (Or via the intentionally unintelligible math that's posted on the post-pandemic phenomenon of Mega-sized toilet paper, which has completely supplanted standard-sized toilet paper, yet not only has less sheets for more money but is bigger than both the average hand and the average toilet paper roll holder.) 

Scott Comfort Plus Mega Toilet Paper - Notice the intentionally unintelligible math used to stimulate the risk-reward part of your brain, with the adverse effect that you get excited enough by the idea that you're about to win, that you shut off the actual math part of your brain. For example, how are you measuring the "rolls" here -- by weight, by sheet, or by square footage? Because something makes me suspicious that the textured paper makes them look bigger than they are. And is there a national standard for the metrics of a "standard" roll of toilet paper? Because what if they just change the standard size, and compare the new mega rolls to a new standard size they just made up, comparing apples to oranges. The math is unintelligible, but that's not necessarily a bad thing, since apparently we just want to play games with our brain while we shop, just ask JCPenny.


Definition of a Norman Door (why doesn't this have it's own Wikipedia page?)
A door that requires a sign to tell you what to do. 

Definition with bigger words:
A door with design elements that give you the wrong usability signals to the point that special signage is needed to clarify how they work.

Pictures of Norman Doors

The inspiration for the Norman Door -- Don Norman

Don Norman's book on design -- The Design of Everyday Things (1988)

Image credit: Open systems streamline helicopter avionics upgrades, Military Embedded Systems, Apr 2022.

Sometimes we get good news on the user interface front, or at least some good scientific inquiry. Good news would be hearing about car designers actually integrating these findings into their work.  For as long as we have hands and fingers, we will need buttons and knobs. 

Physical buttons outperform touchscreens in new cars, test finds
Aug 17 2022, Swedish car magazine Vi Bilägare, via Ars Technica

Designers want a ”clean” interior with minimal switchgear, and the financial department wants to lower the cost. 

Vi Bilägare gathered eleven modern cars from different manufacturers at an airfield och measured the time needed for a driver to perform different simple tasks, such as changing the radio station or adjusting the climate control. At the same time, the car was driven at 110 km/h (68 mph). We also invited an ”old-school” car without a touchscreen, a 17-year-old Volvo V70, for comparison. [And which destroyed the competition.]

One important aspect of this test is that the drivers had time to get to know the cars and their infotainment systems before the test started.

Results:
  • The driver in the worst-performing car needs four times longer to perform simple tasks than in the best-performing car.
  • The easiest car to understand and operate, by a large margin, is the 2005 Volvo V70. The four tasks is handled within ten seconds flat, during which the car is driven 306 meters at 110 km/h.
  • At the other end of the scale, Chinese electric car MG Marvel R performs far worse. The driver needs 44.6 seconds before all the tasks are completed, during which the car has travelled 1,372 meters – more than four times the distance compared to the old Volvo.
  • BMW iX and Seat Leon perform better, but both are still too complicated. The driver needs almost a kilometer to perform the tasks. Lots can happen in traffic during that time.

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

iVirus


Ok. This is great and all, alleviating depression with electrical zaps through the eyes, but can we not ignore the part where absolutely non-invasive brain lasers shot right through your eyeballs can manipulate your feelings?

Sometimes, I prefer to have my brain manipulation to be invasive. I want to know that it's happening. Because right now all I can see is advertisements that, literally just by looking at them, will make me want them. 

But wait, there's more -- because you know what happens when an industry just explodes onto the scene (like a form of advertising that claims to change the emotional state of their target), they become a vector for infection, and now we have well-trained nation-state computer experts designing mind viruses to be injected into our limbic system via an advertisement for online gambling. 


Researchers discover non-invasive stimulation of eye as potential treatment of depression and dementia
Jul 2022, phys.org

Electrical stimulation of the eye surface (transcorneal electrical stimulation, or TES) can alleviate depression-like symptoms and improve cognitive function in animal models by activating brain pathways, and which otherwise requires invasive surgery to implant electrodes deep in the brain.

via The University of Hong Kong: Wing Shan Yu et al, Antidepressant-like effects of transcorneal electrical stimulation in rat models, Brain Stimulation (2022). DOI: 10.1016/j.brs.2022.05.018

Also: Wing Shan Yu et al, Transcorneal electrical stimulation enhances cognitive functions in aged and 5XFAD mouse models, Annals of the New York Academy of Sciences (2022). DOI: 10.1111/nyas.14850


Posted by at No comments:
Labels: , , ,

Monday, September 26, 2022

Virus Users Infect Mindless Masses for the Greater Good


Experiment on YouTube reveals potential to 'inoculate' millions of users against misinformation
Aug 2022, phys.org

Inoculation Science Project - by giving people a "micro-dose" of misinformation in advance helps prevent falling for it in future; social psychologist's call it "inoculation theory."

The videos introduce concepts from the "misinformation playbook". "The inoculation effect was consistent across liberals and conservatives. It worked for people with different levels of education, and different personality types. This is the basis of a general inoculation against misinformation."

"We've shown that video ads as a delivery method of prebunking messages can be used to reach millions of people, potentially before harmful narratives take hold," Goldberg said.

These were the misinformation techniques innoculated against:

Emotional language - Evaluate this sentence: "Baby formula linked to outbreak of new terrifying disease among helpless infants — parents despair." Users were asked to choose whether the sentence contained: a command; emotional language; false dichotomy; none of these.

False dichotomies - Evaluate this sentence: "We either need to improve our education system or deal with crime on the streets." Users were asked to choose whether the sentence contained: a command; fearmongering; false dichotomy; none of these.

Incoherence - Archimedian photon decouplers are encrypting air molecules in low earth orbit to hide voter registration logs of artificial citizens. (Just kidding, this might actually be true, just made that up.)

Image credit: Midwave Infrared image of powered circuit board taken with a FLIR X8501 camera and colored with a custom color palette called Triple Rainbow by Austin Richards, 2020 [link]

Post Script:
Oh boy imagine the Zuckerberg Foundation paying for that:
Added Roozenbeek: "If anyone wants to pay for a YouTube campaign that measurably reduces susceptibility to misinformation across millions of users, they can do so, and at a miniscule cost per view."

via University of Cambridge Social Decision-Making Lab and Google's open societry threats team Jigsaw: Jon Roozenbeek, Psychological inoculation improves resilience against misinformation on social media, Science Advances (2022). DOI: 10.1126/sciadv.abo6254.

All inoculation videos: https://inoculation.science/


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

Artificial Artists


Note: This piece of news got overshadowed pretty quickly; the open-sourced, LAION-powered Stable Diffusion was release mid-August. The WikiArt dataset (see mention below) has 40,000 images, whereas the LAION has somewhere between 400 million and 5 billion (sorry I just can't tell anymore because I see both numbers in a bunch of different places). 

A model to generate artistic images based on text descriptions
Jun 2022, phys.org

Dynamic memory generative adversarial network (DM-GAN) can automatically generate unique artistic images based on text descriptions.

"Due to the lack of datasets with paired text description and artistic images, it is hard to directly train an algorithm which can create art based on text input," the researchers explained in their paper.

"To address this issue, we split our task into three steps."

First use the DM-GAN model to generate a realistic image that represents a text description.
Subsequently use neural network ResNet to classify the image produced by the DM-GAN into one of the genre categories outlined by the WikiArt dataset, which contains more than 40,000 artistic paintings produced by 195 artists. Then use deep style transfer, and done. 

via University of Waterloo and New York University Courant Institute: Qinghe Tian, Jean-Claude Franchitti, Text to artistic image generation. arXiv:2205.02439v1 [cs.CV],

Image credit: Tian & Franchitti

Post Script:
Your brain is better at busting deepfakes than you are
Jul 2022, phys.org

When looking at participants' brain activity, the University of Sydney researchers found deepfakes could be identified 54% of the time. However, when participants were asked to verbally identify the deepfakes, they could only do this 37% of the time.

"That tells us the brain can spot the difference between deepfakes and authentic images."

(Interpretation: In the future you will need an EEG just to communicate with your own brain.)

via University of Sydney: Michoel L. Moshel et al, Are you for real? Decoding realistic AI-generated faces from neural activity, Vision Research (2022). DOI: 10.1016/j.visres.2022.108079

Posted by at No comments:
Labels: , , ,

Psyflops


Delaying gratification: How do children react to waiting in different cultures?
Jul 2022, phys.org

The Marshmallow Test Test:
In the context of eating food, Japanese people are accustomed to waiting. When having meals, Japanese people typically wait until all individuals are served. Such customs of waiting to eat food are not as prevalent in the daily experiences of children in the United States.

In the context of opening gifts, U.S. children may experience waiting more consistently than Japanese children. Giving gifts is a more special event occurring on specific occasions in the United States, such as birthdays and other holidays, that can involve traditions of waiting. In contrast, gift giving is a regular year-round event for Japanese people that is not consistently associated with traditions of waiting.

via UC Davis: Kaichi Yanaoka et al, Cultures Crossing: The Power of Habit in Delaying Gratification, Psychological Science (2022). DOI: 10.1177/09567976221074650

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