This offers a very simple way to understand bias in neural networks. First, an explanation from the press article: In recent years, researchers have found that AI apps used to approve mortgage and other loan applications are biased, for example, in favor of white males. This, researchers found, was because the dataset used to train the system mostly comprised white male profiles. In this new effort, the researchers wondered if the same might be true for AI systems used to assist doctors in diagnosing patients.
And their findings: In looking at their results, the researchers found that there was a definite bias—when the data was mostly male, the error rates for processing female profiles rose. The same was true if the ratios were reversed. They also found that over-representing one gender or the other did not confer an advantage—the error rates remained relatively stable.Post Script
It's precisely its 'spontaneous' quality, its transparency, its 'naturalness', its refusal to be made to examine the premises on which it is founded, its resistance to change or to correction, its effect of instant recognition, and the closed circle in which it moves which makes common sense, at one and the same time, 'spontaneous', ideological and unconscious. You cannot learn, through common sense, how things are: you can only discover where they fit into the existing scheme of things. In this way, its very taken-for-grantedness is what establishess it as a medium in which its own premises and presuppositions are being rendered invisible by its apparent transparency.
-Dick Hebdige quoting Stuart Hall in Subculture: The Meaning of Style, 1979 (p11)
Post Script:
Ideology has very little to do with 'consciousness'. It is profoundly unconscious...Ideology is indeed a system of representation, but in the majority of cases these representations have nothing to do with 'consciousness': they are usually images and occasionally concepts, but it is above all as structures-that they impose on the vast majority of men, not via their 'consciousness'. They are perceived-accepted-suffered cultural objects and they act functionally on men via a process that escapes them.
-Abridged, from Dick Hebdige quoting Louis Althusser in Subculture: The Meaning of Style, 1979 (p12)
Those concerns led OpenAI in February 2019 to take the unusual step of declining to release the early version, GPT-2, citing fears potential misuse could be dangerous.
GPT-3, about 100 times more powerful than GPT-2, has performed admirably in tests, according to the OpenAI report. It tackled reading compression exercises requiring filling in word blanks, tackling "on-the-fly reasoning," and generating compositions up to 500 words.
The closer participants felt to someone, the more similarly their brain represented them throughout the social brain.I mean if this doesn't say something about identity -- You are a collective of those closest to you, not a person so much as a multi-person chimera.
The popular fascination for sustainably glowing foliage is being realized through advances in designer genetics.
The scientists revealed that bioluminescence found in some mushrooms is metabolically similar to the natural processes common among plants. By inserting DNA obtained from the mushroom, the scientists were able to create plants that glow much brighter than previously possible.
When an egg cell of almost any sexually reproducing species is fertilized, it sets off a series of waves that ripple across the egg's surface. These waves are produced by billions of activated proteins that surge through the egg's membrane like streams of tiny burrowing sentinels, signaling the egg to start dividing, folding, and dividing again, to form the first cellular seeds of an organism.
When experimenting on starfish eggs, They observed that each wave emerged in a spiral pattern, and that multiple spirals whirled across an egg's surface at a time. Some spirals spontaneously appeared and swirled away in opposite directions, while others collided head-on and immediately disappeared.
The behavior of these swirling waves, the researchers realized, is similar to the waves generated in other, seemingly unrelated systems, such as the vortices in quantum fluids, the circulations in the atmosphere and oceans, and the electrical signals that propagate through the heart and brain.
It's a manifestation of this very universal wave pattern.
The researchers are particularly interested in the waves' similarity to ideas in quantum computing. Just as the pattern of waves in an egg convey specific signals, in this case of cell division, quantum computing is a field that aims to manipulate atoms in a fluid, in precise patterns, in order to translate information and perform calculations.
"Perhaps now we can borrow ideas from quantum fluids, to build minicomputers from biological cells," Fakhri says. "We expect some differences, but we will try to explore [biological signaling waves] further as a tool for computation."
Topological turbulence in the membrane of a living cell, Nature Physics (2020). DOI: 10.1038/s41567-020-0841-9. https://nature.com/articles/s41567-020-0841-9
Stratospheric aerosol geoengineering is the idea that adding a layer of aerosol particles to the upper atmosphere can reduce climate changes caused by greenhouse gases such as carbon dioxide.
Previous research shows that solar geoengineering could be achieved using commercially available aircraft technologies to deliver the particles at a cost of a few billion dollars per year and would reduce global average temperatures. However, the question remains whether this approach could reduce important climate hazards at a regional level.
Results from a new study by UCL and Harvard researchers suggest that even a crude method like injecting sulphur dioxide in the stratosphere could reduce many important climate hazards without making any region obviously worse off.
"However, if instead only half the warming is offset, then we find that stratospheric aerosol geoengineering could still reduce climate change overall but would only exacerbate change over 1.3% of the land area."Notes:
The presence of humans in a room or in other indoor environments can alter the propagation of RF signals in several ways. By pre-processing RF channel measurements, the researchers were able to create 'images' summarizing the signals, which could in turn be analyzed to detect the presence of humans in a given environment.
They then trained a CNN on a large amount of data containing both magnitude and phase information, two key properties of RF signals. Over time, the deep learning algorithm learned to distinguish when an environment is populated by humans and when it is free from them by analyzing what is known as channel state information (CSI).
"Exploiting the ubiquity of ambient RF signals such as WiFi, Bluetooth or cellular signals for situational awareness information provides added value to existing RF infrastructure," Chen said. "Occupancy detection, for example, is an application where RF sensing can be a low-cost and infrastructure-free alternative or complement to existing approaches."
The drug works in a different way to existing antibacterials and is the first of its kind to be found by setting AI loose on vast digital libraries of pharmaceutical compounds.
“In terms of antibiotic discovery, this is absolutely a first,” said Regina Barzilay, a senior researcher on the project and specialist in machine learning at Massachusetts Institute of Technology (MIT).
“I think this is one of the more powerful antibiotics that has been discovered to date,” added James Collins, a bioengineer on the team at MIT. “It has remarkable activity against a broad range of antibiotic-resistant pathogens.”
To find new antibiotics, the researchers first trained a “deep learning” algorithm to identify the sorts of molecules that kill bacteria. To do this, they fed the program information on the atomic and molecular features of nearly 2,500 drugs and natural compounds, and how well or not the substance blocked the growth of the bug E coli.
Once the algorithm had learned what molecular features made for good antibiotics, the scientists set it working on a library of more than 6,000 compounds under investigation for treating various human diseases. Rather than looking for any potential antimicrobials, the algorithm focused on compounds that looked effective but unlike existing antibiotics. This boosted the chances that the drugs would work in radical new ways that bugs had yet to develop resistance to.
Jonathan Stokes, the first author of the study, said it took a matter of hours for the algorithm to assess the compounds and come up with some promising antibiotics. One, which the researchers named “halicin” after Hal, the astronaut-bothering AI in the film 2001: A Space Odyssey, looked particularly potent.
"Immune modulation isn't currently part of the strategy with any of these severe infections," said Dr. Manish Butte, the report's senior author, who holds the E. Richard Stiehm Endowed Chair in Pediatric Allergy, Immunology and Rheumatology at the David Geffen School of Medicine at UCLA. "Our case suggests that rather than hoping to get the upper hand with more and more antibiotics or antifungals, we can have some success by combining these established approaches with the new idea of programming the patient's immune response to better fight the infection."Synthetic red blood cells mimic natural ones, and have new abilities
Taking a cue from the self-cleaning properties of the lotus leaf, researchers at Ben-Gurion University of the Negev have shed new light on microscopic forces and mechanisms that can be optimized to remove dust from solar panels to maintain efficiency and light absorption. The new technique removed 98 percent of dust particles.-image source: I really hate it when websites don't credit their artwork (almost as much as I hate the threeway image clusterfu**ing olympics performed by Google-Getty-Pinterest triangle). This image came from the site at this link, but they list no artist credits.
In a new study published in Langmuir, the researchers confirmed that modifying the surface properties of solar panels may greatly reduce the amount of dust remaining on the surface, and significantly increase the potential of solar energy harvesting applications in the desert.
Particle removal increased from 41 percent on hydrophilic smooth Si wafers to 98 percent on superhydrophobic Si-based nanotextured surfaces.
"We determined that the reason for the increased particle removal is not low friction between the droplets and the superhydrophobic surfaces," Heckenthaler says. "Rather, it is the increase in the forces that can detach particles from the surfaces. The experimental methods we used and the criterion for particle removal we derived can be implemented to engineer self-cleaning surfaces exhibiting different chemistries and/or textures."
 |
| How is eigencat not a thing? Try that search for yourself, "youtube eigencat" |
As proof of principle, they demonstrated an inverse-designed, inhomogeneous medium to perform English vowel classification based on raw audio signals as their waveforms scattered and propagated through it.You read that correctly. There are no circuits, no electricity. The physical system itself performs a recurrent processing dynamic, based on the way the waves move through space.
-source
The algorithm, called the Fast Loaded Dice Roller (FLDR), was created by MIT graduate students and research scientists.
FLDR can use up to 10,000 times less memory storage space than the Knuth-Yao approach, while taking no more than 1.5 times longer per operation.Growing crystals to generate random numbers
"Should tech companies pay more for dark-skinned subjects because they’re underrepresented in training data? If our bodies are commodities, what’s a fair price, and who should set it? The data-ownership idea is, fundamentally, limited: Even if we manage, with the help of Hughes or Yang or state legislatures, to negotiate a high price for our data, we’re still for sale.Smart home tech can help evict renters, surveillance company tells landlords
— Sidney Fussell, The Atlantic
_-_WGA05708.jpg){kind=link}