Computers Math

The science community has responded to the COVID-19 pandemic with such a flurry of research studies that it is hard for anyone to digest them all, underscoring a long-standing need to make scientific publication more accessible, transparent and accountable, two artificial intelligence experts assert in a data science journal.
The rush to publish results has resulted in missteps, say Ganesh Mani, an investor, technology entrepreneur and adjunct faculty member in Carnegie Mellon University’s Institute for Software Research, and Tom Hope, a post-doctoral researcher at the Allen Institute for AI. In an opinion article in today’s issue of the journal Patterns, they argue that new policies and technologies are needed to ensure relevant, reliable information is properly recognized.
Those potential solutions include ways to combine human expertise with AI as one way to keep pace with a knowledge base that is expanding geometrically. AI might be used to summarize and collect research on a topic, while humans serve to curate the findings, for instance.
“Given the ever-increasing research volume, it will be hard for humans alone to keep pace,” they write.
In the case of COVID-19 and other new diseases, “you have a tendency to rush things because the clinicians are asking for guidance in treating their patients,” Mani said. Scientists certainly have responded — by mid-August, more than 8,000 preprints of scientific papers related to the novel coronavirus had been posted in online medical, biology and chemistry archives. Even more papers had been posted on such topics as quarantine-induced depression and the impact on climate change from decreased transportation emissions.
At the same time, the average time to perform peer review and publish new articles has shrunk; in the case of virology, the average dropped from 117 to 60 days.

advertisement

This surge of information is what the World Health Organization calls an “infodemic” — an overabundance of information, ranging from accurate to demonstrably false. Not surprisingly, problems such as the hydroxycholoroquine controversy have erupted as research has been rushed to publication and subsequently withdrawn.
“We’re going to have that same conversation with vaccines,” Mani predicted. “We’re going to have a lot of debates.”
Problems in scientific publication are nothing new, he said. As a grad student 30 years ago, he proposed an electronic archive for scientific literature that would better organize research and make it easier to find relevant information. Many ideas continue to circulate about how to improve scientific review and publication, but COVID-19 has exacerbated the situation.
Some of the speed bumps and guard rails that Mani and Hope propose are new policies. For instance, scientists usually emphasize experiments and therapies that work; highlighting negative results, on the other hand, is important for clinicians and discourages other scientists from going down the same blind alleys. Identifying the best reviewers, sharing review comments and linking papers to related papers, retraction sites or legal rulings are among other ideas they explore.
Greater use of AI to digest and consolidate research is a major focus. Previous attempts to use AI to do so have failed in part because of the often figurative and sometimes ambiguous language used by humans, Mani noted. It may be necessary to write two versions of research papers — one written in a way that draws the attention of people and another written in a boring, uniform style that is more understandable to machines.
Mani said he and Hope have no illusions that their paper will settle the debate about improving scientific literature, but hope that it will spur changes in time for the next global crisis.
“Putting such infrastructure in place will help society with the next strategic surprise or grand challenge, which is likely to be equally, if not more, knowledge intensive,” they concluded.

Story Source:
Materials provided by Carnegie Mellon University. Original written by Byron Spice. Note: Content may be edited for style and length. More

Netflix is often criticized as a Hollywood-style entertainment behemoth crushing all competition and diminishing local content, but an academic says that’s a simplistic view. A media studies expert said there is a lot of misunderstanding about the world’s biggest internet-distributed video service which has proved a game-changer for entertainment. Media studies expert Professor Amanda Lotz, from QUT’s Digital Media Research Centre, said there is a lot of misunderstanding about the world’s biggest internet-distributed video service.
“Netflix must be examined as a zebra among horses,” said Professor Lotz who is in the middle of a three-year Australian Research Council Discovery Project — Internet-distributed television: Cultural, industrial and policy dynamics. She recently published an article in the International Journal of Cultural Studies — ‘In Between the Global and the Local: Mapping the Geographies of Netflix as a Multinational Service.’
“Few recognize the extent to which Netflix has metamorphosed into a global television service. Unlike services that distribute only US-produced content, Netflix has funded the development of a growing library of series produced in more than 27 countries, across six continents, including Australia.
“Netflix has regional offices now in Singapore, Amsterdam, and São Paulo. Last year it opened its Australian headquarters in Sydney.”
Along with QUT’s Distinguished Professor Stuart Cunningham and Dr Ramon Lobato, Senior Research Fellow, RMIT, Professor Lotz is investigating the impact of global subscription video-on-demand platforms on national television markets.
“Internet-distributed video services such as Netflix, have completely transformed the entertainment landscape and the competitive field in which free-to-air television operates, as well as turned the definition of ‘pay TV’ on its head,” Professor Lotz said.

advertisement

“But the Netflix model has been the real gamechanger. Previously, the core business of channels like the BBC, ABC or NBC that commission and pay the lion’s share of production fees for series has been nation bound, even if those shows would someday be available to audiences in many countries.
“Netflix’s propensity to commission series in multiple countries, and then make them available to the full 150-some million subscribers simultaneously, is unprecedented and something no television channel could do.
“A local example of this is Hannah Gadsby: Nanettewhich has given the Australian comedian a new global profile. She now has a second Netflix show — Hannah Gadsby: Douglas.
“And although many believe Netflix competes with the likes of Amazon Prime Video, Apple TV+, Stan and Disney+, none of these services show evidence of supporting multinational production at a scale comparable to Netflix.
“Our research project has compiled a database of series commissioned by Netflix (in whole or part) and their country of origin. We have found more than half of the titles are produced outside the US and initial analysis of Netflix original films suggests a similar pattern.”
However, Professor Lotz said Netflix could never develop the depth of content necessary to replace national providers, especially public service broadcasters central to cultural storytelling.

advertisement

“It is difficult to appreciate whether some of Netflix’s peculiarity results from its global reach, business model, or distribution technology, but these are crucial questions to ask. And do these characteristics lead to the availability of stories, characters, and places not readily available? If so, this is a notable benefit to audiences,” she said.
“We should also ask how these characteristics affect opportunities available for writers, producers, and actors who might be rethinking the kind of stories that must be told to sell internationally.
“Appealing to audiences outside a commissioning channel’s country is increasingly necessary. Even if Netflix is unlikely to eliminate national providers, it is reconfiguring the competitive landscape.”
Professor Lotz also posted a blog series, Netflix 30 Q&A, in recent months that examines the differences of the SVOD business and how it allows Netflix different program strategies than linear, ad-supported channels.
“The long term and global rights the company seeks in its commissions have required significant changes in the remuneration norms for those who make its series, and it remains unclear whether the new norms amount to lower pay,” she said.
“National broadcasters worry about keeping up with the escalating fees Netflix can support for its prestige series’ and complain of an unfair playing field where Netflix isn’t subject to the same local content rules and other requirements.
“But business and cultural analysts must stop trying to shoehorn Netflix into the same category as linear channels and streaming services aimed at pushing US content abroad. Over its 23-year-existence, Netflix has evolved repeatedly. Perhaps this steady change fuels its misperception.” More

For quantum optical technologies to become more practical, there is a need for large-scale integration of quantum photonic circuits on chips.
This integration calls for scaling up key building blocks of these circuits — sources of particles of light — produced by single quantum optical emitters.
Purdue University engineers created a new machine learning-assisted method that could make quantum photonic circuit development more efficient by rapidly preselecting these solid-state quantum emitters.
The work is published in the journal Advanced Quantum Technologies.
Researchers around the world have been exploring different ways to fabricate identical quantum sources by “transplanting” nanostructures containing single quantum optical emitters into conventional photonic chips.
“With the growing interest in scalable realization and rapid prototyping of quantum devices that utilize large emitter arrays, high-speed, robust preselection of suitable emitters becomes necessary,” said Alexandra Boltasseva, Purdue’s Ron and Dotty Garvin Tonjes Professor of Electrical and Computer Engineering.

advertisement

Quantum emitters produce light with unique, non-classical properties that can be used in many quantum information protocols.
The challenge is that interfacing most solid-state quantum emitters with existing scalable photonic platforms requires complex integration techniques. Before integrating, engineers need to first identify bright emitters that produce single photons rapidly, on-demand and with a specific optical frequency.
Emitter preselection based on “single-photon purity” — which is the ability to produce only one photon at a time — typically takes several minutes for each emitter. Thousands of emitters may need to be analyzed before finding a high-quality candidate suitable for quantum chip integration.
To speed up screening based on single-photon purity, Purdue researchers trained a machine to recognize promising patterns in single-photon emission within a split second.
According to the researchers, rapidly finding the purest single-photon emitters within a set of thousands would be a key step toward practical and scalable assembly of large quantum photonic circuits.

advertisement

“Given a photon purity standard that emitters must meet, we have taught a machine to classify single-photon emitters as sufficiently or insufficiently ‘pure’ with 95% accuracy, based on minimal data acquired within only one second,” said Zhaxylyk Kudyshev, a Purdue postdoctoral researcher.
The researchers found that the conventional photon purity measurement method used for the same task took 100 times longer to reach the same level of accuracy.
“The machine learning approach is such a versatile and efficient technique because it is capable of extracting the information from the dataset that the fitting procedure usually ignores,” Boltasseva said.
The researchers believe that their approach has the potential to dramatically advance most quantum optical measurements that can be formulated as binary or multiclass classification problems.
“Our technique could, for example, speed up super-resolution microscopy methods built on higher-order correlation measurements that are currently limited by long image acquisition times,” Kudyshev said.

Story Source:
Materials provided by Purdue University. Note: Content may be edited for style and length. More

The search is on to discover new states of matter, and possibly new ways of encoding, manipulating, and transporting information. One goal is to harness materials’ quantum properties for communications that go beyond what’s possible with conventional electronics. Topological insulators — materials that act mostly as insulators but carry electric current across their surface — provide some tantalizing possibilities.
“Exploring the complexity of topological materials — along with other intriguing emergent phenomena such as magnetism and superconductivity — is one of the most exciting and challenging areas of focus for the materials science community at the U.S. Department of Energy’s Brookhaven National Laboratory,” said Peter Johnson, a senior physicist in the Condensed Matter Physics & Materials Science Division at Brookhaven. “We’re trying to understand these topological insulators because they have lots of potential applications, particularly in quantum information science, an important new area for the division.”
For example, materials with this split insulator/conductor personality exhibit a separation in the energy signatures of their surface electrons with opposite “spin.” This quantum property could potentially be harnessed in “spintronic” devices for encoding and transporting information. Going one step further, coupling these electrons with magnetism can lead to novel and exciting phenomena.
“When you have magnetism near the surface you can have these other exotic states of matter that arise from the coupling of the topological insulator with the magnetism,” said Dan Nevola, a postdoctoral fellow working with Johnson. “If we can find topological insulators with their own intrinsic magnetism, we should be able to efficiently transport electrons of a particular spin in a particular direction.”
In a new study just published and highlighted as an Editor’s Suggestion in Physical Review Letters, Nevola, Johnson, and their coauthors describe the quirky behavior of one such magnetic topological insulator. The paper includes experimental evidence that intrinsic magnetism in the bulk of manganese bismuth telluride (MnBi2Te4) also extends to the electrons on its electrically conductive surface. Previous studies had been inconclusive as to whether or not the surface magnetism existed.
But when the physicists measured the surface electrons’ sensitivity to magnetism, only one of two observed electronic states behaved as expected. Another surface state, which was expected to have a larger response, acted as if the magnetism wasn’t there.

advertisement

“Is the magnetism different at the surface? Or is there something exotic that we just don’t understand?” Nevola said.
Johnson leans toward the exotic physics explanation: “Dan did this very careful experiment, which enabled him to look at the activity in the surface region and identify two different electronic states on that surface, one that might exist on any metallic surface and one that reflected the topological properties of the material,” he said. “The former was sensitive to the magnetism, which proves that the magnetism does indeed exist in the surface. However, the other one that we expected to be more sensitive had no sensitivity at all. So, there must be some exotic physics going on!”
The measurements
The scientists studied the material using various types of photoemission spectroscopy, where light from an ultraviolet laser pulse knocks electrons loose from the surface of the material and into a detector for measurement.
“For one of our experiments, we use an additional infrared laser pulse to give the sample a little kick to move some of the electrons around prior to doing the measurement,” Nevola explained. “It takes some of the electrons and kicks them [up in energy] to become conducting electrons. Then, in very, very short timescales — picoseconds — you do the measurement to look at how the electronic states have changed in response.”
The map of the energy levels of the excited electrons shows two distinct surface bands that each display separate branches, electrons in each branch having opposite spin. Both bands, each representing one of the two electronic states, were expected to respond to the presence of magnetism.

advertisement

To test whether these surface electrons were indeed sensitive to magnetism, the scientists cooled the sample to 25 Kelvin, allowing its intrinsic magnetism to emerge. However only in the non-topological electronic state did they observe a “gap” opening up in the anticipated part of the spectrum.
“Within such gaps, electrons are prohibited from existing, and thus their disappearance from that part of the spectrum represents the signature of the gap,” Nevola said.
The observation of a gap appearing in the regular surface state was definitive evidence of magnetic sensitivity — and evidence that the magnetism intrinsic in the bulk of this particular material extends to its surface electrons.
However, the “topological” electronic state the scientists studied showed no such sensitivity to magnetism — no gap.
“That throws in a bit of a question mark,” Johnson said.
“These are properties we’d like to be able to understand and engineer, much like we engineer the properties of semiconductors for a variety of technologies,” Johnson continued.
In spintronics, for example, the idea is to use different spin states to encode information in the way positive and negative electric charges are presently used in semiconductor devices to encode the “bits” — 1s and 0s — of computer code. But spin-coded quantum bits, or qubits, have many more possible states — not just two. This will greatly expand on the potential to encode information in new and powerful ways.
“Everything about magnetic topological insulators looks like they’re right for this kind of technological application, but this particular material doesn’t quite obey the rules,” Johnson said.
So now, as the team continues their search for new states of matter and further insights into the quantum world, there’s a new urgency to explain this particular material’s quirky quantum behavior. More

Geneticists could identify the causes of disorders that currently go undiagnosed if standard practices for collecting individual genetic information were expanded to capture more variants that researchers can now decipher, concludes new Johns Hopkins University research.
The laboratory of Johns Hopkins biomedical engineering professor Alexis Battle has developed a technique to begin identifying potentially problematic rare genetic variants that exist in the genomes of all people, particularly if additional genetic sequencing information was included in standard collection methods. The team’s findings are published in the latest issue of Science and are part of the Genotype-Tissue Expression (GTEx) Program funded by the National Institutes of Health.
“The implications of this could be quite large. Everyone has around 50,000 variants that are rare in the population and we have absolutely no idea what most of them are doing,” Battle said. “If you collect gene expression data, which shows which proteins are being produced in a patient’s cells at what levels, we’re going to be able to identify what’s going on at a much higher rate.”
While approximately 8% of U.S. citizens, mostly children, suffer from genetic disorders, the genetic cause has not been found for about half of the cases. What’s even more frustrating, according to Battle, is that even more people are likely living with more subtle genetically-influenced health ailments that have not been identified.
“We really don’t know how many people are out there walking around with a genetic aberration that is causing them health issues,” she said. “They go completely undiagnosed, meaning we cannot find the genetic cause of their problems.”
The field of personalized genomics is unable to characterize these rare variants because most genetic variants, specifically variants that are in “non-coding” parts of the genome that do not specify a protein, are not tested. Doing so would represent a major advance in a growing field that is focused on the sequencing and analysis of individuals’ genomes, she said
The Battle Lab developed a computational system called “Watershed” that can scour reams of genetic data along with gene expression to predict the functions of variants from individual’s genomes. They validated those predictions in the lab and applied the findings to assess the rare variants captured in massive gene collections such as the UK Biobank, the Million Veterans Program and the Jackson Heart Study. The results have helped to show which rare variants may be impacting human traits.
“Any improvement we can make in this area has implications for public health,” Battle said. “Even pointing to what the genetic cause is gives parents and patients a huge sense of relief and understanding and can point to potential therapeutics.”
Battle’s team worked in collaboration with researchers from Scripps Translational Science Institute, the New York Genome Center, the Massachusetts Institute of Technology and Stanford, Harvard and Columbia universities.
“Looking at the cross-tissue transcriptional footprint of rare genetic variants across many human tissues in GTEx data also helps us better understand the gaps and the potential of these analyses for clinical diagnostics,” said Pejman Mohammadi, a co-author and professor of integrative structural and computational biology at Scripps Research.
The grant numbers involved in the research include: R01MH109905, 1R01HG010480, Searle Scholar Program, R01HG008150.

Story Source:
Materials provided by Johns Hopkins University. Note: Content may be edited for style and length. More

Scientists at Tokyo Institute of Technology (Tokyo Tech) use commercially available quantum annealers, a type of quantum computer, to experimentally probe the validity of an important mechanism from nonequilibrium physics in open quantum systems. The results not only shed light into the extent of applicability of this mechanism and an extension of it, but also showcase how quantum annealers can serve as effective platforms for quantum simulations.
It is established that matter can transition between different phases when certain parameters, such as temperature, are changed. Although phase transitions are common (like water turning into ice in a freezer), the dynamics that govern these processes are highly complex and constitute a prominent problem in the field of nonequilibrium physics.
When a system undergoes a phase transition, matter in the new phase has many possible energetically equal “configurations” to adopt. In these cases, different parts of the system adopt different configurations over regions called “domains.” The interfaces between these domains are known as topological defects and reducing the number of these defects formed can be immensely valuable in many applications.
One common strategy to reduce defects is easing the system through the phase transition slowly. In fact, according to the “Kibble-Zurek” mechanism (KZM), it is predicted that the average number of defects and the driving time of the phase transition follow a universal power law. However, experimentally testing the KZM in a quantum system has remained a coveted goal.
In a recent study published in Physical Review Research, a team of scientists led by Professor Emeritus Hidetoshi Nishimori from Tokyo Institute of Technology, Japan, probed the validity of the KZM in two commercially available quantum annealers, a type of quantum computer designed for solving complex optimization problems. These devices, known as D-Wave annealers, can recreate controllable quantum systems and control their evolution over time, providing a suitable experimental testbed for the KZM.
First, the scientists checked whether the “power law” between the average number of defects and the annealing time (driving time of the phase transition) predicted by the KZM held for a quantum magnetic system called the “one-dimensional transverse-field Ising model.” This model represents the orientations (spins) of a long chain of “magnetic dipoles,” where homogenous regions are separated by defects seen as neighboring spins pointing in incorrect directions.
While the original prediction of the KZM regarding the average number of defects was valid in this system, the scientists took it a step further: although this extension of the KZM was originally intended for a completely “isolated” quantum system unaffected by external parameters, they found good agreement between its predictions and their experimental results even in the D-Wave annealers, which are “open” quantum systems.
Excited by these results, Prof Nishimori remarks: “Our work provides the first experimental test of universal critical dynamics in a many-body open quantum system. It also constitutes the first test of certain physics beyond the original KZM, providing strong experimental evidence that the generalized theory holds beyond the regime of validity theoretically established.”
This study showcases the potential of quantum annealers to perform simulations of quantum systems and also helps gain insight on other areas of physics. In this regard, Prof Nishimori states: “Our results leverage quantum annealing devices as platforms to test and explore the frontiers of nonequilibrium physics. We hope our work will motivate further research combining quantum annealing and other universal principles in nonequilibrium physics.” Hopefully, this study will also promote the use of quantum annealers in experimental physics. After all, who doesn’t love finding a new use for a tool?

Story Source:
Materials provided by Tokyo Institute of Technology. Note: Content may be edited for style and length. More

ACM, the Association for Computing Machinery, recently released its eighth annual Study of Non-Doctoral Granting Departments in Computing (NDC study). With the aim of providing a comprehensive look at computing education, the study includes information on enrollments, degree completions, faculty demographics, and faculty salaries. For the first time, this year’s ACM NDC study includes enrollment and degree completion data from the National Student Clearinghouse Research Center (NSC).
In previous years, ACM directly surveyed Computer Science departments, and would work with a sample of approximately 18,000 students. By accessing the NSC’s data, the ACM NDC study now includes information on approximately 300,000 students across the United States, allowing for a more reliable understanding of the state of enrollment and graduation in Bachelor’s programs. Also for the first time, the ACM NDC study includes data from private, for-profit institutions, which are playing an increasingly important role in computing education.
“By partnering with the NSC, we now have a much fuller picture of computing enrollment and degree production at the Bachelor’s level,” explained ACM NDC study co-author Stuart Zweben, Professor Emeritus, Ohio State University. “The NSC also gives us more specific data on the gender and ethnicity of students. This is an important tool, as increasing the participation of women and other underrepresented groups has been an important goal for leaders in academia and industry. For example, having a clear picture of the current landscape for underrepresented people is an essential first step toward developing approaches to increase diversity.”
“The computing community has come to rely on the ACM NDC study to understand trends in undergraduate computing education,” added ACM NDC study co-author Jodi Tims, Professor, Northeastern University. “At the same time, using our previous data collection methods, we were only capturing about 15-20% of institutions offering Bachelor’s degrees in computing. The NSC data gives us a much broader sample, as well as more precise information about enrollment and graduation in specific computing disciplines — such as computer science, information systems, information technology, software engineering, computer engineering and cybersecurity. For example, we’ve seen a noticeable increase in cybersecurity program offerings between the 2017/2018 and 2018/2019 academic years, and we believe this trend will continue next year. Going forward, we also plan to begin collecting information on data science offerings in undergraduate education. Our overall goal will be to maintain the ACM NDC study as the most up-to-date and authoritative resource on this topic.”
As with previous NDC studies, information on faculty salaries, retention, and demographics was collected by sending surveys to academic departments across the United States. Responses were received from 151 departments. The average number of full-time faculty members at the responding departments was 12.
Important findings of the ACM NDC study include:
-Between the 2017/2018 and the 2018/2019 academic years, there was a 4.7% increase in degree production across all computing disciplines. The greatest increases in degree production were in software engineering (9% increase) and computer science (7.5% increase)
-The representation of women in information systems (24.5% of degree earners in the 2018/2019 academic year) and information technology (21.5% of degree earners in the 2018/2019 academic year) is much higher than in areas such as computer engineering (12.2% of degree earners in the 2018/2019 academic year).
-Bachelor’s programs, as recorded by the ACM NDC study, had a stronger representation of African American and Hispanic students than PhD programs, as recorded by the Computer Research Association’s (CRA) Taulbee Survey. For example, during the 2018/2019 academic year, the ACM NDC records that 15.6% of enrollees in Bachelor’s programs were African American, whereas the CRA Taulbee survey records that 4.7% of enrollees in PhD programs were African American.
-In some disciplines of computing, African Americans and Hispanics are actually over-represented, based on their percentage of the US population.
-Based on aggregate salary data from 89 non-doctoral-granting computer science departments (including public and private institutions), the average median salary for a full professor was $109,424.
– Of 40 non-doctoral granting departments reporting over 56 faculty departures, only 10.7% of faculty departed for non-academic positions. Most departed due to retirement (46.4%) or other academic positions (26.9%).

In addition to Stuart Zweben, and Jodi Tims, the ACM NDC study was co-authored by Yan Timanovsky, Association for Computing Machinery. By employing the NSC data in future ACM NDC studies, the co-authors are confident that an even fuller picture will emerge regarding student retention with respect to computing disciplines, gender and ethnicity. More

Androids, or robots with humanlike features, are often more appealing to people than those that resemble machines — but only up to a certain point. Many people experience an uneasy feeling in response to robots that are nearly lifelike, and yet somehow not quite “right.” The feeling of affinity can plunge into one of repulsion as a robot’s human likeness increases, a zone known as “the uncanny valley.”
The journal Perception published new insights into the cognitive mechanisms underlying this phenomenon made by psychologists at Emory University.
Since the uncanny valley was first described, a common hypothesis developed to explain it. Known as the mind-perception theory, it proposes that when people see a robot with human-like features, they automatically add a mind to it. A growing sense that a machine appears to have a mind leads to the creepy feeling, according to this theory.
“We found that the opposite is true,” says Wang Shensheng, first author of the new study, who did the work as a graduate student at Emory and recently received his PhD in psychology. “It’s not the first step of attributing a mind to an android but the next step of ‘dehumanizing’ it by subtracting the idea of it having a mind that leads to the uncanny valley. Instead of just a one-shot process, it’s a dynamic one.”
The findings have implications for both the design of robots and for understanding how we perceive one another as humans.
“Robots are increasingly entering the social domain for everything from education to healthcare,” Wang says. “How we perceive them and relate to them is important both from the standpoint of engineers and psychologists.”
“At the core of this research is the question of what we perceive when we look at a face,” adds Philippe Rochat, Emory professor of psychology and senior author of the study. “It’s probably one of the most important questions in psychology. The ability to perceive the minds of others is the foundation of human relationships. ”

advertisement

The research may help in unraveling the mechanisms involved in mind-blindness — the inability to distinguish between humans and machines — such as in cases of extreme autism or some psychotic disorders, Rochat says.
Co-authors of the study include Yuk Fai Cheong and Daniel Dilks, both associate professors of psychology at Emory.
Anthropomorphizing, or projecting human qualities onto objects, is common. “We often see faces in a cloud for instance,” Wang says. “We also sometimes anthropomorphize machines that we’re trying to understand, like our cars or a computer.”
Naming one’s car or imagining that a cloud is an animated being, however, is not normally associated with an uncanny feeling, Wang notes. That led him to hypothesize that something other than just anthropomorphizing may occur when viewing an android.
To tease apart the potential roles of mind-perception and dehumanization in the uncanny valley phenomenon the researchers conducted experiments focused on the temporal dynamics of the process. Participants were shown three types of images — human faces, mechanical-looking robot faces and android faces that closely resembled humans — and asked to rate each for perceived animacy or “aliveness.” The exposure times of the images were systematically manipulated, within milliseconds, as the participants rated their animacy.
The results showed that perceived animacy decreased significantly as a function of exposure time for android faces but not for mechanical-looking robot or human faces. And in android faces, the perceived animacy drops at between 100 and 500 milliseconds of viewing time. That timing is consistent with previous research showing that people begin to distinguish between human and artificial faces around 400 milliseconds after stimulus onset.
A second set of experiments manipulated both the exposure time and the amount of detail in the images, ranging from a minimal sketch of the features to a fully blurred image. The results showed that removing details from the images of the android faces decreased the perceived animacy along with the perceived uncanniness.
“The whole process is complicated but it happens within the blink of an eye,” Wang says. “Our results suggest that at first sight we anthropomorphize an android, but within milliseconds we detect deviations and dehumanize it. And that drop in perceived animacy likely contributes to the uncanny feeling.”

Story Source:
Materials provided by Emory Health Sciences. Original written by Carol Clark. Note: Content may be edited for style and length. More