#Austin, #US; ArtificialIntelligence; #FutureResearch; #SelfInvolving
Austin (U.S.), Apr 14 (Canadian-Media): Artificial intelligence (AI) is evolving—literally. Researchers have created software that borrows concepts from Darwinian evolution, including “survival of the fittest,” to build AI programs that improve generation after generation without human input, University of Texas, Austin reported.
Artificial Intelligence. Image credit: Twitter
The program replicated decades of AI research in a matter of days, and its designers think that one day, it could discover new approaches to AI.
“While most people were taking baby steps, they took a giant leap into the unknown,” says Risto Miikkulainen, a computer scientist at the University of Texas, Austin, who was not involved with the work. “This is one of those papers that could launch a lot of future research.”
Building an AI algorithm takes time. Take neural networks, a common type of machine learning used for translating languages and driving cars. These networks loosely mimic the structure of the brain and learn from training data by altering the strength of connections between artificial neurons. Smaller subcircuits of neurons carry out specific tasks—for instance spotting road signs—and researchers can spend months working out how to connect them so they work together seamlessly.
In recent years, scientists have sped up the process by automating some steps. But these programs still rely on stitching together ready-made circuits designed by humans. That means the output is still limited by engineers’ imaginations and their existing biases.
So Quoc Le, a computer scientist at Google, and colleagues developed a program called AutoML-Zero that could develop AI programs with effectively zero human input, using only basic mathematical concepts a high school student would know. “Our ultimate goal is to actually develop novel machine learning concepts that even researchers could not find,” he says.
The program discovers algorithms using a loose approximation of evolution. It starts by creating a population of 100 candidate algorithms by randomly combining mathematical operations. It then tests them on a simple task, such as an image recognition problem where it has to decide whether a picture shows a cat or a truck.
In each cycle, the program compares the algorithms’ performance against hand-designed algorithms. Copies of the top performers are “mutated” by randomly replacing, editing, or deleting some of its code to create slight variations of the best algorithms. These “children” get added to the population, while older programs get culled. The cycle repeats.
The system creates thousands of these populations at once, which lets it churn through tens of thousands of algorithms a second until it finds a good solution. The program also uses tricks to speed up the search, like occasionally exchanging algorithms between populations to prevent any evolutionary dead ends, and automatically weeding out duplicate algorithms.
In a preprint paper published last month on arXiv, the researchers show the approach can stumble on a number of classic machine learning techniques, including neural networks. The solutions are simple compared with today’s most advanced algorithms, admits Le, but he says the work is a proof of principle and he’s optimistic it can be scaled up to create much more complex AIs.
Still, Joaquin Vanschoren, a computer scientist at the Eindhoven University of Technology, thinks it will be a while before the approach can compete with the state-of-the-art. One thing that could improve the program, he says, is not asking it to start from scratch, but instead seeding it with some of the tricks and techniques humans have discovered. “We can prime the pump with learned machine learning concepts.”
That’s something Le plans to work on. Focusing on smaller problems rather than entire algorithms also holds promise, he adds. His group published another paper on arXiv on 6 April that used a similar approach to redesign a popular ready-made component used in many neural networks.
But Le also believes boosting the number of mathematical operations in the library and dedicating even more computing resources to the program could let it discover entirely new AI capabilities. “That’s a direction we’re really passionate about,” he says. “To discover something really fundamental that will take a long time for humans to figure out.”
#Apple; #Google; #HarnessPhones; #VirusInfectionTracking; #Covid19Pandemic
New York, Apr 12 (Canadian-Media): Apple and Google fueled hopes for digital technology's promise against a fast-moving, invisible killer, announcing a joint effort to help public health agencies worldwide leverage smartphones to contain the COVID-19 pandemic, phys.org/news reports said.
This March 19, 2018 file photo shows a Google app in Baltimore. Apple and Google launched a major joint effort, Friday, April 10, 2020, to leverage smartphone technology contain the COVID-19 pandemic. New software the companies plan to add to phones would make it easier to use Bluetooth wireless technology to track down people who may have been infected by coronavirus carriers. Image credit: AP Photo/Patrick Semansky, File
New software the companies plan to add to phones would make it easier to use Bluetooth wireless technology to track down people who may have been infected by coronavirus carriers. The idea is to help national, state and local governments roll out apps for so-called "contact tracing" that will run on iPhones and Android phones alike.
The technology works by harnessing short-range Bluetooth signals. Using the Apple-Google technology, contact-tracing apps would gather a record of other phones with which they came into close proximity.
Such data can be used to alert others who might have been infected by known carriers of the novel coronavirus, typically when the phones' owners have installed the apps and agreed to share data with public-health authorities.
Developers have already created such apps in countries including Singapore and China to try to contain the pandemic. In Europe, the Czech Republic says it will release an app after Easter. Britain, Germany and Italy are also developing their own tracing tools.
No such apps have yet been announced in the United States, but Gov. Gavin Newsom of California said Friday that state officials have been in touch with the companies as they look ahead at how to reopen and lift stay-at-home orders.
"We were on the phone just this morning, for example, with Apple," he said at a news conference.
Privacy and civil liberties activists have warned that the apps need to be designed so governments cannot abuse them to track their citizens. Apple and Google said in a rare joint announcement that user privacy and security are baked into the design of their plan.
The technology might serve as a stopgap until there is widespread testing for the novel coronavirus, which in the U.S. remains limited after production problems and limited federal coordination of the tests' production and distribution.
"It's not a replacement for just having widespread testing, which would be more accurate," said Tiffany Li, a visiting law professor at Boston University who studies privacy and technology. "But clearly we have a huge shortage of tests."
Bluetooth signal tracking, as Google and Apple plan to use it, can protect privacy far better than other options such as GPS or cell-tower based location data, which allow centralized authorities access to the information.
In this Dec. 26, 2018, file photo, an Apple logo is seen in raindrops on a window outside an Apple Store at the Country Club Plaza shopping district in Kansas City, Mo. Apple and Google launched a major joint effort, Friday, April 10, 2020, to leverage smartphone technology contain the COVID-19 pandemic. New software the companies plan to add to phones would make it easier to use Bluetooth wireless technology to track down people who may have been infected by coronavirus carriers. Image credit: AP Photo/Charlie Riedel, File
But Li noted it could still lead to numerous mistaken alerts—for instance, if someone were in full protective gear or in an adjacent apartment while physically close to an infected person.
Pam Dixon, executive director of the World Privacy Forum, said a conversation with Apple's senior director for global privacy, Jane Horvath, assured her that the initiative will protect people's privacy.
Sensitive information will stay on individual phones in encrypted form—no personally identifiable data would be collected—and alerts would be handled by public health agencies, not the tech companies, according to briefing paper seen by The Associated Press. It says location data for users won't be used and the identity of people who may have been infected will be protected by encryption and anonymous identifier beacons that change frequently.
"I think they've taken care of some of the really big problems," Dixon said, noting the companies say they can turn off the system when it's no longer needed. "The government is not going to have identity information of those testing positive."
Asked about the Google-Apple effort at his daily news briefing, President Donald Trump called it "very interesting," but expressed concern that "a lot of people worry about it in terms of a person's freedom. We're going to take a look at that."
Security experts note that technology alone cannot effectively track down and identify people who may have been infected by COVID-19 carriers. Such efforts will require other tools and teams of public health care workers to locate people in the physical world, they say. In South Korea and China, such efforts have included the use of credit-card and public-transit records.
In general, epidemiologists say contact tracing won't be effective without widely available testing. In the Czech Republic, the plan is to have soldiers perform testing; medical students have been trained to staff call centers for notifying people at high risk of infection.
The Czech app will use both Bluetooth technology and geolocation data from wireless carriers and banks to create "memory maps" that trace the movement of infected people. That will help them identify others they came into close proximity with in the five to 10 days before they tested positive.
The hope is to quickly isolate people who may be affected so the virus can be contained and restrictions on movement relaxed. The app builds on a popular cellular-location mapping app used by one in 10 Czechs, who number 10 million.
The Google-Apple solution will also be voluntary—or opt-in—but with far greater privacy protections, something the European Commission specified as a central requirement of any such apps in a policy recommendation this week for the 27-nation bloc.
Given the great need for effective contact-tracing—a tool epidemiologists have long employed to contain infectious disease outbreaks—Google and Apple will roll out their changes in two phases. In May, they will release software that will support public-health apps for both Android and iOS phones. In coming months, they will also build the functionality directly into the underlying phone operating systems.
On Friday, the companies released preliminary technical specifications for the effort, which they called "Privacy-Preserving Contact Tracing."
#Zoom; #videoConferencingTool; #remoteWork&Study; #FBI; #MaliciousAttacks; #Investigation
Ottawa, Apr 3 (Canadian-Media): Millions of people being forced to stay home due to the COVID-19 pandemic outbreak, Zoom, with its video conferencing tool for remote work and study, has emerged as an indispensable tool all across the world, media reports said.
Zoom meeting. Image credit: Twitter
Zoom Video Communications is an American remote conferencing services company, founded by Eric Yuan and headquartered in San Jose, California.
Zoom's remote conferencing service combines video conferencing, online meetings, chat, and mobile collaborations.
Eric Yuan. Image credit: Twitter
But warnings have been issued by Federal Bureau of Investigation (FBI) due to a growing number of so-called Zoom-bombing incidents.
Warnings had also been issued from the victims themselves.
Zoom-bombing is the term used when people participating in meetings and lessons via video conference platforms like Zoom can find their screens hijacked by malicious actors who can put words, images, racial and sexist slurs, on the screen and in the chat box that creates havoc with the audio.
Increase in frequency of Zoom-bombing in Canada, which experts termed as "video teleconferencing (VTC) hijacking," prompted a warning from the FBI earlier this week.
The agency also released a tip sheet advising participants to keep VTC meetings private by issuing users a password or employing the "waiting room" function, which requires the host to invite each guest individually; not sharing invitation links on social media; keep software updated to stay on top of any security features provided by VTC companies.
In the meantime Zoom's shares have doubled in price since the COVID-19 crisis erupted in January and has experienced record downloads.
Having been forced to apologize for not being forthcoming about its security limitations, Zoom says it's providing safety guidance for virtual classrooms and meetings.
But users have not been provided with additional controls to prevent harassment and online attacks.
As these malicious attacks were mainly targeted to marginalized groups, there were suggestions that these be investigated as hate speech.
The FBI said these types of security invasions can have a lasting effect on people personally, but children in particular, can have a tough time understanding what happened and why.
Tech & Innovation