At the start of the pandemic, remembers MIT Technology Review's senior editor for AI, the community "rushed to develop software that many believed would allow hospitals to diagnose or triage patients faster, bringing much-needed support to the front lines — in theory. "In the end, many hundreds of predictive tools were developed. None of them made a real difference, and some were potentially harmful." That's the damning conclusion of multiple studies published in the last few months. In June, the Turing Institute, the UK's national center for data science and AI, put out a report summing up discussions at a series of workshops it held in late 2020. The clear consensus was that AI tools had made little, if any, impact in the fight against covid. This echoes the results of two major studies that assessed hundreds of predictive tools developed last year. Laure Wynants, an epidemiologist at Maastricht University in the Netherlands who studies predictive tools, is lead author of one of them, a review in the British Medical Journal that is still being updated as new tools are released and existing ones tested. She and her colleagues have looked at 232 algorithms for diagnosing patients or predicting how sick those with the disease might get. They found that none of them were fit for clinical use. Just two have been singled out as being promising enough for future testing. "It's shocking," says Wynants. "I went into it with some worries, but this exceeded my fears." Wynants's study is backed up by another large review carried out by Derek Driggs, a machine-learning researcher at the University of Cambridge, and his colleagues, and published in Nature Machine Intelligence. This team zoomed in on deep-learning models for diagnosing covid and predicting patient risk from medical images, such as chest x-rays and chest computer tomography (CT) scans. They looked at 415 published tools and, like Wynants and her colleagues, concluded that none were fit for clinical use. "This pandemic was a big test for AI and medicine," says Driggs, who is himself working on a machine-learning tool to help doctors during the pandemic. "It would have gone a long way to getting the public on our side," he says. "But I don't think we passed that test...." If there's an upside, it is that the pandemic has made it clear to many researchers that the way AI tools are built needs to change. "The pandemic has put problems in the spotlight that we've been dragging along for some time," says Wynants. The article suggests researchers collaborate on creating high-quality (and shared) data sets — possibly by creating a common data standard — and also disclose their ultimate models and training protocols for review and extension. "In a sense, this is an old problem with research. Academic researchers have few career incentives to share work or validate existing results. "To address this issue, the World Health Organization is considering an emergency data-sharing contract that would kick in during international health crises."

Read more of this story at Slashdot.

情報元へのリンク

Slashdot reader fahrbot-bot quotes the New York Times: Two red things are hiding in a part of the solar system where they shouldn't be. The space rocks may have come from beyond Neptune, and potentially offer hints at the chaos of the early solar system. Scientists led by Sunao Hasegawa from JAXA, the Japanese space agency, reported in The Astrophysical Journal Letters on July 26, 2021 that two objects, called 203 Pompeja and 269 Justitia, spotted in the asteroid belt between Mars and Jupiter appear to have originated beyond Neptune. The discoveries could one day provide direct evidence of the chaos that existed in the early solar system. "If true it would be a huge deal," says Hal Levison, a planetary scientist at the Southwest Research Institute in Colorado, who was not involved in the research... "People have been talking about some fraction of asteroids coming from the Kuiper belt for quite a while now," said Josh Emery, a planetary scientist from Northern Arizona University who was not involved in the paper. He said the research "definitely takes a step" toward finding evidence to support that hypothesis. Not everyone is convinced just yet. Dr. Levison, who was also not involved in the paper, says objects should become less red as they approach the sun. "It seems to be inconsistent with our models," said Dr. Levison, who is the head of NASA's Lucy mission, which is scheduled to launch in October to study Jupiter's Trojans [asteroids captured in its orbit]. Michaël Marsset from the Massachusetts Institute of Technology, a co-author on the paper, agrees that it's not clear why they would be so red, but it is possibly related to how long it took them to become implanted into the asteroid belt. Some Trojans may also be as red, but haven't been found yet. To truly confirm the origin of 203 Pompeja and 269 Justitia, a spacecraft would likely need to visit them.

Read more of this story at Slashdot.

情報元へのリンク

"It was pretty strange when Russia decided to announce last week that it had successfully run tests between June 15 and July 15 to show it could disconnect itself from the internet," writes an associate professor of cybersecurity policy at Tufts Fletcher School of Law and Diplomacy. The tests seem to have gone largely unnoticed both in and outside of Russia, indicating that whatever entailed did not involve Russia actually disconnecting from the global internet... since that would be impossible to hide. Instead, the tests — and, most of all, the announcement about their success — seem to be intended as some kind of signal that Russia is no longer dependent on the rest of the world for its internet access. But it's not at all clear what that would even mean since Russia is clearly still dependent on people and companies in other countries for access to the online content and services they create and host — just as we all are... For the past two years, ever since implementing its "sovereign internet law" in 2019, Russia has been talking about establishing its own domestic internet that does not rely on any infrastructure or resources located outside the country. Presumably, the tests completed this summer are related to that goal of being able to operate a local internet within Russia that does not rely on the global Domain Name System to map websites to specific IP addresses. This is not actually a particularly ambitious goal — any country could operate its own domestic internet with its own local addressing system if it wanted to do so instead of connecting to the larger global internet... The Center for Applied Internet Data Analysis at the University of California San Diego maintains an Internet Outage Detection and Analysis tool that combines three data sets to identify internet outages around the world... The data sets for Russia from June 15 through July 15, the period of the supposed disconnection tests, shows few indications of any actual disconnection other than a period around July 5 when unsolicited traffic from Russia appears to have dropped off. Whatever Russia did this summer, it did not physically disconnect from the global internet. It doesn't even appear to have virtually disconnected from the global internet in any meaningful sense. Perhaps it shifted some of its critical infrastructure systems to rely more on domestic service providers and resources. Perhaps it created more local copies of the addressing system used to navigate the internet and tested its ability to rely on those. Perhaps it tested its ability to route online traffic within the country through certain chokepoints for purposes of better surveillance and monitoring. None of those are activities that would be immediately visible from outside the country and all of them would be in line with Russia's stated goals of relying less on internet infrastructure outside its borders and strengthening its ability to monitor online activity. But the goal of being completely independent of the rest of the world's internet infrastructure while still being able to access the global internet is a nonsensical and impossible one. Russia cannot both disconnect from the internet and still be able to use all of the online services and access all of the websites hosted and maintained by people in other parts of the world, as appears to have been the case during the monthlong period of testing... Being able to disconnect your country from the internet is not all that difficult — and certainly nothing to brag about. But announcing that you've successfully disconnected from the internet when it's patently clear that you haven't suggests both profound technical incompetence and a deep-seated uncertainty about what a domestic Russian internet would actually mean.

Read more of this story at Slashdot.

情報元へのリンク