An anonymous reader shares an excerpt from an Ars Technica article: In November, people started polishing a Nobel prize for a group of physicists who seemed to have found new boson. [...] This result has been cooking for quite some time. The first experimental results date back to 2015, with publication in 2016. Essentially, the scientists took some lithium and shot protons at it. By choosing the energy of the protons correctly, Beryllium in a particular excited state is produced, which quickly decays back to lithium by emitting an electron and a positron. Now, in these experiments, energy and momentum must be conserved. The lithium nucleus is quite a complicated beast and can rattle around in all sorts of ways, meaning that the electron and positron have a certain amount of freedom in the direction in which they are emitted. By contrast, the researchers observed that some electrons and positrons seem to be correlated in their emission direction. Computer modeling confirmed that this was not due to their equipment and could not be explained by the nuclear physics of beryllium, lithium, or any known background process. The correlation could, however, be explained by a new boson that decayed by emitting a positron and an electron. As long as the production was reasonably inefficient, and the mass was about 17MeV (million electron volts), then the data was beautifully explained. It is always possible to extend our models of the Universe to include new particles, including new bosons and new forces. But, it isn't good enough to match a single experimental result. You have to match all of them. The end results are particles that look a bit like a backyard panel-beating job. Yeah, the paint matches, but you can still see the wavy patches where the filler hasn't been sanded flat. The problems arise from the mass -- 17MeV is at the low end of well-explored territory. So, why did this story flare back up again? A new paper, by the same scientists that published the beryllium results. This time, they measured electron-positron emissions from excited helium. Same experiment, different atom, but the same 17MeV boson was found. The new result is pretty strong evidence. "If the experiment has some kind of systematic error in it, then we would expect that the 'new' particle would change mass between helium and beryllium," adds Ars Technica. "It doesn't, though; the results are very consistent between experiments. That means that if it is an error, it is an unfortunately flukey one."

Read more of this story at Slashdot.

I was driving around this afternoon, and there were school buses everywhere. Why are the school buses out on a weekend I wondered. Drove past the school, huge numbers of buses and cars and people. Then it hit me. It's Friday! Oy.

Companies are trying to capitalize on the threat of space junk with new technology to clean it up, but it's not clear who will pay for the service. From a report: Today, thousands of pieces of space junk -- ranging from tiny fragments of destroyed satellites to spent rocket bodies and defunct spacecraft -- orbit around Earth, threatening operational satellites and astronauts. As thousands of new satellites are slated for launch in the coming years, operators are desperate to find ways to track, remove and prevent the creation of more rogue debris in orbit. The market for in-orbit satellite services is projected to reach about $4.5 billion by 2028, according to Northern Sky Research. [...] Experts agree space junk is a major threat to keeping space usable and open for nations and companies around the world, but it's not clear who is or should be responsible for cleaning it up, complicating the business case for these companies.

Read more of this story at Slashdot.

No NJFF this year unless there's an epic BitTorrent dump of the new releases (it has happened in past years). I live too far out of the city to get to openings. So that means one thing -- binges! Here's what I have queued up.

• Watchmen. Everyone says it's great, so I have to watch it.
• The Morning Show. I watched the first two episodes when it first came out and waited until the full series was out, and as of tonight, it is. It got terrible reviews, but I loved it.
• True Detective. I skipped the second season, but the third is good, they say.
• Mr Robot finale. On Sunday. The final episode of the final season. The last episode was the big reveal, as far as we know right now, Elliot and Darlene weren't really the stars of the show, White Rose was. But there's still another hour to go, and the writing and everything has been so superb, they got me.
• Handmaid's Tale. I tried watching the new season earlier, but I didn't remember enough from the previous season to make sense of it. I have it on my list to try again.
• The Expanse? I know so many people who love it, but I watched the first season and it never got me. But it looks like it might be a Battlestar Galactica type show, which I loved.
• Impeachment. Oh I hope there are leaks. McConnell deserves it. If the witnesses won't testify maybe they can share some of what they know anyway. Wouldn't it serve him right if he tried to control the flow of info, and it just flowed around him. I'd love to see his face then!

That's it for now. I'll keep you posted.

Kyle Wiggers, writing for VentureBeat: AI and machine learning algorithms are quite skilled at generating works of art -- and highly realistic images of apartments, people, and pets to boot. But relatively few have been tuned to singing synthesis, or the task of cloning musicians' voices. Researchers from Amazon and Cambridge put their collective minds to the challenge in a recent paper in which they propose an AI system that requires "considerably" less modeling than previous work of features like vibratos and note durations. It taps a Google-designed algorithm -- WaveNet -- to synthesize the mel-spectrograms, or representations of the power spectrum of sounds, which another model produces using a combination of speech and signing data. The system comprises three parts, the first of which is a frontend that takes a musical score as input and produces note embeddings (i.e., numerical representations of notes) to be sent to an encoder. The second is a model that is modified to accept the aforementioned embeddings, whose decoder produces mel-specrograms. As for the third and final component -- the WaveNet vocoder, which mimics things like stress and intonation in speech -- it synthesizes the spectrograms into song. The frontend performs linguistic analysis on the score lyrics, allowing for three possible vowel levels of stress and ignoring punctuation. In time, it discovers which phonemes (perceptually distinct units of sound) correspond to each note of the score using syllabification information specified in the score itself. It also computes the expected duration in seconds of each note, as well as the tempo and time signature of the score, which it combines into embeddings.

Read more of this story at Slashdot.

At its developer conference in June this year, Apple introduced Project Catalyst that aims to help developers swiftly bring their iOS apps to Macs. Developers have had more than half a year to play with Catalyst. Here's where things stand currently: The crux of the issue in my mind is that iOS and Mac OS are so fundamentally different that the whole notion of getting a cohesive experience through porting apps with minimal effort becomes absurd. The problem goes beyond touch vs pointer UX into how apps exist and interact within their wider OSes. While both Mac OS and iOS are easy to use, their ease stem from very different conventions. The more complicated Mac builds ease almost entirely through cohesion. Wherever possible, Mac applications are expected to share the same shortcuts, controls, windowing behavior, etc... so users can immediately find their bearings regardless of the application. This also means that several applications existing in the same space largely share the same visual and UX language. Having Finder, Safari, BBEdit and Transmit open on the same desktop looks and feels natural. By comparison, the bulk of iOS's simplicity stems from a single app paradigm. Tap an icon on the home screen to enter an app that takes over the entire user experience until exited. Cohesion exists and is still important, but its surface area is much smaller because most iOS users only ever see and use a single app at a time. For better and worse, the single app paradigm allows for more diverse conventions within apps. Having different conventions for doing the same thing across multiple full screen apps is not an issue because users only have to ever deal with one of those conventions at a given time. That innocuous diversity becomes incongruous once those same apps have to live side-by-side. Columnist John Gruber of DaringFireball adds: I think part of the problem is Catalyst itself -- it just doesn't feel like nearly a full-fledged framework for creating proper Mac apps yet. But I think another problem is the culture of doing a lot of nonstandard custom UI on iOS. As Wellborn points out, that flies on iOS -- we UI curmudgeons may not like it, but it flies -- because you're only ever using one app at a time on iOS. It cracks a bit with split-screen multitasking on iPadOS, but I've found that a lot of the iPad apps with the least-standard UIs don't even support split-screen multitasking on iPadOS, so the incongruities -- or incoherences, to borrow Wellborn's well-chosen word -- don't matter as much. But try moving these apps to the Mac and the nonstandard UIs stick out like a sore thumb, and whatever work the Catalyst frameworks do to support Mac conventions automatically doesn't kick in if the apps aren't even using the standard UIKit controls to start with. E.g. scrolling a view with Page Up, Page Down, Home, and End. Further reading: Apple's Merged iPad, Mac Apps Leave Developers Uneasy, Users Paying Twice (October 2019).

Read more of this story at Slashdot.

More than 267 million Facebook users allegedly had their user IDs, phone numbers and names exposed online, according to a report from Comparitech and security researcher Bob Diachenko. From a report: That info was found in a database that could be accessed without a password or any other authentication, and the researchers believe it was gathered as part of an illegal scraping operation or Facebook API abuse. Dianchenko says he reported the database to the service provider managing the IP address of the server, but the database was exposed for nearly two weeks. In the meantime, he says, the data was posted as a download in a hacker forum. That's a lot of personal data to be floating around in the wild, and as Comparitech notes, it could be used to carry out phishing scams and other foul play.

Read more of this story at Slashdot.

The unmanned debut test flight of the Starliner failed to reach its required altitude to dock with the International Space Station.

Raging bushfires in Australia reignite the debate over climate change, with the government accused of not doing enough.

How an overlooked feature of deep learning networks can turn into a major breakthrough for AI. From a report: Traditionally, deep learning algorithms such as deep neural networks (DNNs) are trained in a supervised fashion to recognize specific classes of things. In a typical task, a DNN might be trained to visually recognize a certain number of classes, say pictures of apples and bananas. Deep learning algorithms, when fed a good quantity and quality of data, are really good at coming up with precise, low error, confident classifications. The problem arises when a third, unknown object appears in front of the DNN. If an unknown object that was not present in the training set is introduced, such as an orange, then the network will be forced to "guess" and classify the orange as the closest class that captures the unknown object -- an apple! Basically, the world for a DNN trained on apples and bananas is completely made of apples and bananas. It can't conceive the whole fruit basket. While its usefulness is not immediately clear in all applications, the idea of "nothing" or a "class zero" is extremely useful in several ways when training and deploying a DNN. During the training process, if a DNN has the ability to classify items as "apple," "banana," or "nothing," the algorithm's developers can determine if it hasn't effectively learned to recognize a particular class. That said, if pictures of fruit continue to yield "nothing" responses, perhaps the developers need to add another "class" of fruit to identify, such as oranges. Meanwhile, in a deployment scenario, a DNN trained to recognize healthy apples and bananas can answer "nothing" if there is a deviation from the prototypical fruit it has learned to recognize. In this sense, the DNN may act as an anomaly detection network -- aside from classifying apples and bananas, it can also, without further changes, signal when it sees something that deviates from the norm. As of today, there are no easy ways to train a standard DNN so that it can provide the functionality above.

Read more of this story at Slashdot.

Social media giant says the tweets were amplifying pro-Saudi messages as well as discussions about sanctions on Iran.

Xi reportedly told Trump he is concerned about the US's 'negative rods and deeds' on issues like Hong Kong and Xinjiang.

Deadly protests jolt northern India amid widespread anger over legislation deemed discriminatory by critics.

Iranian president holds talks with Japanese PM Abe in Tokyo on future of crumbling landmark 2015 agreement.

The venture capital firm First Round Capital conducts an annual "State of Startups" survey that gets passed around widely in Silicon Valley. Its 2019 findings, published this week, are grim. From a report: Over two-thirds of startup founders, more than ever before, believe that we are in a tech "bubble." Sixty-five percent of founders believe that it's going to be harder for them to raise money next year, up 20 percent from last year's survey. Across the country on Wall Street, there are those who share these entrepreneurs' new pessimism. In a note, Bank of America's research division suggested that recent shifts in the market could produce a lot of pain and "volatility" in the coming year. Let's call it the "WeWork effect." For some context, new companies, especially startups from Silicon Valley, were able to raise substantially more money from private investors in the past decade than they were in previous years. This continued to be the case even as these companies, like Uber and WeWork, got bigger and bigger, approaching the size at which they'd traditionally need make a public stock offering in order to raise the necessary cash to keep growing. As Bloomberg columnist Matt Levine frequently says, "private markets are the new public markets," meaning that these companies are able to tap the same kinds of large investors as they would if they were publicly-traded without actually going public.

Read more of this story at Slashdot.