Monthly Archives: April 2020

Mandy Moore showed off some extremely '90s memorabilia from her early career

A pandemic isn’t the ideal time to release your first new album in a decade. Fortunately, Mandy Moore is using the time she would have been on tour to dig up some memorabilia from her early career.

Moore showed off her finds to Jimmy Kimmel on Thursday, which included an extremely ’90s photo with the Backstreet Boys and an unauthorised Mandy Moore book.

“It’s an actual book that was for sale for $12.95 that someone wrote without me knowing about it,” said Moore. “But it has lots of fun facts. Like, this page here is ’10 Things We Love About Mandy,’ and the very first fact is ‘Mandy is proud of her big feet.'”  Read more…

More about Jimmy Kimmel Live, Backstreet Boys, Jimmy Kimmel, 90s Nostalgia, and Mandy Moore

'Failed to provide any leadership': Seth Meyers examines how Trump left U.S. states on their own

The coronavirus pandemic has the U.S. is in shambles right now, but what else is new? Trump has continued to Trump throughout the crisis, stubbornly blustering on with his ill-informed ideas in defiance of all medical advice or logic. Now BuzzFeed News reports a man with no background in medical supplies was awarded a $69 million ventilator contract after tweeting at Trump that he could provide them. Shockingly, not a single one has arrived.

“There’s literally a law that allows him to compel companies to make ventilators, and he’s combing through his Twitter replies like he’s putting a band together,” quipped Late Night host Seth Meyers on Thursday. “Honestly, how long until he replies to a spam email offering free boner pills?” Read more…

More about Donald Trump, Seth Meyers, Late Night With Seth Meyers, Coronavirus, and Ventilators

The 'Parks and Recreation' reunion song was the most beautiful thing since Lil' Sebastian

Communities and friends band together in tough times — if anyone knows that, it’s the former employees of the Parks and Recreation department of Pawnee, Indiana. The core cast of Parks and Recreation reunited for a socially distant special on Thursday to raise money for Feeding America.

When Leslie (Amy Poehler) lets on that phone trees and media outreach might be slightly tiring her out both mentally and emotionally, Ron (Nick Offerman) knows exactly what to do. He merges their video call to include all of Leslie’s best friends as they join in a chorus of “5,000 Candles In The Wind.” The rendition quickly put a smile on Leslie’s face and ours as well. Sometimes a quick singalong with friends — even fictitious ones — is all you need for a little uplift.  Read more…

More about Entertainment, Parks And Recreation, Entertainment, and Movies Tv Shows

iPhone sales may be down, but business is booming for Apple Music and the App Store

Consumers aren’t rushing to buy as many iPhones, iPads, or other Apple products as before the pandemic, but the company’s services seem to be doing just fine. 

During Thursday’s earnings call, Apple disclosed that its services category, which includes the App Store and Apple TV+, hit an all-time revenue record of $13.3 billion for the second quarter.

The company saw strong performance within the App Store (for both downloads and search ads), Apple Music, video, and cloud services. App Store revenue also grew by double digits, as people continue to make in-app purchases and opt into subscriptions.  Read more…

More about Apple, Ipad, App Store, Macbook Air, and Revenue

The ‘PuffPacket’ could help researchers learn when, how, and why people vape

Vaping is a controversial habit: it certainly has its downsides, but anecdotally it’s a fantastic smoking cessation aid. The thing is, until behavioral scientists know a bit more about who does it, when, how much, and other details, its use will continue to be something of a mystery. That’s where the PuffPacket comes in.

Designed by Cornell engineers, the PuffPacket is a small device that attaches to e-cigarettes (or vape pens, or whatever you call yours) and precisely measures their use, sharing that information with a smartphone app for the user, and potentially researchers, to review later.

Some vaping devices are already set up with something like this, to tell a user when the cartridge is running low or a certain limit has been reached. But generally when vaping habits are studied, they rely on self-report data, not proprietary apps.

“The lack of continuous and objective understanding of vaping behaviors led us to develop PuffPacket to enable proper measurement, monitoring, tracking and recording of e-cigarette use, as opposed to inferring it from location and activity data, or self-reports,” said PhD student Alexander Adams, who led the creation of the device, in a Cornell news release.

The device fits a number of e-cigarette types, fitting between the mouthpiece and the heating element. It sits idle until the user breathes in, which activates the e-cigarette’s circuits and the PuffPacket’s as well. By paying attention to the voltage, it can tell how much liquid is being vaporized, as well as simpler measurements like the duration and timing of the inhalation.

An example using real data of how location and activity could be correlated with vaping.

This data is sent to the smartphone app via Bluetooth, where it is cross-referenced with other information, like location, motion, and other metadata. This may lead to identifiable patterns, like that someone vapes frequently when they walk in the morning but not the afternoon, or after coffee but not meals, or far more at the bar than at home — that sort of thing. Perhaps even (with the proper permissions) it could track use of certain apps — Instagram and vape? Post-game puff?

Some of these might be obvious, others not so much — but either way, it helps to have them backed up by real data rather than asking a person to estimate their own usage. They may not know, understand, or wish to admit their own habits.

“Getting these correlations between time of day, place and activity is important for understanding addiction. Research has shown that if you can keep people away from the paths of their normal habits, it can disrupt them,” said Adams.

No one is expecting people to voluntarily stick these things on their vape pens and share their info, but the design — which is being released as open source — could be used by researchers performing more formal studies. You can read the paper describing PuffPacket here.


Microsoft opens registration for its free, online Build 2020 developer conference

Microsoft has now opened the registration for the virtual edition of its online-only Build 2020 developer conference, which will take place from May 19 to 20.

Typically, the event draws over 6,000 developers, but because of the coronavirus pandemic, that’s obviously not an option. In contrast to Google, which completely scrapped its I/O developer conference this year, Microsoft decided to go ahead with the virtual event, though. But this will be a very different kind of Build — and not only because it’s online-only.

Not only will the keynotes be shorter (though there will still be Day 1 and Day 2 keynotes). but in response to feedback from developers that have attended previous events, the Microsoft team also decided to focus solely on that audience. In previous years, Microsoft often used Build to announce consumer products, just like Google does at I/O. But that won’t happen this year. And instead of using the keynotes to put an early spotlight on features that won’t be available for half a year or more, the event will be more about providing content that’s immediately useful for developers and new products that are either immediately available or only a couple of months out from getting into the hands of developers.

That also likely means that even though Microsoft CEO Satya Nadella will still keynote, there will be less talk about big picture company philosophy and more about developer tools and APIs.

Some of the keynotes and demos will be live, some will be pre-recorded, but overall, the look and feel of the event shouldn’t be all that different from what developers who previously watched Build from afar experienced. But it will be shorter and more focused than in previous years, which isn’t necessarily a bad thing.

Attendees sit in pods during the Microsoft Developers Build Conference in Seattle, Washington, U.S., on Monday, May 7, 2018. The Build conference, marking its second consecutive year in Seattle, is expected to put emphasis on the company’s cloud technologies and the artificial intelligence features within those services. Photographer: Grant Hindsley/Bloomberg via Getty Images


A new pro bono portal just launched for lawyers looking to help people hit hard by the pandemic

The coronavirus pandemic has laid low a lot of Americans, more than 62,000 of whom have already died since the beginning of March and 30 million more who are now out of work owing to the resultant shutdown of most businesses and public gathering places.

The ensuant crisis it has created is so massive that any type of coordinated effort is a challenge to pull together. Fortunately, that hasn’t stopped the American Bar Association and a justice tech company called Paladin that we introduced to readers last year. On the contrary, the SaaS startup — which helps legal teams sign up for pro bono opportunities, then makes their work and its impact easier to track — has teamed up with the ABA to help lawyers find pro bono opportunities specifically to help people affected by the coronavirus pandemic and other natural disasters.

It’s a way to accelerate work that the ABA has been doing for that last 13 years through its Young Lawyers Division’s Disaster Legal Services Program, and time, right now, is of the essence.

We were in touch yesterday with Paladin cofounder Kristen Sonday to learn more about the new portal, which helps lawyers filter opportunities by practice area, communities to serve, type of engagement and the ability to work remotely. We asked how it might better hep attorneys and those in need of their services to connect faster.

TC: How did this project come together? Who brought in whom to do what?

KS: The Paladin team saw a similar response to COVID-19 as previous legal emergencies in how volunteer attorneys were being recruited, which is mostly through online forms and spreadsheets [and] manual and ad hoc.

We had already built a version of a centralized opportunity guide for the Chicago Bar Foundation, and this seemed like a great opportunity to leverage that technology to help legal services organizations across the country recruit and manage disaster response volunteers seamlessly.

TC: Two other companies are involved here: LegalZoom and Clio. What role are they playing?

We approached LegalZoom and Clio to get involved given their reach on both the client and solo/small firm sides. They’re covering the costs of researching and developing this online resource, and we’re excited about the partnership because both organizations are well-known for their commitment to using technology to increase access to justice and improve legal services. filter opportunities by practice area, communities to serve, type of engagement and the ability to work remotely.

TC: Who are some of the people, or businesses, you see this helping right now? What are some common cases?

KS: Vulnerable individuals are experiencing a range of legal issues at unprecedented levels. Common cases include individuals filing for unemployment benefits; navigating housing issues and unlawful evictions; victims of domestic violence who have sheltered-in-place with an abuser; nonprofits and small businesses navigating cancelled contracts; and delays in court proceedings affecting thousands of Americans.

TC: And how did you put together this database?

KS: Paladin partnered its technology with the ABA DLS’s network to provide a centralized portal for state hotlines and disaster-related opportunities from legal services organizations across the country. All Legal Services Corporation grantees are able to post on the platform, and we look forward to seeing more organizations participating every day.

TC: Were there privacy concerns that needed to be addressed first?

KS: Opportunities have confidential information removed prior to posting, and attorneys are typically only provided confidential case information after the’ve been vetted by the referring organization.

TC: How do attorneys make themselves available on this portal?

KS: They sign up for specific pro bono opportunities that align with their interests and experience. Once they submit a form outlining their background to the referring organization, that organization will screen them and pair them with the most appropriate pro bono client.

TC: Are you generating any revenue from this project?

KS: Our mission is to increase access to justice by supporting pro bono legal services, so this project seemed like the perfect way to leverage Paladin’s expertise and increase impact. There is no cost for attorneys to use, nor cost to the legal services organizations posting opportunities. It is a great way to raise awareness of Paladin’s work more broadly.

Lawyers who are willing to provide pro bono legal services can sign up and view cases at here.


Amazon Q1 beats on net sales of $75.5B but posts net income of $2.5B, down $1B on a year ago

Amazon has been one of the biggest names synonymous with how the consumer masses are experiencing life under lockdown: its site lets you buy anything from soup to nuts, from books to baking pans for all your sourdough; and via its streaming services, it gives you many ways stay entertained. But it can also be a source of major frustration, when you find yourself unable to book slots for deliveries, or are facing an army of sellers trying to price gouge you for hot items like masks or toilet paper.

Today, the company reported first quarter earnings that bore out the first of these in spades, but at a cost to profitability as it works to serve a public under a whole new set of challenging conditions.

The company reported net sales of $75.5 billion, up 26% on a year ago, a huge boost on the $59.7 billion it made in net sales in the first quarter a year ago. $41 billion of the sum was attributable to product sales and $33 billion to services (which includes AWS, but also streaming and other non-physical goods).

But earnings per share took a hit, with basic EPS at $5.09 and diluted EPS at $5.01, and net income declining down to $2.535 billion versus $3.561 billion a year ago.

Operating income was also down to $4 billion versus operating income of $4.4 billion in the same quarter a year ago.

Analysts on average were expecting EPS of $6.25 on revenues of $73.61 billion in sales. It was a repeat of the pattern we saw from eBay’s earnings yesterday, albeit on a much, much bigger scale.

Jeff Bezos, Amazon’s colourful founder and CEO, acknowledged the challenges even the mighty Amazon is facing, but also said that the company plans to double down on spending to face up to serving people during the COVID-19 pandemic, whatever it might bring. It’s a long statement (in what is a very, very wordy press release overall):

“From online shopping to AWS to Prime Video and Fire TV, the current crisis is demonstrating the adaptability and durability of Amazon’s business as never before, but it’s also the hardest time we’ve ever faced,” he said in a statement. “We are inspired by all the essential workers we see doing their jobs — nurses and doctors, grocery store cashiers, police officers, and our own extraordinary frontline employees. The service we provide has never been more critical, and the people doing the frontline work — our employees and all the contractors throughout our supply chain — are counting on us to keep them safe as they do that work. We’re not going to let them down. Providing for customers and protecting employees as this crisis continues for more months is going to take skill, humility, invention, and money. If you’re a shareowner in Amazon, you may want to take a seat, because we’re not thinking small.

“Under normal circumstances, in this coming Q2, we’d expect to make some $4 billion or more in operating profit. But these aren’t normal circumstances. Instead, we expect to spend the entirety of that $4 billion, and perhaps a bit more, on COVID-related expenses getting products to customers and keeping employees safe. This includes investments in personal protective equipment, enhanced cleaning of our facilities, less efficient process paths that better allow for effective social distancing, higher wages for hourly teams, and hundreds of millions to develop our own COVID-19 testing capabilities. There is a lot of uncertainty in the world right now, and the best investment we can make is in the safety and well-being of our hundreds of thousands of employees. I’m confident that our long-term oriented shareowners will understand and embrace our approach, and that in fact they would expect no less.”

Of note: Amazon Web Services accounted for $10.2 billion in sales, up 33% on the same quarter a year ago. North America accounted for about $44 billion of the company’s net sales, versus $19 billion for the international segment.

At a time when we’ve seen tens of thousands of people laid off across the technology sector, Amazon has been one of the few companies to hire, specifically to staff up with 100,000 extra workers across warehouses and its logistics network to meet surging demand from buyers. That has not always been smooth sailing however, with accusations of poor and potentially health-threatening working conditions.

This has been a thorny issue for the company, so it’s no surprise that in its earnings report, it prominently reminded investors that it has made “over 150 significant process changes in our operations network and Whole Foods Market stores to help teams stay healthy — and we conduct daily audits of the measures we’ve put into place.”

It also noted that it has procured 100 million face masks (presumably not on Amazon itself, where economical ones have been very hard to find) and are requiring that they be worn by all associates, drivers, and support staff in our operations network. “We purchased more than 1,000 thermal cameras and 31,000 thermometers, which we are using to conduct mandatory daily temperature checks for employees and support staff throughout our operations sites and Whole Foods Market stores,” it noted.

Those thermal cameras have also, however, been a point of contention: Reuters this week reported that those cameras were sourced from Dahua, a Chinese company currently blacklisted by the US government.

More to come.



OpenAI’s new experiments in music generation create an uncanny valley Elvis

AI-generated music is a fascinating new field, and deep-pocketed research outfit OpenAI has hit new heights in it, creating recreations of songs in the style of Elvis, 2Pac, and others. The results are convincing, but fall squarely in the unnerving “uncanny valley” of audio, sounding rather like good, but drunk, karaoke heard through a haze of drugs.

Jukebox, the organization’s new music-generating system, was detailed in a blog post and paper published today. OpenAI produced some interesting work almost exactly a year ago with MuseNet, a machine learning system that, having ingested a great deal of MIDI-based music, was able to mix and match genres and instruments.

But MIDI is a simpler format than final recorded music with live instruments, since the former consists of discrete notes and key presses rather than complex harmonics and voices.

If you wanted an AI to examine the structure of a classical piano piece, the timing and key presses might only amount to a couple thousand pieces of information. Recorded audio is far denser, with (usually) 44,100 samples per second.

Machine learning systems that learn and imitate things like instruments and voice work by looking at the most recent words or sounds and predicting the next few, but they generally operate on the order of tens or a hundred pieces of data — the last 30 words or notes predict what the next 30 will be, for instance. So how can a computer learn how a tiny fraction of a waveform 10 seconds and 440,000 samples into a song compare with a sample 90 seconds and 4 million samples in?

OpenAI’s solution is to break down the song into more digestible parts — not quite key and chord, but something like that, a machine-palatable summary of 1/128th of a second of the song, picked from a “vocabulary” of 2048 options. To be honest it’s hard to create an analogy because this is so unlike the way humans remember or understand things — as far as we even understand that.

It doesn’t actually use color swatches – that’s just to indicate that it’s breaking the waveform down into pieces.

The end result is that the AI agent has a reliable way to break down a song into digestible bits that are big enough that there aren’t too many to track, but small enough that they can reliably reconstruct the sound of a song. The process is much more complex than it sounds here; Reliably breaking down a song to a series of “words” and then reconstructing it from them is the core of the new research, but the technical details I’ll let the OpenAI team to explain in their paper.

The system also had to learn how to parse the lyrics in a song, which like most things in this domain is more complicated than it sounds. Our ability to remember and use vocal patterns is partly innate and partly learned, and we tend to take for granted how powerful it is. Computers have no such ability and must learn how to pick out a voice from a mix, understand what it’s saying, and match that to lyrics that are nothing more than a series of words with no information on key, tempo, and all the rest. Nevertheless the OpenAI system does it to a satisfactory degree.

Jukebox is able to accomplish a variety of musical tasks, and while the results aren’t what you might call single material, it must be kept in mind that there’s very little like this out there now, able to rebuild a song from scratch that’s recognizable as being like the target artist. Trained on 1.2 million songs, the system in the end has one multifaceted ability it accomplishes these tasks with: essentially, improvising a song given lyrics and the style it’s learned from ingesting others by that artist.

So given its knowledge of how Ella Fitzgerald sings and the way instruments generally accompany her, it can sing a rendition of “At Long last Love” in a way that sounds like her but definitely isn’t what Cole Porter had in mind. (Samples for these examples and more are included near the top of the OpenAI blog post.)

Jukebox can also sing entirely original lyrics in another’s style, like this truly strange Elvis song, “Mitosis,” written by another AI language model:

In case you didn’t catch that:

From dust we came with humble start;
From dirt to lipid to cell to heart.
With [mitosis] with [meiosis] with time,
At last we woke up with a mind.
From dust we came with friendly help;
From dirt to tube to chip to rack.
With S. G. D. with recurrence with compute,
At last we woke up with a soul.

Yes, it’s “Elvis” using cell division as a metaphor for life, as imagined by an AI. What a world we live in.

Lastly there’s the “completion” task, where Jukebox learns (in addition to the base learning from its library) from the first 12 seconds of a song and uses that to generate the rest in a similar style. The switch from original to AI-generated sounds a bit like the ether just kicked in.

While MuseNet could be played with more or less in real time due to its lesser complexity, Jukebox is hugely computation intensive, taking hours to generate a single second of music. “We shared Jukebox with an initial set of 10 musicians from various genres… these musicians did not find it immediately applicable to their creative process,” the authors note drily. Still, it’s fun and fascinating research and, given the current cadence, we can expect an even further improved version of the OpenAI music effort next April.


Subscribe to our mailing list

Join our online network:

Copyright © 2011-2018 Modulates