• Home
  • About Us
  • Contact Us
  • Privacy Policy
  • Special Offers
Business Intelligence Info
  • Business Intelligence
    • BI News and Info
    • Big Data
    • Mobile and Cloud
    • Self-Service BI
  • CRM
    • CRM News and Info
    • InfusionSoft
    • Microsoft Dynamics CRM
    • NetSuite
    • OnContact
    • Salesforce
    • Workbooks
  • Data Mining
    • Pentaho
    • Sisense
    • Tableau
    • TIBCO Spotfire
  • Data Warehousing
    • DWH News and Info
    • IBM DB2
    • Microsoft SQL Server
    • Oracle
    • Teradata
  • Predictive Analytics
    • FICO
    • KNIME
    • Mathematica
    • Matlab
    • Minitab
    • RapidMiner
    • Revolution
    • SAP
    • SAS/SPSS
  • Humor

Tag Archives: Amazon

Amazon Textract adds handwriting recognition and support for new languages

November 14, 2020   Big Data
 Amazon Textract adds handwriting recognition and support for new languages

Best practices for a successful AI Center of Excellence

A guide for both CoEs and business units

Download Guide

Amazon today announced small enhancements to Textract, its service that extracts printed text and other data from documents, as well as tables and forms, using machine learning. As of today, Textract now supports handwriting in English documents, in addition to files typed in Spanish, Portuguese, French, German, and Italian.

Amazon rightly notes that many documents, like medical intake forms or employment applications, contain a combination of handwritten and printed text. While rivals like Google and Amazon have offered handwriting recognition-as-a-service for some time, Amazon says customer requests spurred the launch of its own solution, which works with both free-form text and text embedded in tables and forms.

Amazon Web Services (AWS) customers can use the Textract handwriting recognition feature in conjunction with Amazon’s Augmented AI (A2I) for improved performance. A2I lets users build workflows for human review of the machine learning system’s predictions, either by employees or AWS Marketplace contractors. Documents can be uploaded on the Amazon Textract console or sent using the AWS Command Line Interface or AWS software development kits.

New features aside, Textract remains less holistic than Google’s newly launched Document AI (DocAI) platform, a console for document processing hosted in Google Cloud. It supports the creation and customization of processing workflows built with a predefined taxonomy without the need to perform additional data mapping or training. DocAI offers general processors, including a form parser, W9 parser, optical character recognition, document splitter, and custom workflows for domain-specific documents. The parsers can classify information in documents like addresses, account numbers, and signatures, as well as extracting data like supplier names, invoice dates, and payment terms.

Companies spend an average of $ 20 to file and store a single document, according to some estimates, and only 18% of companies consider themselves paperless. An IDC report revealed that document-related challenges account for a 21.3% productivity loss, and U.S. companies waste a collective $ 8 billion annually managing paperwork. Beyond AI-powered products from tech giants like Amazon and Google, this has given rise to a cottage industry, with startups like Rossum, Anvil, PandaDoc, and others competing on document processing accuracy and pricing.

Amazon also announced updates to Amazon Lex and Amazon Polly earlier this week. Lex, a platform for building conversational interfaces, newly supports French, Spanish, Italian, and Canadian French. Polly, which turns text into lifelike speech, now features Amazon’s first Australian English voice, Olivia, synthesized with the same neural text-to-speech technology powering Amazon’s Brand Voice service.


Best practices for a successful AI Center of Excellence:

A guide for both CoEs and business units Access here


Let’s block ads! (Why?)

Big Data – VentureBeat

Read More

Hands-on: Amazon Fresh grocery stores tease brick-and-mortar retail’s future

November 2, 2020   Big Data

Maintain your employer brand in a pandemic

Read the VentureBeat Jobs guide to employer branding

Download eBook

There were no lines outside Irvine, California’s new Amazon Fresh grocery store on its opening day last week, despite the fact that it was only the second such location in the world — and the first to be open to the public on day one. But after early visitors discovered the store’s high-tech shopping carts, two lines formed over the weekend, stretching past Amazon’s front doors to adjacent retailers in the suburban plaza. One line was longer and moving slower than the other.

“Do you want to try the Dash Cart?” an employee asked people near the end of the queue. “If not, you can move into the shorter line, and you’ll get in faster.”

We were there specifically for the Dash Cart: Friends told us that it was worth the 10- to 20-minute wait to go hands-on with one of the 25 magical shopping carts, since their integrated touchscreens and cameras were the key to Amazon’s next-generation shopping experience. The Dash Cart felt like the future of brick-and-mortar retail, they said, even though the rest of the store wasn’t that amazing.

Our friends were correct, but there’s more to the Amazon Fresh story than just Dash Carts. Here’s what it’s like to visit the supermarket of the future, today, as it’s been dreamed up and implemented by Amazon.

A boxy, spartan layout, seemingly by design

Unlike Whole Foods, the high-end supermarket chain Amazon acquired in 2017, Amazon Fresh stores look like small warehouses, and have all the charm of Walmart’s grocery sections, minus two-thirds the people and half the choices. From the signage to the aisles and specialty counters, Amazon’s latest store feels as if it was designed largely by engineers, and conceived to be as easy as possible to retrofit inside another retailer’s abandoned space — in this case, the 40,000 feet formerly occupied by a Babies R Us store.

Apart from the produce, nothing about the environment feels organic: Floors are spartan, displays are boxy, and everything looks to have been optimized for customers by computers, rather than humans. There are places to purchase whole cooked chickens for $ 4.97 and pizzas for $ 8.99, but nowhere to sit and eat them. A staffed customer service area is in the back, not the front, which instead allocates a lot of interior space to managing shopping carts.

Even the baked goods, which in other stores flow attractively off the edges of store shelves, seem to have been assigned to a specific corner of Amazon Fresh and told to stay firmly within the lines. If it wasn’t for a cadre of friendly greeters, walking through Amazon Fresh would feel more like visiting a warehouse or Costco than shopping in a typical supermarket of its size.

That feeling extends to how Amazon Fresh uses — and doesn’t use — people in its operations. Instead of having employees answer inventory-related questions, Amazon scatters Alexa terminals throughout the store, offering AI guidance on item locations, wine pairings, and measurement unit conversions. On our first visit, the Alexa terminals were both working and helpful, accurately pointing us towards items we wanted to locate. But on our second visit, all of the terminals were experiencing “connectivity issues,” perhaps the closest Amazon Fresh stores will get to a business-disrupting employee strike.

The Alexa terminals suggest that Amazon wants to staff Fresh stores as leanly as possible, even if it’s liberally using employees during the launch phase to address potential customer pain points. There were lots of Fresh staffers — too many, really — constantly restocking shelves while otherwise keeping to themselves, plus the aforementioned greeters at the front doors to help get people in and out of the store. In traditional supermarkets, all of these employees might be floaters who move from place to place as needed, alternating between helping customers and restocking shelves. But at Amazon Fresh, Alexa could help reduce the need for greeters as customers become familiar with the technology, and shelving recalibrations could reduce the need for such frequent stock replenishment.

Lower labor costs could translate directly into lower prices. And half of the new store’s appeal is reasonable pricing — that’s the single biggest problem with Whole Foods, which offers well-heeled customers an impressive selection of high-end foods and beverages that just aren’t affordable to the masses. By contrast, Amazon Fresh is clearly aimed at middle-income shoppers who still want to do some of their purchasing and browsing in person instead of on a computer screen. There are a handful of fancy items on the shelves, such as $ 10 pints of McConnell’s ice cream, but most of the signage is directed towards selling 15-cent bananas and 89-cent loaves of bread, rather than champagne and caviar.

Dash Cart as a solution and a problem

The more exciting part of Amazon Fresh is the Dash Cart, a shopping cart that uses sensors and smartphone technology to replace checkout lanes and standalone produce scales. As mentioned above, you don’t have to use a Dash Cart to shop at Amazon Fresh, and despite its speedy name, you’ll likely get in and out of the store faster without waiting for one. But without a Dash Cart, the shopping experience isn’t hugely different from any old small suburban supermarket you’ve previously visited.

Once you make it through the Dash Cart waiting line, you’ll get a three-minute human tutorial that explains how to link your cart to your Amazon app with a QR code, scan packaged items by dropping them into one of two included paper bags, and add produce items by inputting four-digit PLU codes into the cart’s tablet screen. These steps are supposed to eliminate the need for employees to check you out and bag your purchases; instead, the cart’s cameras and scale track everything you place in the bags, so when you leave the store, your Amazon account is automatically charged for whatever you bagged yourself. It’s an evolution of what Amazon pioneered with much smaller Amazon Go stores years ago.

Dash Carts are cool in concept, but their execution leaves a lot to be desired. On a positive note, their cameras and software did a good job with accurately scanning items we placed in the bags, and automatically removing items if we pulled them from the cart. The tablet-like touchscreen worked as expected, and though the scale inside the cart wasn’t fast, it could — with practice — be faster than walking over to a standalone produce scale and printing out a label for each item.

On the other hand, the Dash Carts had limitations that beg to be resolved in future iterations. Each cart is limited to two bags, which restricts your ability to complete a full shopping trip, and limits Amazon’s maximum take per shopper. You can’t overfill the bags, lest the cart’s cameras become incapable of seeing what’s inside. Additionally, Amazon is so concerned about theft or damage that it swaps each Dash Cart for a regular one before customers leave for the parking lot, or hands you the bags to carry to whatever distant parking space you selected. These are the sorts of practical inconveniences that could kill Dash Cart’s utility for some people.

Real-world glitches also undermined our Dash Cart experience. One of our bags ripped and needed to be replaced during the cart-to-cart transfer. We also had to go through a manual checkout line — including rescanning and rebagging every item — because our cart’s integrated code scanner couldn’t recognize an Amazon coupon. Staff said that the carts were somewhat finicky and had been experiencing hiccups like this.

Whenever one of these issues with the Dash Cart popped up, we felt as if we were holding up people who were waiting behind us, even though the issues weren’t really our fault. Other delays, such as learning how to enter PLU codes and weigh produce, caused Dash Cart users to abruptly stop mid-aisle and fidget with the tablet’s screen. We noticed some customers without the high-tech carts becoming visibly frustrated with other customers’ Dash-inspired touch interactions, but cart users seemed to be too focused on their screens to notice.

Could data make all the difference for future retailers?

It’s easy to overlook a key element of this retail experience — the intersection between Amazon.com and Amazon Fresh — because it’s so muddled at the moment. But it could wind up being a critical differentiator for Amazon’s brick-and-mortar ventures going forward.

As part of the initial onboarding experience, Amazon openly encourages Dash Cart users to digitally manage their shopping lists with the cart and browse current in-store specials using their phones. This is a mess for two reasons: The cart’s integrated shopping list management software is extremely limited, and the idea of asking users to check not just one but two touchscreens while they’re shopping is just straight-out crazy. No one wants to be stuck behind that guy who’s blocking shelves or freezers while browsing through lists and brochures. If Larry David ever visits Amazon Fresh, there’s enough material here for an entire Curb Your Enthusiasm sub-plot.

Yet there’s obvious value in tying the internet directly — and more thoughtfully — to a customer’s shopping cart. Your first visit to an Amazon Fresh store could conceivably be your last trip through its aisles: Amazon could just present you with a list of the items you purchased, offer to reorder them, and make them instantly available for either pickup or delivery. That could eliminate the need (and the premium people currently pay) for Instacart. It also could reduce the footprints of future Amazon Fresh stores by lowering the number of people who simultaneously walk through them, enabling many customers to complete transactions using the equivalent of drive-through windows.

Amazon is technically doing some if not most of these things already, but it needs to refine its smartphone and cart software to make the end-to-end experience intuitive and frictionless for customers. Somewhat ironically, the sign that it has succeeded will be if its Fresh grocery stores aren’t packed with people but are still hugely profitable, which is to say that they’ll be moving tons of products without the packed aisles and long lines normally associated with successful supermarkets.

The best of the rest of Amazon, plus coupons

One thing we loved at Amazon Fresh was an area labeled “Customer Service, Returns & Pick Up.” Normally, these things are found very close to the entrance of a supermarket, but at Amazon Fresh, they’re in the back, a decision that was likely made to get returns and pick-ups closer to the store’s storage areas and loading docks. Customers can pick up items from Amazon lockers and drop off Amazon returns — conveniences that simultaneously provide an incentive to do grocery shopping while eliminating the need to visit standalone Amazon shipping and return locations, something we can see ourselves using at least occasionally.

Amazon Fresh also includes a limited selection of the online retailer’s popular gadgets and books. We spotted the José Andrés cookbook Vegetables Unleashed and the Death & Co. cocktail guide on shelves only a short distance away from Fire tablets and Echo speakers, none of which we were looking to purchase at a supermarket — but then, we had already bought some of them online in the past. Over time, new items will replace them, and we might have reason to consider buying non-grocery goods at Amazon Fresh, as well.

At this stage, it would be hard to describe Amazon Fresh as the guaranteed future of brick-and-mortar retailing; the experience currently feels closer to a public beta test than a fully formed and polished business. Visitors can certainly have a normal or even a unique experience in the store, but they’re actually guinea pigs in a grand experiment that runs smoothly — until it doesn’t.

To Amazon’s credit, the speed bumps aren’t too daunting. Moreover, the company is actively addressing problems by handing out coupons to apologize for technical issues, and on one of our two visits, was giving away free cans of sparkling water and refrigerator magnets to everyone exiting the store. Despite the glitches, we didn’t see anyone leaving the store angry, and between the coupons and the small number of bags we left with, we were already planning our next trip to the store as we walked out to our car.

Only Amazon knows whether such a mixed but positive impression counts as “mission accomplished” or whether its early Amazon Fresh grocery customers are just helping it refine a larger campaign to completely dominate the retail world. Thanks to Amazon’s growing scale and unquestionable ambition, the Fresh grocery stores could either become very real challengers to traditional supermarkets — or fizzle out as experiments that made little difference to the company’s bottom line.

If you’re interested in seeing Amazon Fresh for yourself, you can visit the new store at 13672 Jamboree Road in Irvine, or the first location — open to the public since September — at 6245 Topanga Canyon Boulevard in Woodland Hills, California. Both stores are open from 7 a.m. to 10 p.m., seven days a week.


How startups are scaling communication:

The pandemic is making startups take a close look at ramping up their communication solutions. Learn how


Let’s block ads! (Why?)

Big Data – VentureBeat

Read More

U.S. Congress calls for antitrust reforms to limit powers of Amazon, Apple, Facebook, and Google

October 7, 2020   Big Data
 U.S. Congress calls for antitrust reforms to limit powers of Amazon, Apple, Facebook, and Google

Members of Congress investigating the activity of Amazon, Apple, Facebook, and Google say antitrust law reform is needed to safeguard democracy and “ensure that our economy remains vibrant and open in the digital age.” The findings comes from a document released today (PDF), the culmination of a 16-month long investigation carried out by the antitrust subcommittee, a part of the House Judiciary committee.

The report concludes that although each Big Tech company maintains different kinds of monopolies, each effectively acts as a gatekeeper in digital markets today with the power to pick winners and acquire or dispose of competitors. The document details monopolization and anticompetitive behavior by each of the companies. Members of Congress say that power is used to extract concessions and dictate terms to competitors in ways that wouldn’t be possible in a competitive market, like Apple’s high App Store fee or Amazon third-party seller fees for millions of small businesses.

“To put it simply, companies that once were scrappy, underdog startups that challenged the status quo have become the kinds of monopolies we last saw in the era of oil barons and railroad tycoons,” the report reads. “Although these firms have delivered clear benefits to society, the dominance of Amazon, Apple, Facebook, and Google has come at a price. These firms typically run the marketplace while also competing in it—a position that enables them to write one set of rules for others, while they play by another, or to engage in a form of their own private quasi regulation that is unaccountable to anyone but themselves.”

The report recommends Congress and the antitrust subcommittee pass legislation to strengthen enforcement of existing antitrust law, and supports data portability and interoperability so users can transfer data to another platform in order to promote competition. Also among recommendations:


You can’t solo security

COVID-19 game security report: Learn the latest attack trends in gaming. Access here


– Strengthen portions of the Sherman Act which addresses competition and monopoly by adding a prohibition of abuse of dominance or monopoly leveraging.

– Bring back stronger oversight of antitrust enforcement by Congress; multiple incidents were found where regulators failed to stop monopolists from consolidating market share and eliminating competitors

– Put rules in place to prevent favoritism, discrimination, or placing for example an Amazon product ahead of a third-party seller using Amazon’s platform

– Increase scrutiny of merger and acquisition activity to ensure monopolies don’t consume competitors. The report found that since 1998, the four companies have acquired more than 500 startups. On the topic of the 2012 acquisition of Instagram by Facebook, an unidentified former senior Instagram employee who testified to the committee last week said “It was collusion, but within an internal monopoly. If you own two social media utilities, they should not be allowed to shore each other up. It’s unclear to me why this should not be illegal.”

– Shift the burden of proving a merger isn’t anticompetitive from regulators to Big Tech companies with merger presumptions. As the report reads: “Under this change, any acquisition by a dominant platform would be presumed anticompetitive unless the merging parties could show that the transaction was necessary for serving the public interest and that similar benefits could not be achieved through internal growth and expansion.”

– Give news publishers the ability to collectively bargain with Facebook and Google

VentureBeat reached out to Amazon, Apple, Facebook, and Google for response to the report and its conclusions. In a statement shared with VentureBeat, Apple argues that the company does not maintain dominant market share in any category, and that “we have always said that scrutiny is reasonable and appropriate but we vehemently disagree with the conclusions reached in this staff report with respect to Apple.”

An Amazon blog post in response to the report Tuesday warned against “misguided interventions” and criticized what it calls “fringe notions of antitrust.” A Google spokesperson told VentureBeat in a statement that Google competes fairly and that “Americans simply don’t want Congress to break Google’s products or harm the free services they use  every day.”

A Facebook company spokesperson told VentureBeat in a statement that a competitive landscape existed when Instagram and WhatsApp acquisitions took place, exists today, and that “Regulators thoroughly reviewed each deal and rightly did not see any reason to stop them at the time.”

Antitrust law in the U.S. began to rein in the power of companies dominant in industries like steel and railroads starting in the late 1800s, however U.S. regulatory officials have been far less likely to act in recent decades to ensure open, competitive markets. The report titled “Investigation of Competition in Digital Markets” says each of the companies maintain different  monopolies: Facebook in social media and advertising, Google in search and advertising, Amazon in online retail, and Apple through the App Store. The report states that online markets became particularly vulnerable to market concentration and monopolization in the past decade. Hundreds of mergers and acquisitions by Big Tech companies in recent years are outlined in the report alongside general declines in early-stage startup funding and business formation.

User data, leveraging existing monopolies, acquisitions, and anticompetitive practices helped defend the dominance of Big Tech businesses. All four of the tech firms have recently focused on acquiring startups in artificial intelligence other emerging technology markets in order to “control the technology of tomorrow,” the report finds. Members of the subcommittee and staff also describe growing political power for tech companies.

“Through a combination of direct lobbying and funding think tanks and academics, the dominant platforms have expanded their sphere of influence, further shaping how they are governed and regulated,” the report reads.

Authors of the report also assert that though companies like Facebook offer services with no monetary cost, people pay a cost due to “diminished consumer choice, eroded innovation and entrepreneurship in the U.S. economy, weakened the vibrancy of the free and diverse press, and undermined Americans’ privacy.”

The report concludes that every member of the committee led by Democratic members of Congress may not agree with all investigations results or recommendations. Around the same time that the committee report was released, five Republican members of the antitrust committee released their own report. The 28-page staff document calls says “Big Tech is out to get conservatives” and calls for reform of Section 230 liability protection extended to social media platforms online. Prior to the report’s expected release on Monday, multiple news outlets reported that the report’s release was postponed in order to include additional Republican feedback, and address portions of the draft document considered untenable to some committee members.

Antitrust is important to small business prosperity, startup innovation, competitive business practices, and democracy. Economists fear the dominance of large companies will grow even larger in the U.S. as a large number of small businesses are wiped out by a recession sparked by COVID-19. By contrast, since July Big Tech companies reported record profits in quarterly earnings. Currently, 8 of the 10 largest companies in the world are in technology.

This is the first significant antitrust investigation by a congressional committee in decades. Antitrust hearings held in the late 1990s preceded a US v. Microsoft lawsuit and settlement, which some say enabled the growth of companies like Amazon, Facebook, and Google. Legislation stemming from the report may not be proposed until a new term begins in 2021, but the document is meant to lay out a menu of options for Congress to regulate Big Tech and antitrust law reform for the coming months and years.

The investigatory process leading up to the release of the report today consisted of seven hearings and more than a million document. The process reached a high point in July when Amazon CEO Jeff Bezos, Apple CEO Tim Cook, Google and Alphabet CEO Sundar Pichai, and Facebook CEO Mark Zuckerberg testified before the House Judiciary committee.

Speaking with antitrust experts at Yale Law School Sunday, Antitrust subcommittee chair David Cicilline (D-RI) said he believes some basic recommendations in the report are attainable like enacting separations of power that prevent the owner of a platform from favoring their own products over others, and giving federal agencies that enforce antitrust law — like the Federal Trade Commission (FTC) — resources for more robust enforcement. He also said that any meaningful legislation from Congress will require the help of the American people due to the power and resources available to Amazon, Apple, Facebook, and Google.

In the realm of other actions being taken in Washington DC to limit powerful businesses, the Senate Commerce committee held hearings related to antitrust last month, and recently subpoenaed CEOs of Facebook, Google, and Twitter to testify in the coming weeks about potential Section 230 reforms. President Trump called for Section 230 reforms earlier today after Twitter labeled his tweet as misleading and potentially harmful information about COVID-19. A Cornell University analysis last week have named Trump the biggest source of COVID-19 misinformation in the U.S. today.

The U.S. Department of Justice is expected to launch a case against Google later this week.

Outside of antitrust activity in the U.S., Big Tech companies continue to face antitrust lawsuits and backlash in Australia and the European Union. According to multiple reports last week, a draft of the Digital Services Act being considered by the European Parliament will require dominant tech companies to share some data with rivals and place limits on how companies can use consumer data. Chinese officials are also considering antitrust action against Google due to Android’s dominance of smartphone markets, a person familiar with the matter told Reuters.

Updated 6:29 p.m. to include responses from Amazon, Apple, and Google, and 7:01 p.m. to include a Facebook response.

Let’s block ads! (Why?)

Big Data – VentureBeat

Read More

AI Weekly: Amazon went wide with Alexa; now it’s going deep

September 26, 2020   Big Data
 AI Weekly: Amazon went wide with Alexa; now it’s going deep

Automation and Jobs

Read our latest special issue.

Open Now

Amazon’s naked ambition to become part of everyone’s daily lives was on full display this week at its annual hardware event. It announced a slew of new Alexa-powered devices, including a home surveillance drone, a suite of Ring-branded car alarm systems, and miscellany like an adorable little kids’ Echo device. But it’s clear Amazon’s strategy has shifted, even if only for a product cycle, from going wide to going deep.

Last year, Amazon baked its virtual assistant into any household device that could accommodate a chip. Its list of new widgets with Alexa seemed a mile long and included a menagerie of home goods, like lamps and microwaves. The company also announced device partnerships that ensure Alexa would live on some devices alongside other virtual assistants, tools to make it easier for developers to create Alexa skills, networking devices and capabilities, and wearables. It was a volume play and an aggressive bid to build out its ecosystem in even more markets.

This year, Amazon had fewer devices to announce, but it played up ways it has made Alexa itself better than ever. That’s the second prong of the strategy here: Get Alexa everywhere, then improve the marquee features such that the experience for users eclipses anything the competition offers.

As is always the case at these sorts of events, Amazon talked big and dreamy about all the new Alexa features. Users will find out for themselves whether this is the real deal or just hype when Amazon rolls out updates over the course of the next year (they’re landing on smart home devices first). But on paper and in the staged demos, Alexa’s new capabilities certainly seem to bring it a step closer to the holy grail of speaking to a virtual assistant just like talking to a person.

That’s the crux of what Amazon says it has done to improve Alexa, imbuing it with AI to make it more humanlike. This includes picking up nuances in speech and adjusting its own cadence, asking its human conversation partner for clarifications to fill in knowledge, and using feedback like “Alexa, that’s wrong” to learn and correct itself.

Amazon is particularly proud of the new natural turn-taking capabilities, which help Alexa understand the vagaries of human conversation. For example, in a staged demo two friends talked about ordering a pizza through an Alexa device. Like normal humans, they didn’t use each other’s names in the conversation, they paused to think, they changed their minds and adjusted the order, and so on. Alexa “knew” when to chime in, as well as when they were talking to each other and not to the Alexa device.

At the event, Alexa VP and head scientist Rohit Prasad said this required “real invention” and that the team went beyond just natural language processing (NLP) to embrace multisensory AI — acoustic, linguistic, and visual cues. And he said those all happen locally, on the device itself.

This is thanks to Amazon’s new AZ1 Neural Edge processor, which is designed to accelerate machine learning applications on-device instead of in the cloud. In the event liveblog, Amazon said: “With AZ1, powerful inference engines can run quickly on the edge — starting with an all-neural speech recognition model that will process speech faster, making Alexa even more responsive.” There are scant details available about the chip, but it likely portends a near future when Alexa devices are able to do more meaningful virtual assisting without an internet connection.

Given the utter lack of information about the AZ1, it’s impossible to say what it can or can’t do. But it would a potential game changer if it was able to handle all of Alexa’s new tricks on devices as simple as an Echo smart speaker. There could be positive privacy implications, too, if users were able to enjoy a newly powerful Alexa on-device, keeping their voice recordings from Amazon’s cloud.

But for Amazon, going deep isn’t just about a more humanlike Alexa; it involves pulling people further into its ecosystem, which Amazon hopes is the sum of adding device and service ubiquity to more engaging user experiences.

Part of that effort centers on Ring devices, which now include not just front-door home security products but also car security products and a small autonomous drone for the inside of your home. They’re essentially surveillance devices — and taken together, they form an ecosystem of surveillance devices and services that Amazon owns, and that connects to law enforcement. You can buy into it as deeply as you want, creating a surveillance bubble inside your home, around your home, and on board your vehicles, regardless of where you’ve parked them. The tension over Ring devices — what and who they record, where those recordings go, and who uses them for what purpose — will only be amplified by this in-home drone and the car alarm and camera.

Whether Amazon goes deep or wide, what hasn’t changed is that it wants to be omnipresent in our lives. And with every event’s worth of new devices and capabilities, the company takes another step closer to that goal.

Let’s block ads! (Why?)

Big Data – VentureBeat

Read More

Antitrust experts weigh in on breaking up Amazon, Apple, Facebook, and Google

August 2, 2020   Big Data
 Antitrust experts weigh in on breaking up Amazon, Apple, Facebook, and Google

VB Transform

Watch every session from the AI event of the year

On-Demand

Watch Now

Gary Reback is perhaps best known as the lawyer who helped convince the U.S. Department of Justice to bring an antitrust lawsuit against Microsoft in the 1990s. He watched the majority of the House Judiciary Committee hearing earlier this week with Amazon CEO Jeff Bezos, Apple CEO Tim Cook, Google CEO Sundar Pichai, and Facebook CEO Mark Zuckerberg. Reback still remembers a Senate hearing that put pressure on the government and led to the filing of United States v. Microsoft Corporation in 1998. That’s why he focused more on lines of questioning this week than on answers from CEOs of companies that have amassed unprecedented wealth and power.

“To me, the core turning point was right at the beginning, basically,” he said, pointing to sharp and detailed questioning in reference to a range of documents obtained in the course of the committee’s assessment.

The question on Wednesday was whether tech giants have grown too powerful, and if you watched the hearings, many members of Congress leave no doubt that they believe the answer is yes. In his opening remarks, for example, U.S. House Antitrust subcommittee chair David Cicilline (D-RI) said tech giants enjoy the power to pick winners and losers in the private market and that consumers have “no escape from surveillance” because there are no alternatives. He added “What’s at stake is whether we let ourselves be governed by private monopolies.”

VentureBeat spoke with Reback and other antitrust experts about what stood out during the hearing and what should happen next. Each favors some form of action or legislation to address the litany of anticompetitive practice allegations that emerged over nearly six hours of testimony.

‘We may have run out of options’

During the hearing, a number of Democratic lawmakers argued that Washington needs to take steps to ensure a fair market. Rep. Jim Sensenbrenner (R-WI) insisted that existing antitrust law is fine but needs enforcement. Reback pointed out that convoluted antitrust law makes it difficult to bring cases to court. He said Sensenbrenner was correct in the sense that new legislation might not have been necessary if action had been taken earlier, but he stressed that there isn’t much antitrust enforcement happening today.

Reback blamed the outsized growth of tech giants in part on the Obama administration’s failure to take action. Nobody who understands antitrust law would suggest implementing restrictions cavalierly, Reback said, but regulators have already waited so long that it’s tough to see any viable alternative to reform.

“By not doing anything for so long, we may have run out of options,” Reback said. “The way anticompetitive practices work, if you deal with them quickly, if you deal with them in a rifle shot, you fix it and the companies go on competing and everything’s fine. But if you don’t fix it, then market power builds up and builds up and, in the case we have here, there are no competitors basically for at least three of these companies. They bought their competitors, put them out of business, whatever. And so that then is a problem in terms of how you expect a free market to fix this.”

For context: Apple reported nearly $ 60 billion in profits and became the world’s most valuable company on Friday. On Thursday, Amazon reported revenue up 40% to $ 88.9 billion, while Google’s parent company Alphabet and Facebook reported revenue above analysts’ estimates.

Each of these companies has been accused of anticompetitive practices that harm democracy, consumers, and small businesses. And each enjoys majority control in a number of industry verticals in the U.S. (and much of the world). Areas of dominance include:

  • Facebook’s control of social media platforms Facebook and Instagram
  • Apple and Google’s control of mobile app markets
  • Google’s control in search
  • Facebook and Google’s control of online advertising
  • Amazon’s online shopping platform, which records nearly 75% of online sales
  • Facebook’s and Google’s control of digital advertising, which has hurt journalism industry revenue

Stacy Mitchell is codirector of the Institute for Local Self-Reliance (ILSR), a nonprofit organization that champions distributed, local control over corporate power. She found the committee to be very knowledgeable about how the four companies conduct business and argued that Amazon has a monopolistic hold on small businesses. Last week, Mitchell resigned her position as a fellow at the Yale University Thurmond Arnold Project, a group studying antitrust issues, after finding out that director Fiona Scott Morton is a paid advisor to Amazon and Apple. ILSR also researches Amazon’s anticompetitive practices and advocates breaking the ecommerce giant into separate companies.

Mitchell watched the hearing with interest.

“I found it thrilling, really,” she said. “Like, ‘Oh, this is what it looks like when a democratic government actually addresses fundamental issues of power and control and … takes an aggressive stance toward CEOs and really in some ways alters the power dynamics.’”

She said an exchange between Rep. Mary Gay Scanlon (D-PA) and Jeff Bezos during this week’s antitrust hearing demonstrated the need for this kind of intervention. Under questioning, Bezos confirmed that the Buy Box algorithm favors products shipped with Prime and Amazon fulfillment services. Bezos also said Amazon ties the use of its fulfillment services to winning the Buy Box. Mitchell highlighted the obvious pressure this puts on sellers. “This is a pretty serious matter because effectively what Amazon has done is it has compelled sellers to use Amazon’s fulfillment services in order to get the kind of placement on the site that actually results in sales.”

Scanlon also cited a report ILSR released last week that found Amazon pockets about 30% of sales from independent sellers, up from 19% five years ago. The fee brought in nearly twice the revenue Amazon Web Services did in 2019, and Mitchell said Amazon uses the money to subsidize other company divisions and dominate new industry verticals. That’s part of why ILSR supports breaking major Amazon divisions into individual companies.

“Much of its power, and abuse of its power, comes from it being able to leverage one part of its operation to compel someone to do something in another part of its operation, like we were talking about with the fulfillment. A lot of the abuse lives in the in-between, and I think that’s a pretty powerful case,” Mitchell said.

In testimony Wednesday, Cicilline noted that Amazon internally refers to third-party sellers as “internal competitors.”

Bezos said Wednesday that he “cannot guarantee” Amazon employees have not used data about independent sellers to create the company’s own products, which would be a violation of company policy.  Lawmakers argued this practice undercuts small businesses that have no option but to sell their goods on Amazon’s online marketplace. The Wall Street Journal reported in April that Amazon employees used data about individual sellers to create competing Amazon products.

Anticompetitive behavior accusations have also been lobbed at Amazon by startups who participated in its Alexa Fund, including Nucleus, which says Amazon stole its smart display and home intercom concept. AI startup DefinedCrowd and more than two dozen startup founders detailed similar practices last week. Rep. Joe Neguse (D-CO) referred to this practice as an “innovation kill zone” that puts fear into smaller businesses and makes fair competition impossible.

During the hearing, Apple, Facebook, and Google were also accused of stealing ideas from other companies. (Each of the tech giants has growing investment fund arms.)

“Three of the companies [Amazon, Facebook, and Google] basically have the same set of accusations against them, which is that competitors come on your platform, you take their information, you preference your own results, compete against them, drive them out of business, or buy them cheap,” Reback said.

Big tech too big for free markets?

Both Mitchell and Reback repeatedly emphasized the contrast between the tone of the hearing Wednesday and congressional hearings held in 2018 following the Cambridge Analytica scandal with Mark Zuckerberg. In that earlier hearing, members of Congress demonstrated an inability to understand technology or how Facebook applications work.

What seemed to be missing back then, Mitchell and Reback said — but could be seen in exchanges Wednesday — was the ability to ask followup questions. Without a grasp of how a company operates, or its products and services, any line of questioning will yield little beyond the initial question. CEOs can then run out a portion of the five minutes allotted to each elected official by explaining technology or offering platitudes about how much their companies value things like privacy.

Mitchell said the recent antitrust hearing had a different tone. “This clearly was proceeding as part of an investigation … and sort of had the feel of ‘We’ve gathered all this data or information, and the process is that we’re going to give you a chance to answer for it,’” Mitchell said.

Sally Hubbard is director of enforcement strategy at the Open Markets Institute and testified before a House Judiciary subcommittee on antitrust at its first big tech hearing in June 2019.

Last summer, Hubbard testified that Facebook and Google use their digital advertising monopoly in multiple ways to disadvantage independent journalism and newspapers that are deemed competitors. Both companies directly compete with the free press for attention while controlling the main traffic highways of search engines, smartphones, and social media. She said the hearing Wednesday in part restored her faith in democracy and institutions and gave her hope that meaningful action is possible.

“To me, it was the most impressive hearing that I can remember seeing in my lifetime of the members of Congress being so highly prepared, so in-the-know about complex issues, and willing to take on the most powerful companies in America and the world,” Hubbard said.

Break up big tech?

Hubbard said she could see the rollback of acquisitions like Google’s AdMob and DoubleClick or Facebook’s acquisition of Instagram and WhatsApp as part of potential solutions, but she pointed out that such measures aren’t the only option. She wants lawsuits filed when anticompetitive behavior like the kind described in the hearing occurs, ongoing monitoring of exclusionary conduct, and scrutiny of mergers, especially acquisitions of small companies that may be viewed as a competitive threat. She also recommended passing legislation to support Senator Elizabeth Warren’s (D-MA) suggestion that you can own a platform or sell goods on it, but you can’t do both.

Reback agrees.

“If we can’t get antitrust enforcement on [these companies] … we need legislation like the type Elizabeth Warren is suggesting, where if you own the platform then you can’t own anything on it. That would be a big change, but if there’s no way to police the [current] situation so that competitors get a fair shot and you don’t run everybody else out of business by using their data against them, if there’s no way to police that, then you don’t have any alternative but new legislation,” he said.

On the topic of the digital ad market’s negative impact on journalism in the United States, Rep. Pramila Jayapal (D-WA) talked Wednesday with Google and Alphabet CEO Sundar Pichai about Google’s AdExchange operating in a way that resembles insider trading, adding in the context of independent news outlets.

There’s nothing revolutionary about stopping a single company from controlling both sides of the market to prevent adverse outcomes that resemble insider trading, and it’s past time to bring the same concepts to the digital economy, Hubbard said.

“The point that I like to make a lot is that the online world somehow has avoided regulation, but it’s not the cute little internet of the ’90s anymore. The online world has basically eaten the offline world, so if we don’t have any rules or regulations governing the online world or fair competition, then you don’t have them in the offline world either,” Hubbard said, adding that the hearing will add momentum to other antimonopoly efforts and will help educate the public.

Mitchell concluded that the hearing itself was just one step in the process and the committee’s final report will convey the actual substance of the debate.

Let’s block ads! (Why?)

Big Data – VentureBeat

Read More

Aisha Tyler To Host ‘Prime Rewind: Inside The Boys’ For Amazon Prime- The After-Show Of The Hit ‘The Boys’

July 28, 2020   Humor
 Aisha Tyler To Host ‘Prime Rewind: Inside The Boys’ For Amazon Prime  The After Show Of The Hit ‘The Boys’

“The Boys” will be back once again — Amazon Prime Video has renewed the superhero and vigilante drama for a third season, the streamer announced Thursday.

Additionally, the second season will deliver an after-show entitled “Prime Rewind: Inside The Boys,” hosted and executive produced by Aisha Tyler. While the second season of “The Boys” premieres on Sept. 4, delivering three episodes at launch with the subsequent five dropping weekly through Oct. 9, the after-show will start on Aug. 28 with a look back at the series so far.

“Eric Kripke and the incredible cast of ‘The Boys’ continue to deliver a wild and action-packed series full of surprises and can’t-miss moments that have made the series a global hit,” said Vernon Sanders, co-head of television, Amazon Studios. “We couldn’t be more excited to see where Eric takes ‘The Boys’ and the supes in Season 3, and to have Aisha Tyler join ‘The Boys’ family as the host of ‘Prime Rewind: Inside The Boys.’”

Added Jeffrey Frost, president, Sony Pictures Television Studios: “‘The Boys’ is unlike any other show on television. It’s incredibly entertaining and compelling and elevates genre programming to a whole new level. We’re very excited about Season 3 as we know Eric Kripke and this incredibly talented cast are going to raise the bar even higher. And ‘Prime Rewind: Inside The Boys’ is the perfect show to bring it all home.”

In a statement creator and showrunner Eric Kripke said the writers’ room is already up and running (virtually) for the new season.

“The world has given us way too much material. We hope to be shooting in early 2021, but that’s up to a microscopic virus,” he said.

Addressing the after-show, Kripke noted that the “double entendre” of the title was intended. “It’s a deep dive into how we make this insane thing. Seriously, thanks to Sony, Amazon and the fans. We love making this show so much, and we’re thrilled we get to make more.”

Added Tyler: “‘The Boys’ is one of the smartest, most irreverent, unapologetically badass shows streaming. I became a fan during Season 1 and this season I’m stoked to be flying fans into the heart of the show as host of ‘Prime Rewind: Inside The Boys’ Season 2 is bigger, badder, and more audacious than ever before, so join me after every episode as we dig through the rubble pile of our feelings. I promise, we’ll get through it like ‘The Boys’ — dysfunctional, but together.”

The second season of “The Boys” sees the titular vigilantes on the run and hunted by the superheroes (or “supes,” as the show calls them). Hughie (Jack Quaid), Mother’s Milk (Laz Alonso), Frenchie (Tomer Capon) and Kimiko (Karen Fukuhara) are in hiding after the explosive events of the first season finale, while Butcher (Karl Urban) is off on his own for a bit. Meanwhile, Homelander (Antony Starr) is trying to expand his power in a few new ways and Starlight (Erin Moriarty) is trying to navigate her own place in the Seven as the group shifts. Stormfront (Aya Cash), a social-media savvy new supe, comes in with an agenda of her own, while a larger threat looms, leaving Vought seeking to capitalize on the nation’s paranoia.

“Our show wants you to question superheroes, and therefore fascism and authoritarianism, because they’re closely related,” Kripke tells Variety. “Strongmen tell the public they’re the only ones who can save them. They use the media and their own celebrity to manipulate the masses and shout down any dissenters. I think one downside of the glut of superhero media is that it trains the audience to adulate figures who seem strong, to wait for these figures to solve their problems. When the truth is, anyone who says they can save you is a liar and a con man. Real heroism is a thousand tiny gestures, boring and painful, performed by us all. We have to save ourselves.”

“The Boys” is based on the best-selling comic by Garth Ennis and Darick Robertson, and produced by Amazon Studios and Sony Pictures Television Studios with Point Grey Pictures, Kripke Enterprises and Original Film. Dominique McElligott, Jessie T. Usher, Chace Crawford and Nathan Mitchell also star, with Claudia Doumit, Goran Visnijc, Malcolm Barrett, Colby Minifie, Shantel VanSanten, Cameron Crovetti, PJ Byrne, Laila Robbins and Giancarlo Esposito recurring.

‘Prime Rewind: Inside The Boys” is executive produced by Michael Davies, Tyler, Julia Cassidy, Eden Sutley and Jennifer Ryan. Amazon Studios, Embassy Row and Sony Pictures Television produce.

Source: Variety

Share this:

Like this:

Like Loading…

Damon Williams Lands Unscripted Comedy Stand-Up On New Streaming Service UrbanFlixTV

Let’s block ads! (Why?)

The Humor Mill

Read More

Teradata Strengthens Cloud Offerings on Amazon Web Services and Microsoft Azure

July 11, 2020   BI News and Info
teradata logo social Teradata Strengthens Cloud Offerings on Amazon Web Services and Microsoft Azure

Continued investments in the cloud improve performance, security, availability, operations, ease-of-use and integration
 
As-a-service offerings for Teradata Vantage will soon be available across the top three public cloud providers, including Google Cloud Platform, in 2H 2020

Teradata (NYSE: TDC), the cloud data and analytics company, today announced a myriad of enhancements to its as-a-service offerings for Teradata Vantage on public cloud providers Amazon Web Services (AWS) and Microsoft Azure. The improvements will also apply to Vantage on Google Cloud Platform (GCP), which will begin limited availability in July 2020. The collection of upgrades, as well as the expansion of public cloud offerings, demonstrate Teradata’s ongoing, strong cloud commitment and underscore the company’s critical role in migrating enterprises to the cloud with easy and low-risk hybrid cloud deployments.

Business leaders are increasingly looking to cloud services to access and use software rather than take on the responsibility, cost and effort involved with owning and maintaining IT infrastructure. With the cloud offering new standards of speed, massive (and low cost) storage and scale – in addition to greater flexibility and more expansive service capabilities – it’s easy to understand the cloud’s rapid growth. Teradata enables enterprises to take full advantage of the cloud with quick and low-risk migrations and expansions, offering the industry’s most scalable hybrid cloud and multi-cloud deployments. Because Teradata Vantage software is consistent from one environment to the next, the process is greatly simplified and risk to analytics performance, security, availability and operations is mitigated.

“Teradata is heavily invested in the cloud and the sheer number of improvements we are introducing is a testament to our determination to continue providing customers with the lowest risk, easiest path to the cloud,” said Hillary Ashton, Executive Vice President of Teradata Products. “For many companies, hybrid cloud deployments are an ideal way to benefit from the cloud’s speed and scale without abandoning on-premises investments or regulatory compliance. Hybrid cloud and multi-cloud offerings remain an area of strength for Teradata and these updates advance our cloud offerings across the board.”

With Vantage delivered as-a-service, in the cloud, companies can start small and elastically scale compute and storage independently, and only when needed. The combination of pricing model flexibility with software tier choice enables Teradata’s cloud customers to match features and performance with diverse business requirements. 

Through the enhancements announced today for Vantage on AWS, Azure, and soon also GCP, Teradata cloud customers gain access to the following improvements:

  • Performance: Customers can experience reduced network latency via Teradata’s growing global footprint. With upgrades to compute instances and network performance, Teradata customers will also see improved query response times. 
  • Security: Teradata now offers support for customer-managed keys for both Vantage on AWS and Vantage on Azure. This alternative to the default of Teradata-managed encryption keys increases security options for Teradata customers.
  • Availability: The service level agreement (SLA) for availability is now 99.9% for every as-a-service offering. Guaranteed, higher uptime leads to improved business productivity and increased customer satisfaction.
  • Operations: Quicker compressed data migration times are the result of Teradata’s new data transfer utility (DTU) tool, yielding improved reliability and 20% faster transfers. Improved automation accelerates the submission and resolution of cloud customer requests.
  • Ease of Use: Teradata upgraded the self-service web-based console with expanded options for monitoring and managing as-a-service environments, providing greater ease of use for as-a-service customers.
  • Integration: Through co-development with engineering teams at AWS and Microsoft, Teradata provides greater convenience for customers via the integration of Vantage with multiple first party cloud services, including:  
  • Amazon Kinesis (data streaming);
  • Amazon QuickSight (visualization);
  • Amazon S3 (low-cost object store);
  • Amazon SageMaker (machine learning);
  • AWS Glue (ETL pipeline);
  • Amazon Comprehend Medical (natural language processing);
  • Azure Blob (low-cost object store);
  • Azure Data Factory (ETL pipeline);
  • Azure Databricks (Spark analytics);
  • Azure ML Studio (machine learning);
  • Microsoft Power BI Desktop (visualization); and more to come over time.

Teradata Vantage is the leading hybrid cloud data analytics software platform that enables ecosystem simplification by unifying analytics, data lakes and data warehouses. With Vantage delivered as-a-service, in the cloud, enterprise-scale companies can eliminate silos and cost-effectively query all their data, all the time, regardless of where the data resides – in the cloud, on multiple clouds, on-premises or anywhere in-between – to get a complete, integrated view of their business. Vantage delivered as-a-service subscriptions include Vantage software, high-performance infrastructure, and environment management in one convenient bundle.

Availability
Enhancements to the as-a-service offerings for Vantage on AWS and Vantage on Azure are available now to all Teradata cloud customers on these public cloud providers. Similar enhancements will be available to Teradata customers in the Vantage on GCP limited availability program starting July 2020. 

Let’s block ads! (Why?)

Teradata United States

Read More

Amazon launches monitoring service AWS IoT SiteWise in general availability

July 9, 2020   Big Data
 Amazon launches monitoring service AWS IoT SiteWise in general availability

VB Transform

The AI event for business leaders

Hosted Online

July 14 – 17

Register Today

Amazon today announced that AWS IoT SiteWise, its managed monitoring service for industrial customers, is generally available two years after its unveiling at AWS re:Invent 2018. SiteWise is live for Amazon Web Services (AWS) customers as of this week in the US East (N. Virginia), US West (Oregon), Europe (Frankfurt), and Europe (Ireland) regions, with additional server regions to follow in the coming months.

Manufacturers, energy utilities, and food processors often struggle to collect, process, and analyze data from their hundreds to thousands of internet of things (IoT) sensors. Extracting data across different locations is time-consuming and expensive, because this data is typically stored in local, specialized servers that lack a common data format. Even with a data-collection pipeline in place, information like equipment type, facility location, and relationships with other equipment must be annotated manually; customers have to write custom apps to calculate and compare performance metrics across facilities to derive insights.

SiteWise automates data ingestion from the plant floor, structuring and labeling the data and generating real-time metrics to inform managerial decisions. In SiteWise, customers begin by modeling their industrial equipment, processes, and facilities by adding context to the collected data. They then define common industrial performance metrics (e.g., overall equipment effectiveness and uptime) using SiteWise’s built-in library of mathematical functions. Once a customer’s environment is modeled and their data processed in AWS, the service automatically computes the metrics at an interval defined by the customer. Uploaded data and computed metrics are sent to a fully managed time-series database designed to store and retrieve time-stamped data with low latency.

From within the SiteWise console, customers can create custom web apps without coding to visualize KPIs across end-user devices. And in addition to using software running on an edge device, SiteWise provides interfaces for collecting data from modern industrial apps through MQ Telemetry Transport (MQTT) messages or its APIs.

VB Transform 2020 Online – July 14-17. Join leading AI executives: Last chance to register!

Amazon says that customers like Volkswagen Group are using SiteWise to ingest manufacturing shop floor data into the cloud, model and organize those different machine assets within its plants, and visualize operational data from its cylinder production line in a web application. Another customer — Bayer — deployed SiteWise across nine corn production plants in North America, collecting data from the plant floor and then measuring and analyzing the overall equipment effectiveness of its machinery to identify production inefficiencies. The company claims that SiteWise enables it to onboard a crop site in less than a few hours versus a few weeks.

Amazon maintains pole position in the IoT segment, which is anticipated to be worth $ 212 billion by the end of 2019. According to a survey conducted by the Eclipse Foundation in 2018, AWS was by far the most popular cloud platform for IoT developers, growing in popularity by 21% from 2017 to a 51.8% share, compared with Azure’s 31.21% share (up from 17% in 2017). Amazon CTO Werner Vogels told VentureBeat in a recent interview that AWS customers deploy upwards of hundreds of thousands of sensors.

According to a recent survey published by Tech Pro Research, over 80% of industrial manufacturing companies are using or planning to use IoT devices.

Let’s block ads! (Why?)

Big Data – VentureBeat

Read More

Researchers identify dozens of words that accidentally trigger Amazon Echo speakers

July 6, 2020   Big Data

As voice assistants like Google Assistant and Alexa increasingly make their way into internet of things devices, it’s becoming harder to track when audio recordings are sent to the cloud and who might gain access to them. To spot transgressions, researchers at the University of Darmstadt, North Carolina State University, and the University of Paris Saclay developed LeakyPick, a platform that periodically probes microphone-equipped devices and monitors subsequent network traffic for patterns indicating audio transmission. They say it identified “dozens” of words that accidentally trigger Amazon Echo speakers.

Voice assistant usage might be on the rise — as of 2019, there were an estimated 4.25 billion assistants being used in devices around the world, according to Statista — but privacy concerns haven’t abated. Reporting has revealed that accidental activations have exposed contract workers to private conversations. The risk is such that law firms including Mischon de Reya have advised staff to mute smart speakers when they talk about client matters at home.

LeakyPick is designed to identify hidden voice audio recordings and transmissions as well as to detect potentially compromised devices. The researchers’ prototype, which was built on a Raspberry Pi for less than $ 40, operates by periodically generating audible noises when a user isn’t home and monitoring traffic using a statistical approach that’s applicable to a range of voice-enabled devices.

 Researchers identify dozens of words that accidentally trigger Amazon Echo speakers


VB Transform 2020 Online – July 15-17. Join leading AI executives: Register for the free livestream.

LeakyPick — which the researchers claim is 94% accurate at detecting speech traffic — works for both devices that use a wakeword and those that don’t, like security cameras and smoke alarms. In the case of the former, it’s preconfigured to prefix probes with known wakewords and noises (e.g., “Alexa,” “Hey Google”), and on the network level, it looks for “bursting,” where microphone-enabled devices that don’t typically send much data cause increased network traffic. A statistical probing step serves to filter out cases where bursts result from non-audio transmissions.

To identify words that might mistakenly trigger a voice recording, LeakyPick uses all words in a phoneme dictionary with the same or similar phoneme count compared with actual wakewords. (Phonemes are the perceptually distinct units of sound in a language that distinguish one word from another, such as p, b, d, and t in the English words pad, pat, bad, and bat.) It also verbalizes random words from a simple English word list.

The researchers tested LeakyPick with an Echo Dot, a Google Home, a HomePod, a Netatmo Welcome and Presence, a Nest Protect, and a Hive Hub 360, and Hive View to evaluate its performance. After creating baseline burst and statistical probing data sets, they monitored the eight devices’ live traffic and randomly selected a set of 50 words out of the 1,000 most-used words in the English language combined with a list of known wakewords of voice-activated devices. Then, they had users in three households interact with the three smart speakers — the Echo Dot, HomePod, and Google Home — over a period of 52 days.

The team measured LeakyPick’s accuracy by recording timestamps of when the devices began listening for commands, taking advantage of indicators like the LED ring around the Echo Dot. A light sensor enabled LeakyPick to mark each time the devices were activated, while a 3-Watt speaker connected to the Pi via an amplifier generated sound and a Wi-Fi USB dongle captured network traffic.

In one experiment intended to test LeakyPick’s ability to identify unknown wakewords, the researchers configured the Echo Dot to use the standard “Alexa” wakeword and had LeakyPick play different audio inputs, waiting for two seconds to ensure the smart speaker “heard” the input. According to the researchers, the Echo Dot “reliably” reacted to 89 words across multiple rounds of testing, some of which were phonetically very different than “Alexa,” like “alachah,” “lechner,” and “electrotelegraphic.”

 Researchers identify dozens of words that accidentally trigger Amazon Echo speakers

All 89 words streamed audio recordings to Amazon — findings that aren’t surprising in light of another study identifying 1,000 phrases that incorrectly trigger Alexa-, Siri-, and Google Assistant-powered devices. The coauthors of that paper, which has yet to be published, told Ars Technica the devices in some cases send the audio to remote servers where “more robust” checking mechanisms also mistake the words for wakewords.

“As smart home IoT devices increasingly adopt microphones, there is a growing need for practical privacy defenses,” the LeakyPick creators wrote. “LeakyPick represents a promising approach to mitigate a real threat to smart home privacy.”

Let’s block ads! (Why?)

Big Data – VentureBeat

Read More

Amazon launches AI-powered code review service CodeGuru in general availability

June 30, 2020   Big Data
 Amazon launches AI powered code review service CodeGuru in general availability

Amazon today announced the general availability of CodeGuru, an AI-powered developer tool that provides recommendations for improving code quality. It was first revealed during the company’s Amazon Web Services (AWS) re:Invent 2019 conference in Las Vegas, and starting today, it’s available with usage-based pricing.

Software teams perform code reviews to check the logic, syntax, and style before new code is added to an existing application codebase — it’s an industry-standard practice. But it’s often challenging finding enough developers to perform reviews and monitor the apps post-deployment. Plus, there’s no guarantee those developers won’t miss problems, resulting in bugs and performance issues.

CodeGuru ostensibly solves this with a component that integrates with existing integrated development environments (IDEs) and taps AI algorithms trained on over 10,000 of the most popular open source projects to evaluate code as it’s being written. Where there’s an issue, CodeGuru proffers a human-readable comment that explains what the issue is and suggests potential remediations. The tool also finds the most inefficient and unproductive lines of code by creating a profile that takes into account things like latency and processor utilization.

It’s a two-part system. CodeGuru Reviewer — which uses a combination of rule mining and supervised machine learning models — detects deviation from best practices for using AWS APIs and SDKs, flagging common issues that can lead to production issues such as detection of missing pagination, error handling with batch operations, and the use of classes that are not thread-safe. Developers commit their code as usual to the repository of their choice (e.g. GitHub, GitHub Enterprise, Bitbucket Cloud, and AWS CodeCommit) and add Reviewer as one of the code reviewers. Reviewer then analyzes existing code bases in the repository, identifies bugs and issues, and creates a baseline for successive code reviews by opening a pull request. The service also provides a dashboard that lists information for all code reviews, which reflects feedback solicited from developers.

VB Transform 2020 Online – July 15-17. Join leading AI executives: Register for the free livestream.

CodeGuru Profiler delivers specific recommendations on issues like extravagant recreation of objects, expensive deserialization, usage of inefficient libraries, and excessive logging. Users install an agent in their app that observes the app run time and profiles the app to detect code quality issues (along with details on latency and CPU usage). Profiler then uses machine learning to automatically identify code and anomalous behaviors that are most impacting latency and CPU usage. The information is brought together in a profile that shows the areas of code that are most inefficient. This profile includes recommendations on how developers can fix issues to improve performance and also estimates the cost of continuing to run inefficient code.

Amazon says that CodeGuru — which encodes AWS’ best practices — has been used internally to optimize 80,000 applications, leading to tens of millions of dollars in savings. In fact, Amazon claims that some teams were able to reduce processor utilization by 325% and lower costs by 39% in just a year.

CodeGuru is available now in US East (N. Virginia), US East (Ohio), US West (Oregon), EU (Ireland), EU (London), EU (Frankfurt), EU (Stockholm), Asia Pacific (Singapore), Asia Pacific (Sydney), and Asia Pacific (Tokyo) with availability expanding to additional regions in the coming months. Early adopters include Atlassian, cloud tech consultancy EagleDream Technologies, enterprise software developer DevFactory, condominium review website operator Renga, and scheduling program startup YouCanBook.me.

Let’s block ads! (Why?)

Big Data – VentureBeat

Read More
« Older posts
  • Recent Posts

    • Database version control: Getting started with Flyway
    • Support CRM with New Dynamics 365 Field Service Mobile App
    • 6 Strategies for Achieving Your Business Goals in the New Year
    • Researchers propose using the game Overcooked to benchmark collaborative AI systems
    • Oracle Launches Version 21c
  • Categories

  • Archives

    • January 2021
    • December 2020
    • November 2020
    • October 2020
    • September 2020
    • August 2020
    • July 2020
    • June 2020
    • May 2020
    • April 2020
    • March 2020
    • February 2020
    • January 2020
    • December 2019
    • November 2019
    • October 2019
    • September 2019
    • August 2019
    • July 2019
    • June 2019
    • May 2019
    • April 2019
    • March 2019
    • February 2019
    • January 2019
    • December 2018
    • November 2018
    • October 2018
    • September 2018
    • August 2018
    • July 2018
    • June 2018
    • May 2018
    • April 2018
    • March 2018
    • February 2018
    • January 2018
    • December 2017
    • November 2017
    • October 2017
    • September 2017
    • August 2017
    • July 2017
    • June 2017
    • May 2017
    • April 2017
    • March 2017
    • February 2017
    • January 2017
    • December 2016
    • November 2016
    • October 2016
    • September 2016
    • August 2016
    • July 2016
    • June 2016
    • May 2016
    • April 2016
    • March 2016
    • February 2016
    • January 2016
    • December 2015
    • November 2015
    • October 2015
    • September 2015
    • August 2015
    • July 2015
    • June 2015
    • May 2015
    • April 2015
    • March 2015
    • February 2015
    • January 2015
    • December 2014
    • November 2014
© 2021 Business Intelligence Info
Power BI Training | G Com Solutions Limited