Home Blog

Fb admits to Messenger Children safety flaw however insists it’s mounted

" "

Fb missed a troubling design flaw in its Messenger Children app that meant youngsters might have communicated with customers who hadn’t been accepted by their dad and mom.

The social networking big launched the app in 2017, touting it as a manner for kids beneath 13 to “safely video chat and message with household and mates.” Dad and mom arrange Messenger Children by authorizing it by way of their very own Fb account after which choosing the customers with whom they’re glad for his or her youngster to attach.

However this safety in some way wasn’t in place for group chats, The Verge found, which means youngsters had been in a position to talk with customers who hadn’t been parent-approved.

" "

Fb stated on Monday that it’s been sending out “hundreds” of alerts to oldsters over the previous week, explaining that it’s conscious of the safety flaw and has closed down affected group chats.

“We not too long ago notified some dad and mom of Messenger Children account customers a couple of technical error that we detected affecting a small variety of group chats,” a Fb spokesperson instructed Digital Developments. “We turned off the affected chats and offered dad and mom with further sources on Messenger Children and on-line security.”

" "

Messenger Children solely lets youngsters choose from accepted customers for one-to-one chats. They will additionally enter a chat group that has been began by an accepted person. However till Fb noticed the bug final week, that chat group might include different people accepted by the dad or mum of the person who began the chat group, although not accepted by the dad or mum of the kid getting into that group.

In such circumstances, youngsters might have been speaking to somebody in a chat group whom the dad or mum knew nothing about. Whereas everybody within the group ought to have been accepted by somebody, the flaw is more likely to concern many dad and mom because the app’s safety was not as tight as Fb had claimed.

The social networking big is but to supply extra detailed data on the incident, together with for the way lengthy the difficulty was current within the app.

The unsettling revelation is available in the identical week that the U.S. Federal Commerce Fee is anticipated to announce a colossal $5 billion settlement with Fb over its dealing with of knowledge belonging to 87 million customers caught up within the Cambridge Analytica privateness scandal. Whether or not there will probably be any fallout in response to this newest slip-up stays to be seen.

New Adobe XD instruments allow you to export prototypes to Alexa-enabled units


Final 12 months throughout its Max 2018 convention in Los Angeles, Adobe took the wraps off a brand new model of Adobe XD with enhancements on the voice assistant design entrance and tighter integration with the corporate’s broader Inventive Cloud suite. Constructing on this groundwork, the San Jose firm as we speak unveiled the Amazon Alexa XD plugin, which might export prototypes on to Alexa units. Moreover, it launched the Adobe XD ability for Alexa, which allows designers to entry mentioned prototypes by way of Alexa-enabled good shows and audio system just like the Amazon Echo Dot and Lenovo Sensible Show.

“Right this moment, we’re excited to introduce a brand new solution to prototype voice experiences utilizing Adobe XD and Amazon Alexa. These … instruments allow you to export and preview voice prototypes on any Alexa-enabled gadget,” wrote Adobe director of product voice UI and UX Mark Webster in a weblog publish. “As voice turns into much more ubiquitous, designers want extra methods to include voice into their processes to create experiences that leverage this new medium.”

The Amazon Alexa XD plugin enables you to preview Adobe XD works-in-progress on an Alexa gadget, whereas the Adobe Ability — which Adobe notes is its first Alexa ability to market — permits you to work together with them. The previous is basically a plug-and-play affair, however the latter requires first guaranteeing the Amazon Alexa plugin is enabled (by way of the XD Plugin Supervisor) after which activating the Adobe XD ability by means of the Alexa Expertise Retailer on-line or by means of the Alexa companion app. Full these steps, although, and also you’ll be golden.

“It’s our objective to offer designers, builders, and creators with the assist they should construct nice voice experiences with the instruments they’re conversant in,” mentioned Amazon head of voice design training Paul Cutsinger. “We’re excited to work with Adobe to assist Adobe XD designers extra simply prototype their Alexa ability experiences, and provides their imaginative and prescient a voice.”

The newest launch of Adobe XD helps voice prototyping, with a collection of instruments aimed toward streamlining improvement for in-car navigation techniques, cellular units, and clever assistants resembling Amazon’s Alexa, Microsoft’s Cortana, and the Google Assistant. It enhances the launch of the primary Adobe UI package created particularly for Amazon’s not too long ago introduced Alexa Presentation Language (APL) framework, which includes a variety of kinds, typography, colours, and parts required for visible Alexa apps operating on good shows like Amazon’s Echo Present and Echo Spot.

APL, for the uninitiated, was unveiled in September alongside 11 new and refreshed Alexa-enabled units. It allows builders to construct expertise with textual content, audio, video, and HTML5 that join with TVs utilizing Fireplace TV, Fireplace tablets, Echo Present, and Echo Spot.

Witnesses say damaged wind turbine triggered a number of hundred acre fireplace



by Megan Magensky  Sunday, July 21st 2019

KLICKITAT COUNTY, Wash. — KLICKITAT COUNTY, Wash. — The Juniper Hearth is now 40 % contained, based on a press launch. The fireplace is generally burning within the Pine Creek Drainage space south of Bickleton, WA.

As of Sunday night, the fireplace has burned 242 acres.

39 constructions are threatened by the fireplace however no constructions have burned. The Pine Creek Drainage space is underneath a degree three evacuation.

Crews combating the fireplace are up in opposition to 28 mph wind gusts and low humidity.

201 whole individuals are working the fireplace. Sources will proceed to reach in a single day, based on a press launch.

Officers stated the reason for the fireplace continues to be underneath investigation.

Full article right here.

Video from Youtube account The MustangFan13

(sorry in regards to the music)

From Oregon Stay

Wildfire in Southern Washington attributable to wind turbine that caught fireplace

Up to date Jul 22, 11:35 AM; Posted Jul 21, 2019

By Peter Talbot | The Oregonian/OregonLive

UPDATED Monday, July 22: S. Washington blaze attributable to wind turbine that caught fireplace now 30% contained


A wildfire in southern Washington that has burned greater than 350 acres was attributable to melting sections of a wind turbine that fell to the bottom after the turbine’s generator caught fireplace, fireplace officers stated.

Round 1:40 p.m. Saturday, smoke was reported to be coming from the generator of a wind turbine south of Bickleton, Washington in Klickitat County, based on a brand new launch from Klickitat County Emergency Administration.

Hearth engulfed the turbine 300 toes above the bottom, inflicting melted items to fall to the bottom, igniting grass and brush, based on the discharge. Gusting winds helped unfold the wildfire, referred to as the Juniper Hearth, to between 350 and 500 acres by Saturday night. By 10 p.m. the fireplace was estimated to be about 20% contained.

Three residences have been placed on degree three evacuation notification by the Klickitat County Sheriff’s Workplace. Firefighters have been stationed on the residences in a single day.

Full story right here

HT/Steve Oregon, and MCR

Like this:

Like Loading…

The Apple Watch Sequence four receives a uncommon low cost on Amazon


Final 12 months noticed the discharge of the superior Apple Watch Sequence four, and though offers on these have been fleeting, Amazon has each the 40mm and 44mm fashions on sale proper now. For those who missed the Prime Day offers, then test these out.

Amazon Alexa scientists increase info retrieval with novel AI coaching approach


The data retrieval fashions underpinning search engines like google and yahoo require reams of manually annotated information in the event that they’re skilled utilizing machine studying. That’s as a result of they have to deal with not solely a variety of queries, however any and all information tangential to those queries. Happily, an method (“Content material-Based mostly Weak Supervision for Advert-Hoc Re-Rating”) detailed by scientists at Amazon’s Alexa division may pave the best way for fashions that require much less guide supervision. These may, in flip, improve the dimensions of coaching information units from tens of 1000’s of entries to thousands and thousands, resulting in better-performing methods sooner or later.

Because the staff explains, AI-based retrieval algorithms are usually skilled on a question and two paperwork: a “related” doc that satisfies a consumer’s seek for info, and a associated however non-relevant one which doesn’t. The paperwork in query are labeled manually by people as both related or non-relevant, and through coaching AI methods study to maximise the distinction between relevancy scores they assign to processed samples.

In contrast, the researchers’ technique leverages the truth that the majority of data retrieval coaching information — information articles and Wikipedia entries — is already related to related texts within the articles and sections they introduce. In different phrases, they hypothesize that headlines and titles can be utilized instead of search strings for coaching functions.

The staff first collected thousands and thousands of document-title pairs from the New York Instances’ on-line repository and from Wikipedia. From every pair, they used the question and the related textual content (together with texts associated to the question however much less related than the related textual content) to coach machine studying fashions. They then tapped a corpus from AOL consisting of buyer queries and search outcomes to determine a baseline, to which they utilized an algorithm that recognized related and non-relevant texts for every question. Lastly, they supplemented the AOL information set with a set of about 25,000 hand-annotated samples and algorithmically chosen samples from the take a look at information.

As a way to suss out the efficacy of their method, the staff skilled AI methods individually on every of the 4 take a look at units — New York Instances, Wikipedia, AOL, and the hand-annotated set — and scored the cumulative relevance of the highest 20 outcomes of every utilizing a metric dubbed “normalized discounted cumulative acquire” (nDCG). They report that, of the baselines, the mixture of the AOL information set and an AI structure known as a position-aware convolutional recurrent relevance community, or PACRR, yielded one of the best outcomes. On the identical system, the New York Instances information set provided a 12% improve in nDCG. And when the system was skilled on examples that had been tough to tell apart from information in a given new goal area, the rating improved by 35%.

“Through the use of our method, one can successfully practice neural rating fashions on new domains with out behavioral information and with solely restricted in-domain information,” wrote the coauthors.

Recent Posts