Earlier in a review, we pitched Bing.com ahead of Google for accurate information in search results, and that's back in January 2016, but come 2018 it seems Bing is getting it all wrong. So what has Microsoft done to its search engine to have resulted these disparate results, that even in areas it has performed remarkably well, that it is now found wanting in accurate local information.

While local search is an interesting area where the search technology must come to bear to be truly relevant, we repeated the minor comparison search with the long-tail keyword for "transfer a domain name to google domains in Nigeria", of which Bing returned the most relevant result in 2016 ahead of Google, but alas Bing results is hugely irrelevant to the search query.

Both Google and Bing are touting of better intelligence of their algorithms, but Google search engine pulls far ahead in this regard, as Bing continues to follow far behind it.

Albeit, Bing had recently improved on natural language search capabilities, which has positively impacted hotel booking online, making available third-party information right on SERP, but its travel search generally seems to lack sufficient data.

Google’s flight portal offers better alternative, and its fully able to provide a ample information on travel related searches, though Bing does quite well in certain flight search recommendations; the best result for a given set of criteria such as searching for flights to New York City simply goes to Google.

Again, in the area of personalized search results, given a user is signed in to the two accounts when conducting a search, Google gives the best search results when searching for branded terminologies and Bing can be found wanting in this regard too.

Bing can be liken to an encyclopedia, with static results and more of responding to the question by further recommendation of related information. Google, On the other hand, goes far ahead, as it provides more accurate results and also giving further richer data through the sidebar area of the main search results.

Microsoft clearly have an uphill task in repositioning its search engine to better compete with Google, as Bing’s success can only be achievable when the web users are duly satisfied by the results of their searches.

Microsoft's search engine, Bing appears to have lost accuracy altogether

Microsoft is harnessing artificial intelligence research with computational power in advanced algorithms to bring about natural interactions between people and machines to extend creative skills.

While Microsoft research aims to infuse computers and systems with the ability to reason, communicate and perform skills akin to humans, with its deep investments in the AI field advancing the state of machine intelligence, and enabling computers to understand what they see; communicate in natural language, and interact with the environment.

The tools and services that result from this project is woven into existing and new products and made open and accessible to the broader community in a bid to accelerate innovation, democratize AI and solve the world’s most pressing challenges.

Microsoft had recently launched a chatbot, XiaoIce in China, which now boasts of 500 million “friends” with over 16 channels for users to interact with it, including through WeChat and other popular messaging platforms.

XiaoIce, is Chinese for “little Bing” and the chatbot has almost convinced some of its users that it is a human being; having its own TV show, ability to write poetry, and does many interesting things, that has lend it a celebrity status.

Albeit, the skill of being able to conjure Chinese poetry from a couple of pictures may sound rather trivial, XiaoIce as Chinese language conversational AI is also able to interact with users in text conversations, and even call people on their phones.

The chatbot marks a notable leap forward in the capacity of AI to undertake such simple tasks; and as interfacing with the actual environment, it tends to help in improving the machine learning model as indistinguishable from its human counterparts.

Microsoft has clearly demonstrated that its Chatbot has the capability to handle basic voice conversations in Chinese, it’s now only a matter of time before it find its way to other language speaking countries.

How Microsoft is taking AI bots to new heights in creating useful skills

Google is expanding the Firebase Cloud Messaging service to include sending notifications, which feature is currently rolling out to all users, along with deeper integration with Google services and new design for the service.

The in-app notification will allow developers to send contextual messages to users of their apps, while the messaging can be customized to look and feel as the developer deems right, and may only include basic text for iOS users and an additional click_action for Android users.

Firebase Predictions and Google Analytics for Firebase are now integrated for improved Cloud Messaging API, with Predictions support from Google’s AI, and a revamped console for managing projects across the app cycle.

Also, Firebase developers will be able to create JIRA issues based on crash reports in Firebase, thanks to the integration of Atlassian’s JIRA to Firebase, which will be rolling out in the coming week.

And with the Crashlytics support in the Firebase, even as Firebase has become more of a platform, developers will be able to build their app crash reports into workflows.

Firebase Release section now features the live data, albeit the new features which have started rolling out today would be available to all users in the coming week.

Google adds in-app notification to Firebase, JIRA integration, and more nifty features

While Microsoft and Amazon were reported to be working on integrating their respective digital assistants, Alexa and Cortana, now the companies have announced the milestone of the two assistants serving as skills in their respective platforms.

The integration was demonstrated earlier at Microsoft’s Build developer conference earlier in the year, as the company sought feedback from the respective communities before bringing it to live for everyone.

What it means is that users on both platforms can now summon either of the assistant irrespective of platform; and for Amazon customers by the simple phrase, "Alexa, open Cortana” to launch Cortana on any Echo device, and on Windows 10 PCs and other Cortana-powered devices, users can summon Alexa by the phrase, “Hey Cortana, open Alexa” to avail the full functions.

As both Cortana and Alexa will be enabled as a skill on each others platform, it affords users the ability to call upon Cortana from an Echo device and get access to same Windows / Office specific features that are formerly exclusive to the digital assistant.

Albeit, this initial integration is rather too basic, as it assiduously requires that users must summon the respective digital assistant to get through to the specific service.

The cross-platform integration, however holds the possibility that these artificial intelligence agents may someday be smart enough to route requests to the best suitable virtual assistant for a particular task without a specific “Alexa” or “Cortana” command.

Finally, Microsoft and Amazon have integrated their respective digital assistants

Oracle, is playing in an unlikely terrain as the heavily locked-down company has open-sourced its tool created to make the deployment of machine learning models in the cloud easy, GraphPipe.

While popular frameworks like Google's TensorFlow and Amazon's Caffe2 are already leading in the machine learning deployment verticals, GraphPipe aims to make it easier to deploy machine learning models for use on mobile apps and IoT devices, and to serve for internal use within organizations.

According to Oracle, GraphPipe is created to solve three challenges: a standard, high-performance protocol for transmitting tensor data over the network, along with simple implementations of clients and servers that make deploying and querying machine learning models from any framework easy.

It can serve models built on TensorFlow, PyTorch, mxnet, CNTK, or caffe2; and developers don’t need custom APIs to deploy AI models or require popular framework to create a model.

The adoption of machine learning in the enterprise has been rather slower than expected, as organizations find it difficult to manage their own machine learning technology, and the models are often deployed using bespoke techniques, which is difficult to manage across servers in different departments.

This challenge is what Oracle is aiming to solve with the new open source, high-performance standard for transmitting tensor data.

GraphPipe is available on GitHub, with documentation, examples, and other relevant content available at this web address: https://oracle.github.io/graphpipe.

GraphPipe, Oracle's open-source standard for deploying Machine Learning models

There's a new report about a wide spread compromise affecting hundreds of users on Instagram, who have reported about their account being hacked.

According to the reports, the victims claim that they were logged out of their account, and their profile details altered by the hackers, with their avatars mostly changed to a Disney or Pixar toons.

As Instagram offers a fun way to explore your creative side in photo shooting and sharing with friends and public alike, it exposes you to a lot of risks around your personal information. The following steps will help you to ensure security for your Instagram account and avoid getting hacked.

Firstly, if you don’t want your account to make the photos you share on the platform public and available to everyone, there's settings for private profile.

Simply go to your Instagram profile window, and click on the three dots in the right corner. Then at the bottom you’ll see the Private Account option, turn this security setting ON, and that's it. Once activated, all media items you share will only been seen by those who follow you online, non will be made available publicly.

Now, back to the business of securing your Instagram account against hacking. The two-factor authentication system offers extra layer of protection to your account, in any case hackers somehow cracks your password, your account will still be inaccessible, as it will need a second factor of authentication before granting them access.

The second factor could be any of these: a one-time code send to your mobile, or a notification to confirm you via Android system. And you’ll be required to confirm it any time you want to login to your Instagram account from a new location or device.

To turn on two-factor authentication from the mobile app, simply click on the Options icon and from the menu on your profile, you’ll see a shortlink for “Two-Factor Authentication” to turn it on and set up using your phone.

It also offer alternative requirements, should you lose your phone or don’t have access to your the mobile number, you can use the Backup Codes. With each Instagram account having five unique codes, which can be accessed in the same settings where you activated the Two-Factor Authentication.

Asides this, it’s still essential to set a strong, unique password for your Instagram account, and to desist from allowing access to apps that are not trustworthy as these apps can have access to your personal information, so could be a potential risk.

Instagram Hack: How you can ensure that your account is safe from hackers

The initial web surfing experience was built on uncharted territory whereby people could spontaneously discover topics online, now Mozilla wants to bring it back to Firefox browser.

The idea is akin to ‘forward button’ which will serve to improve the way content is discovered, and the experiment is dubbed "Advance" - bringing the concept of the recommendation system back to life.

According to Mozilla, Advance is coming at a point where people no longer go backwards in search to move forward to discover new, and relevant content.

The feature brings a two-part recommendations of "Read Next" and "For You" to Firefox window, with the later been recommendation of pages based on a user's browsing history, while the former highlights sites that complement the content of the user's current tab.

Mozilla hopes to recapture those serependitious moment of discovery, which opens people’s eyes to greater awareness of topics they sought after and the idea of a ‘forward button’ is to improve the way content is discovered.

The recommendations will be purely driven by relevance, which is the primary goal of the experiment to give users the best and most timely information.

Albeit, Advance isn't Mozilla's first attempt to bring browsing recommendations to Firefox, as the initial Test Pilot in Context Graph feature, resulted the "Activity Stream" feature, which made debut with Firefox Quantum browser that shipped in November last year.

Mozilla's plan to return the concept of recommendation system to Firefox browser

DeepMind, is an artificial intelligence research firm owned by Google's parent company, Alphabet, and self-described as on a mission to push the boundaries of artificial intelligence (AI) in developing programs that can learn to solve complex problems.

It's specialty ranges from climate change to the need for radical improvements in healthcare; while many such problems are saddled by slow progress, with AI as a multiplier for human ingenuity, the solutions may become more feasible.

The science journal Nature reports that DeepMind, in partnership with Moorfields Eye Hospital in UK, has developed AI that can detect over 50 sight-threatening conditions with the same accuracy as expert clinicians.

DeapMind AI is also quite capable of correctly recommending the appropriate course of treatment for patients and singling out the conditions that require urgent care.

While DeepMind's AI was trained using a particular type of eye scanner, the researchers claim it is compatible with any model and can be used without any hardware restrictions.

According to the report, Google's DeepMind AI diagnoses were accurate for about 94.5 per cent of the time in a trial with Moorfields Eye Hospital, and going by these feats, it could potentially be used to transform how eye examinations are carried out around the world.

And perhaps, DeepMind will eventually cut down the amount of time spent by doctors in diagnosing with OCT scans, being able to recognize 50 common eye problems, which includes three of the most critical: age-related macular degeneration, glaucoma and diabetic retinopathy.

Google's DeepMind AI is now able to diagnose several eye problems

Facebook has introduced extra layer of authorization for Pages with large US audiences, following its influence in the last U.S. election, with Russian meddlers buying ads that focused on divisive social and political rhetoric.

Now, If you're an admin of a Facebook Page that has a large US audience, you will be required to complete extra authorization process to be able to post and manage the page; to ensure such pages are managed by real people, and not imposters or fake accounts.

Under the new authorization, administrators of Facebook pages will have to enable two-factor authentication in order to secure their account and also confirm their primary home address.

And Facebook will show more information about Pages as part of its effort to stamp out fake news; while ensuring authentic information on its platform, the company will also inform followers when a Page they follow has merged with another Page.

With more details in the Info and Ads section of Pages, Facebook will show the primary country locations where Pages are managed and these requirement will also be enforced in its other platforms such as Instagram.

Albeit, these changes will only apply to Pages with a large US audience, but there's the likelihood that Facebook may roll it out to the rest of the world soon.

Facebook's extra layer of Authorization for Pages with large US audience

Kryptowire, a US based security firm has uncovered 38 different vulnerabilities in some Android devices shipped by popular OEMs that can allow for spying and malware reload on the affected devices.

While the US military vetted security firm, Kryptowire specializes in mobile security analysis tools, app marketplace analytics, anti-piracy technologies and Enterprise Mobility Management (EMM) solutions, and was jumpstarted by the DARPA (Defense Advanced Research Projects Agency ) and the Department of Homeland Security.

The vulnerabilities are such that hackers can leverage it to take screenshots, factory reset a device or steal personal information and potentially access the logs of what a person is typing or reading at any given time through the malicious applications.

It's quite alarming that lots of Android users fall victim to malicious apps that pose as harmless tools such as flashlight or battery savers, albeit these apps can't get access to protected files themselves, but they can leverage the flaws in pre-installed applications for access.

According to the researchers, the vulnerabilities were tied to devices from ZTE, Asus, LG and Essential Phone, which are distributed by carriers like Verizon and AT&T in the United States.

These malware are supposedly bundled out of the box, not that it is downloaded and installed by the user, nor is it a flaw in the Android operating system itself, but third party code running on the devices.

Since there's so many different Android phone makers involved, the vulnerabilities are different across the devices, as they come with different pre-installed apps, making it pretty hard for Google to keep track of all the pre-installed malicious apps.

The researchers, however noted that there could be more infected devices out there, outside the named OEMs, considering that not all Android devices were tested, and given that over 24,000 different types of devices were logged in 2015 alone; to run vulnerability scans on every single device will indeed be a herculean task.

How Android devices from popular OEMs come pre-installed with malware