Posted on Leave a comment

Northern Lights partnership is innovating for the future of carbon transport and storage

With the catastrophic effects of climate change knocking at the world’s door, urgent action is needed. Countries and industries are coming together to create economic models that aim to address today’s challenges while providing more sustainable opportunities for growth.

For years Norway’s government has been especially involved in analyzing and investing in forward-thinking technologies and initiatives, not just to address climate problems now, but to envision new industries for the future.

One such project, a 3-year-old partnership called Northern Lights, is a joint effort of the Norwegian government and energy firms Equinor, Shell and Total, each of which has deep roots working with Microsoft. The partnership is seeking to standardize and scale carbon capture and storage, or CCS, across Europe.

CCS has great potential to reduce carbon output, particularly in industries where cutting emissions is more difficult to achieve, and recently the Norwegian government announced an investment proposal in Northern Lights. On Wednesday, Microsoft signed an agreement to explore how it can join the project as a technology partner.

Together, the group will explore how to integrate Microsoft’s digital expertise and work to find ways to invest in the effective development of the project. Microsoft will also look into the of use Northern Lights’ CO2 transport and storage facility as part of its own portfolio of carbon capture, transportation and storage projects.

“This is a challenge that no one government or corporation can solve alone,” says Lucas Joppa, chief environmental officer for Microsoft. “We all need to do more, and those of us who can move faster should. We’re excited by the potential of new approaches like the Northern Lights project. Together with our partners we can work to scale the transportation and storage of captured carbon to help achieve the business needs of a net zero carbon future.”

A permanent storage solution for CO2

Northern Lights has been developing a business model that chains together technologies developed for the energy industry across decades, using them in new ways to provide for the effective transportation, receipt and permanent storage of CO2 in a reservoir in Norway’s North Sea.

According to Irene Rummelhoff, executive vice president of Equinor, the idea is to facilitate the capture and transport of CO2 from industrial emissions, and store it safely without releasing it into the atmosphere. Upon capture, the carbon is liquified and shipped to a Northern Lights facility near Bergen, where it’s pumped 2,600 meters below the sea floor into the pores of a saline aquifer .

“After 40 to 50 years of offshore drilling, instead of taking something out, we are pushing it back down into the earth,” Rummelhoff says. “This project is proving that we can repurpose oil technology in the area of carbon storage in sub-sea reservoirs.”

The plant can initially process up to 1.5 million tons of liquid CO2 each year, and more than 100 million tons over time. The Northern Lights storage facility has gone from an industrial-scale proof of concept to a mature technology, and recently the Norwegian government announced its proposal to invest 16.8 billion Norwegian Krones (1.55 billion Euros) into realizing the Longship CCS value chain, of which Northern Lights is the transport and storage part.

Adding Microsoft brings a technology partner with a global footprint to complement the energy expertise and resources of the other partners. Microsoft’s role is to explore providing a foundation of technology to innovate on, and to work to find ways to help the group further develop the project.

Bringing an ecosystem of innovation to the challenge

In January 2020, Microsoft made a pledge to become carbon negative by 2030 and to remove from the environment all of the carbon it has emitted directly and through electricity consumption since the company was founded by 2050. But company leaders acknowledged that to achieve those goals, they must bet on technologies that have not yet been developed or that are not yet deployed at scale.

Microsoft’s intent is to work with Northern Lights to create a new business ecosystem around carbon management. The company will explore how a software platform based on open-source principles could help foster the technology and business innovation needed to make CCS a reality at an unprecedented scale.

Microsoft can also implement technology in the storage facility itself, using insights and analytics to unlock innovation and blueprint those solutions for use elsewhere in the world. Microsoft’s vast global partner network can be tapped to offer specialized solutions and expertise to fill in the gaps.

“To achieve a net zero carbon future, companies need to be able to transport and store their captured carbon,” Joppa says. “We hope to enable a large-scale value chain and a transport and storage network where there aren’t yet large-scale carbon storage practices in place.”

Exploring the potential of a new industry

Joppa says that joining the project is a win-win situation that reflects Microsoft’s values while serving the common interests of people and countries worldwide. At this initial stage, Microsoft is evaluating what its investment in the initiative might look like, in terms of contributing resources, people, brainpower, research, technology or a combination of all.

“It will take an extraordinary effort to achieve a net zero carbon future,” he says. “We’re going to have to create technologies that don’t exist today at the scale we need them today. We want to explore how we can connect this carbon chain digitally, and we hope to use the Northern Lights project to help reach our own sustainability goals too. We don’t have all the answers at this point, but that is something that we are agreeing to explore together.”

For the world to avoid the catastrophic impacts of climate change, a huge transformation is required for the business value chain itself.

Top photo: The northern lights over Norway’s Lofoten islands. (Photo by Nutexzles/Getty Images)

Posted on Leave a comment

Latest AI breakthrough describes images as well as people do

Novel object captioning

Image captioning is a core challenge in the discipline of computer vision, one that requires an AI system to understand and describe the salient content, or action, in an image, explained Lijuan Wang, a principal research manager in Microsoft’s research lab in Redmond.

“You really need to understand what is going on, you need to know the relationship between objects and actions and you need to summarize and describe it in a natural language sentence,” she said.

Wang led the research team that achieved – and beat – human parity on the novel object captioning at scale, or nocaps, benchmark. The benchmark evaluates AI systems on how well they generate captions for objects in images that are not in the dataset used to train them.

Image captioning systems are typically trained with datasets that contain images paired with sentences that describe the images, essentially a dataset of captioned images.

“The nocaps challenge is really how are you able to describe those novel objects that you haven’t seen in your training data?” Wang said.

To meet the challenge, the Microsoft team pre-trained a large AI model with a rich dataset of images paired with word tags, with each tag mapped to a specific object in an image.

Datasets of images with word tags instead of full captions are more efficient to create, which allowed Wang’s team to feed lots of data into their model. The approach imbued the model with what the team calls a visual vocabulary.

The visual vocabulary pre-training approach, Huang explained, is similar to prepping children to read by first using a picture book that associates individual words with images, such as a picture of an apple with the word “apple” beneath it and a picture of a cat with the word “cat” beneath it.

“This visual vocabulary pre-training essentially is the education needed to train the system; we are trying to educate this motor memory,” Huang said.

The pre-trained model is then fine-tuned for captioning on the dataset of captioned images. In this stage of training, the model learns how to compose a sentence. When presented with an image containing novel objects, the AI system leverages the visual vocabulary to generate an accurate caption.

“It combines what is learned in both the pre-training and the fine-tuning to handle novel objects in the testing,” Wang said.

When evaluated on nocaps, the AI system created captions that were more descriptive and accurate than the captions for the same images that were written by people, according to results presented in a research paper.

Posted on Leave a comment

Join EDUCAUSE virtual conference Oct. 27-29 for the latest higher ed tech solutions

Join Microsoft at EDUCAUSE 2020

This year, the EDUCAUSE Annual Conference takes place online on October 27-29, 2020.

In today’s landscape, it is more important than ever for higher education institutions to reimagine the campus experience. Technology has a critical role in driving this transformation. Please join Microsoft for our online sessions at EDUCAUSE to learn how universities are leading the transformation to recover and reimagine education.

Our sessions showcase how higher education leaders have harnessed technology to innovate across the campus experience. With the use of Microsoft technology, leading universities will share how they are adapting to the changing environment to drive innovative student engagement, transform operations, and ensure a secure, connected campus.

Join us for the following sessions at EDUCAUSE:

table {border-collapse: collapse;}
th, td {border: 1px solid black !important; vertical-align:top; padding: 4px 6px !important;}

Live session

Tuesday, October 27, 12:15 – 1:00 PM EST

Education transformation: From recovery to
reimagine
– presented by Anthony Salcito, Microsoft VP of Education Industry, and featuring University of South Florida

Recent events have dramatically shifted technology’s role in learning, creating a transformation imperative for higher education. Institutions have rapidly transitioned from “Why digital transformation?” to “How much?” and “How quickly?” Join Microsoft Vice President of Education, Anthony Salcito, for a candid exploration of technology’s role in an ever-changing higher education landscape.

[youtube https://www.youtube.com/watch?v=VBGefOJnllc]

Simulive session

Wednesday, October 28, 11:00 – 11:20 AM EST

Security can’t wait: Securing your virtual campus – presented by Micah Linehan, Microsoft Principal Technical Specialist and featuring Kent State University

In the past two years, Kent State has simplified their security architecture and reduced overall IT spend. Learn about Kent State’s sustainable strategy to secure data and protect privacy through modern identity, a modern SIEM, and comprehensive endpoint security to safeguard faculty and student data.

Simulive session

Thursday, October 29, 11:00 – 11:20 AM EST

Reimagine student engagement – presented by Rob Curtin, Director of Americas Higher Education at Microsoft, and featuring Florida State University

Florida State University set out on a student engagement initiative over a year ago, then the pandemic hit. Learn about the Florida State University journey and creation of innovative models to reimagine student engagement and provide a more personalized learning experience in the new world of hybrid education.

For more information, visit our web site Microsoft Higher Education. Learn about campus safety and security, Microsoft Teams, Windows devices for education, and remote learning.

Microsoft Higher EducationMicrosoft Higher Education
Browse affordable devices starting at $219Browse affordable devices starting at $219
Posted on Leave a comment

New nanodegree program offers chance to develop machine learning skills

Earlier this year, we empowered over 10,000 students from all over the world to learn the basics of machine learning over the course of four months. We are excited to announce the next stage of skilling with the availability of an advanced machine learning nanodegree on Udacity. Starting today, students can enroll for the Machine Learning Engineer for Microsoft Azure Nanodegree Program.

This new nanodegree program offers students the opportunity to develop deeper technical skills in machine learning (ML). Students will strengthen their skills by building and deploying sophisticated ML models using Azure Machine Learning. They will learn how to train ML models, manage ML pipelines, and tune hyperparameters to improve model performance. Once the model is ready, students will learn how to operationalize the model with the right MLOps practices, including automation, CI/CD, and monitoring.

Students will get hands-on exposure with built-in Azure labs that are designed to help students put theory into practice, all within Udacity’s classroom environment. To round it up, students will have the opportunity to show off their talents by completing a capstone project based on a real-life data science scenario. By the end of this program, students will also be well-prepared to earn the Azure Data Scientist Associate certification

We also want to congratulate the top 300 students of the introductory ML course who are receiving a scholarship for the Nanodegree program. Here are five such scholars sharing their experiences from the introductory course:

“This is an opportunity to master ML in Azure, get coached by industry experts, and build a solid machine learning portfolio for career advancement. I believe that the scholarship opportunity will bring me a step closer to actualizing my dream,” Ijeoma Ndu said.

Like Ijeoma, many of these students are looking to this nanodegree program to either further their careers or make a career switch. Join our scholarship winners in taking the nanodegree program. Sign up today!

Explore Azure courses on Udacity

Posted on Leave a comment

How fading eyesight led Kenny Singh to embrace tech and become a champion for accessibility

“There’s a lot of granularity in the way I can ‘read’ a screen. I right-click an image and say, ‘OCR it,’” he explains. “Screen readers struggled in the past, but as AI evolves, things are constantly improving. If I am presenting with PowerPoint, I click a button, and it starts transcribing my words. It does a live translation. I can be German if you like!”

Over the horizon, he hopes OCR technology will open doors to even more inclusion. “OCR could be built into Office apps so that text in images with their format can be easily read, copied and edited by a screen reader. OCR technology could also make screen-sharing accessible to screen readers.” He would also like video conferencing technology, such as Microsoft Teams, to be able to recognize and describe to him the facial expressions and emotions of others.

Singh has just been promoted into a new product manager role with responsibility for the performance of Microsoft cybersecurity and compliance products and services. Working closely with industry, academia, and government, he is helping to create a more secure Australia.

“Microsoft Security is a holistic security platform. We have security deeply ‘baked’ into the platform: identity security, endpoint security – including mobile phones, laptops, operational technology, and Internet of Things devices, data security, and security of your apps. All your machines, whether on-premises or in the cloud, are safe – that’s data and emails. We invest more than $1 billion into cybersecurity on a yearly basis globally.”

READ more stories about diversity and inclusion in our region.

In the coming years, he would like to move into even more senior roles “It really gets me going to generate shareholder value, to give people autonomy, mastery and a strong connection with their work.”

He is also mindful of his leadership style. “I have a very strong personal will that’s balanced with humility, empathy, and a growth mindset.”

Man on a exercise machine in a gym
Singh started weight training four years ago.

He also enjoys a work-life balance built around his extended family. He met his partner, Tina, in 1999. They got married in 2006 and now live in suburban Melbourne. Tina is a business transformation leader at the national postal service, Australia Post. “She is also my chauffeur,” he says with a wicked chuckle. Both have parents and other family members living close-by.

About four years ago, he went through another personal transformation: He lost weight, got into shape and started powerlifting.

“I had lost all the fat I wanted to lose, but I then wanted to gain muscle. I started to enjoy powerlifting a lot and seeing my strength increase. It was the measurable performance that really drew me to it,” he says.

“I absolutely love powerlifting from a tenacity perspective. We have very vibrant and full days at Microsoft. At night, I’m thinking, ‘I don’t wanna do this,’ but I always feel great afterward.”

Singh works hard at making progress: whether he is chalking up a new personal best in his home gym or exceeding a business target in the office. His 10th anniversary with Microsoft is coming up soon, and looking back, his nervous first trip to Seattle feels like a lifetime ago.

He has grown professionally, and the attitudes of others toward issues of accessibility have evolved. There is a lot more support, he says. “Now at every event I attend, there’s thought given to captioning. There’s thought given to how to make content more accessible to everyone.

A closeup of a man lifting weights.
“I absolutely love powerlifting from a tenacity perspective.”

“If someone joins (Microsoft) today and they have special needs at work, we have a Special Accommodation Fund. The cost doesn’t even come out of your manager’s fund. There’s no burden on your immediate chain of command; there’s just a box asking what special assistance is required. It takes the trepidation out if it.”

He very much believes in the idea that everyone benefits when designers design with people with disabilities in mind. He champions digital technology as an “intense enabler” for inclusion and accessibility.

“At a people, process and technology level, Microsoft has made a lot of investments that have moved the dial significantly forward. And we’ll keep on pushing for more accessibility for everyone.”

TOP IMAGE: Kenny Singh lifting weights in his garage gym. All images by Penny Stephens. 

Posted on Leave a comment

Play thousands of games from 4 generations of Xbox on Xbox Series X and Xbox Series S

As gamers, we know how important it is to preserve and respect your gaming legacy. We believe your favorite games and franchises, your progression and achievements, your Xbox One gaming accessories and the friendships and communities you create through gaming should all move with you across generations. We also fundamentally believe that not only should you be able to play all of your games from the past without needing to purchase them again, but they should also look, feel and play better on the next generation of Xbox consoles. Preserving and improving the thousands of games you know and love, or have yet to discover, has been a core objective since the beginning of the backward compatibility program in 2015, and with years of learnings and successes – and over 500,000 hours of gameplay testing across four generations of Xbox in the last year alone getting ready for Xbox Series X and Xbox Series S – we couldn’t be more proud of the results you will be able to experience beginning November 10. 

Games play best on Xbox Series X and Series S 

Backward compatible games run natively on the Xbox Series X and S, running with the full power of the CPU, GPU and SSD. No boost mode, no downclocking, the full power of the consoles for each and every backward compatible game. This means that all titles run at the peak performance that they were originally designed for, with significantly higher performance than their original launch platform, resulting in higher and more steady framerates and rendering at their maximum resolution and visual quality. Backward compatible titles also benefit from significant reductions in load times due to the massive leap in performance from our custom NVME SSD at the heart of the Xbox Velocity Architecture

Auto HDR for everyone 

Modern games often implement high dynamic range (HDR) to improve overall visual quality. HDR allows a game to render a much larger range of brightness values and colors. This gives an extra sense of richness and depth to the image when compared to a standard dynamic range (SDR) image. For example, the light from a flashlight looks much brighter and red flowers look much more vibrant. 

However, thousands of Xbox games shipped before HDR was first introduced with Xbox One S, and even for some Xbox One games, developers simply didn’t have the development resources or time to implement HDR. With Xbox Series X and Series S we are introducing a new feature named Auto HDR. Auto HDR automatically adds HDR enhancements to games which only shipped with SDR. Auto HDR enhances the visual quality of an SDR game without changing the original artistic intent of the game. Auto HDR is implemented by the system so developers don’t have to do any work to take advantage of this feature. Also, since Auto HDR is enabled by the console’s hardware, there is absolutely no performance cost to the CPU, GPU or memory and there is no additional latency added ensuring you receive the ultimate gaming experience. 

The below images show how Auto HDR can improve the visual quality without changing the overall look of the game. The HDR images are brighter and more colorful only in the naturally expected areas, while the rest of the image retains its original intent. 

Xbox Series X HDR

Amazing innovations for some of the classics 

By the time the new Xbox consoles launch this November, the team will have spent well over 500,000 hours of gameplay across the entire backward compatibility catalog to ensure your experience is the best possible, no matter which game is your favorite. Beyond the improvements all backward compatible games will see due to increased hardware performance, our team of backward compatibility engineers continue to innovate and push the limits of game preservation and enhancement to make your current game library look and play even better, at no additional cost and with no work from developers while still respecting the artistic intent and vision of the original creators. 

New technology to double framerates 

Xbox Series X and Xbox Series S bring next-gen performance to your favorite games. Higher, steadier framerates make games feel smoother, resulting in more immersive gameplay. Many improvements are the result of the custom designed processor that allows compatible games to play and leverage the increased CPU, GPU and memory from the new consoles. In addition however, the backward compatibility team has developed new methods for effectively doubling the framerate on select titles. While not applicable for many titles due to the game’s original physics or animations, these new techniques the team has developed can push game engines to render more quickly for a buttery smooth experience beyond what the original game might have delivered due to the capabilities of the hardware.  Fallout 4 framerate, shown below, is effectively doubled from 30fps to 60fps on Xbox Series S, delivering a new way to preserve and enjoy this legendary title.   

[youtube https://www.youtube.com/watch?v=6GnvSNA-1DY?version=3&rel=1&fs=1&autohide=2&showsearch=0&showinfo=1&iv_load_policy=1&wmode=transparent&w=640&h=360]

Enhanced visuals you have to see to believe 

On the Xbox One, we were able to provide a curated list of titles that were enhanced for the Xbox One X via the Heutchy method. This allowed titles from Xbox 360 that rendered at 720p and original Xbox games that ran at 360p to play at 4K on Xbox One X, well beyond the capabilities of their original platform. The Heutchy method continues to be used to bring a variety of titles to 1440p on Xbox Series S and 4K on Xbox Series X.   

Xbox Series S BC

Improved texture filtering is also coming to backward compatible titles on both Xbox Series X and Series S. On Xbox One X, a portion of the catalog benefited from increased anisotropic filtering, improving image quality of games. On Xbox Series X and Xbox Series S, 16x anisotropic filtering has been enabled for nearly all backward compatible titles so you can experience the very best visuals the games have to offer.  

Delivering on our promise 

The backward compatibility team has been working hard to make all the games you love and remember available to you on Xbox Series X and Series S. Playing the games will be just as easy and magical as it’s always been. Simply insert your favorite backward compatible Xbox One, Xbox 360 or original Xbox disc into your Xbox Series X, install the game, and you’ll be ready to play. Your digital library will instantly appear and ready for download on the console when you sign in. If you’ve already installed your games to an external drive, you can bring that with you to the Xbox Series X and Xbox Series S. This all happens with no additional cost to you. And, with cloud saves, you’ll jump right back in where you left off.  For those of you still enjoying Xbox 360, cloud saves will soon be free to all Xbox 360 users, making transferring your favorite games to Xbox Series X and Xbox Series S even easier. 

Preserving the games of our past is fundamental to our DNA at Xbox and our continued promise to you. Backward compatibility lets you experience your cherished gaming memories again and in new ways. Your favorite games retain everything the original developer intended, while experiencing enhancements and exciting new features. There’s no need to look back, because your games are ready to move forward with you to the next generation of Xbox. 

Posted on Leave a comment

How Microsoft helped combat ransomware ahead of US elections

Today we took action to disrupt a botnet called Trickbot, one of the world’s most infamous botnets and prolific distributors of ransomware.

As the United States government and independent experts have warned, ransomware is one of the largest threats to the upcoming elections. Adversaries can use ransomware to infect a computer system used to maintain voter rolls or report on election-night results, seizing those systems at a prescribed hour optimized to sow chaos and distrust.

We disrupted Trickbot through a court order we obtained as well as technical action we executed in partnership with telecommunications providers around the world. We have now cut off key infrastructure so those operating Trickbot will no longer be able to initiate new infections or activate ransomware already dropped into computer systems.

In addition to protecting election infrastructure from ransomware attacks, today’s action will protect a wide range of organizations including financial services institutions, government agencies, healthcare facilities, businesses and universities from the various malware infections Trickbot enabled.

The Trickbot botnet

Trickbot has infected over a million computing devices around the world since late 2016. While the exact identity of the operators is unknown, research suggests they serve both nation-states and criminal networks for a variety of objectives.

In the course of Microsoft’s investigation into Trickbot, we analyzed approximately 61,000 samples of Trickbot malware. What makes it so dangerous is that it has modular capabilities that constantly evolve, infecting victims for the operators’ purposes through a “malware-as-a-service” model. Its operators could provide their customers access to infected machines and offer them a delivery mechanism for many forms of malware, including ransomware. Beyond infecting end user computers, Trickbot has also infected a number of “Internet of Things” devices, such as routers, which has extended Trickbot’s reach into households and organizations.

In addition to maintaining modular capabilities for a variety of end purposes, the operators have proven adept at changing techniques based on developments in society. Trickbot’s spam and spear phishing campaigns used to distribute malware have included topics such as Black Lives Matter and COVID-19, enticing people to click on malicious documents or links. Based on the data we see through Microsoft Office 365 Advanced Threat Detection, Trickbot has been the most prolific malware operation using COVID-19 themed lures.

Disruption components and new legal strategy

We took today’s action after the United States District Court for the Eastern District of Virginia granted our request for a court order to halt Trickbot’s operations.

During the investigation that underpinned our case, we were able to identify operational details including the infrastructure Trickbot used to communicate with and control victim computers, the way infected computers talk with each other, and Trickbot’s mechanisms to evade detection and attempts to disrupt its operation. As we observed the infected computers connect to and receive instructions from command and control servers, we were able to identify the precise IP addresses of those servers. With this evidence, the court granted approval for Microsoft and our partners to disable the IP addresses, render the content stored on the command and control servers inaccessible, suspend all services to the botnet operators, and block any effort by the Trickbot operators to purchase or lease additional servers.

To execute this action, Microsoft formed an international group of industry and telecommunications providers. Our Digital Crimes Unit (DCU) led investigation efforts including detection, analysis, telemetry, and reverse engineering, with additional data and insights to strengthen our legal case from a global network of partners including FS-ISACESETLumen’s Black Lotus LabsNTT and Symantec, a division of Broadcom, in addition to our Microsoft Defender team. Further action to remediate victims will be supported by internet service providers (ISPs) and computer emergency readiness teams (CERTs) around the world.

This action also represents a new legal approach that our DCU is using for the first time. Our case includes copyright claims against Trickbot’s malicious use of our software code. This approach is an important development in our efforts to stop the spread of malware, allowing us to take civil action to protect customers in the large number of countries around the world that have these laws in place.

We fully anticipate Trickbot’s operators will make efforts to revive their operations, and we will work with our partners to monitor their activities and take additional legal and technical steps to stop them.

Impact to additional sectors

In addition to its threat to elections, Trickbot is known for using malware to reach online banking websites and steal funds from people and financial institutions. Financial institutions ranging from global banks and payments processors to regional credit unions have been targeted by Trickbot. For this reason, the Financial Services Information Sharing and Analysis Center (FS-ISAC) has been a critical partner and a co-plaintiff in our legal action.

When someone using a Trickbot-infected computer attempts to log onto a financial institutions website, Trickbot executes a series of activities to secretly hijack the user’s web browser, capture the person’s online financial login credentials and other personal information, and send that information to the criminal operators. People are unaware of Trickbot’s activity as the operators have designed it to hide itself. After Trickbot captures login credentials and personal information, operators use that information to access people’s bank accounts. People experience a normal login process and are typically unaware of the underlying surveillance and theft.

Trickbot is also known to deliver the Ryuk crypto-ransomware that has been used in attacks against a wide range of public and private institutions. Ransomware can have devastating effects. Most recently, it crippled the IT network of a German hospital resulting in the death of a woman seeking emergency treatment. Ryuk is a sophisticated crypto-ransomware because it identifies and encrypts network files and disables Windows System Restore to prevent people from being able to recover from the attack without external backups. Ryuk has been attacking organizations, including municipal governments, state courts, hospitals, nursing homes, enterprises and large universities. For example, Ryuk has been attributed to attacks targeting a contractor for the Department of Defense, the North Carolina city of Durham, an IT provider for 110 nursing homes, and a number of hospitals during the COVID-19 pandemic.

Election security and guarding against malware

As we shared last month in the Microsoft Digital Defense Report, ransomware is on the rise. For organizations involved in the elections wanting protection from ransomware and other threats, we offer the threat notification service AccountGuard at no cost which now protects more than two million email accounts around the world. We’ve completed more than 1,500 AccountGuard nation-state attack notifications to AccountGuard enrollees to date. We also offer Microsoft 365 for Campaigns, an easy-to-set-up version of Microsoft 365 that comes with intelligent and secure default settings at an affordable price. Finally, Election Security Advisors provide proactive resiliency services and reactive incident response for campaigns and election officials, also at an affordable price.

Our Digital Crimes Unit will also continue to engage in operations to protect organizations involved in the democratic process and our entire customer base. Since 2010, Microsoft, through the Digital Crimes Unit, has collaborated with law enforcement and other partners on 23 malware and nation-state domain disruptions, resulting in over 500 million devices rescued from cybercriminals. With this civil action, we have leveraged a new legal strategy that allows us to enforce copyright law to prevent Microsoft infrastructure, in this case our software code, from being used to commit crime. As copyright law is more common than computer crime law, this new approach helps us pursue bad actors in more jurisdictions around the world.

To make sure your computer is free of malware, visit support.microsoft.com/botnets.

Tags: , , , ,

Posted on Leave a comment

How a scattered Indian nation kept its songs alive when it couldn’t sing face-to-face

Away from her governance work, Eastwood also has led or participated in many of the virtual cultural sessions. Next to a tabletop in her Anacortes home, she set up a ring light stand, affixed her smart phone to that stand and activated the Teams app, sharing an overhead view of her hands as they created cattail mats and woolen headbands.

Samish citizens who participated in her sessions were shipped boxes of materials (such as dried cattail leaves) to use as they followed Eastwood’s step-by-step instructions from their homes.

Other tribal members have followed her lead. At her home in Seattle, Baker has virtually taught fellow citizens how to use strips of cedar bark to weave a decorative heart and a fish.

And from his place in Anacortes, Wooten has led several remote singing classes, covering “The Bone Game Song” plus Samish flag songs, farewell songs and paddle songs.

A Samish drum rests against a man's left hip as he places a drum stick on the surface.
Tom Wooten plays his traditional Samish drum.

“The songs go way back – before radios and record players,” Wooten says with a smile. “Tribal citizens are hungry to know the past. To move forward, you have to know where you’ve been.

“That platform allows us to reach folks who definitely wouldn’t have been able to come, not just because of COVID, but because our membership is scattered all over the world,” he adds.

Distance is something Samish people have dealt with for generations.

The first bits of archeological evidence linked to the Samish tribe – serrated bison bones and stone butchering tools –are 14,400 years old, carbon dating showed. They were found on Orcas Island, the largest of the San Juan Islands of the Pacific Northwest.

The head and beak of a heron perched in a tree stand out against the orange sun in the sky.
A heron perched above the shoreline in Anacortes.

Since the last Ice Age, the region has served as the traditional Samish homeland. When white settlers arrived there in the late 1800s, they began destroying a large Samish community house. In the ensuing decades, those settlers drove out scores of Samish people, creating a regional diaspora.

During World War II, some of the Samish people who had remained in Anacortes found better-paying work in the airline industry or in shipyards far away, causing the tribe to further separate.

“What made Samish people unique,” Eastwood says, “was we had to dig in and find out how to survive by following opportunities elsewhere but also stay connected with our scattered families.

“Part of our present-day story is based on the fact that the tribe hadn’t ever been given its own reservation,” she adds.

A Samish totem pole stands outside against a smoky sky.
A Samish story pole stands against the sky and trees in Anacortes.

That far-flung citizenry even earned the Samish a legal nickname – one that Eastwood loves.

In 1994, the U.S. Department of Interior conducted a hearing on federal recognition of the Samish as an Indian tribe. Administrative Law Judge David Torbett conducted the hearing.

At that time – the early days of the internet – tribal leaders already were tech adopters, using cell phones, personal email, and faxes to pull together a dispersed people.

Torbett, who ruled in favor of Samish recognition, recognized their technical savvy. In his opinion, he dubbed the Samish the “Cyber Tribe.”

“It gives me chills, even today,” Eastwood says. “Literally, it makes the hair stand up on the back of my neck.”

A woman beads while sitting at a table in front of a laptop.
Leslie Eastwood learns traditional beading via Teams.

That same digital familiarity remains intact. The tribe has a thriving website, Facebook, and Instagram pages. Most citizens are comfortable using smart phones and apps, Eastwood says.

Their tech acumen also led Samish leaders to select a communication platform that ensured only tribal citizens could participate in the virtual sessions – particularly when it came to safeguarding the nation’s business information, Wooten says.

The Samish nation is a member of the Multi-State Information Sharing and Analysis Center (MS-ISAC), which seeks to improve the overall cybersecurity posture of the nation’s state, local, tribal and territorial governments. The MS-ISAC is part of the nonprofit Center for Internet Security (CIS).

“With the internet the way it is today, we were looking for something that had security,” Wooten says.

“It’s been a benefit for folks to interact and know what’s going on. They’re more informed and our participation has gone up during the pandemic,” he adds. “We’re going to continue to utilize Teams well beyond COVID-19. It has saved us time and money and allowed the government to keep working.”

Next to a two-story building, a sign reads
Samish Indian Nation headquarters in Anacortes.

The recommendation to choose Teams over other platforms came from JR Walters, the Samish nation’s IT director. The selection, Walters says, was rooted in frequent headlines about IT breaches and the tribe’s implementation of CIS security controls, a set of cybersecurity best practices.

But it was one early, virtual meeting that Walters never will forget, he says. It took place around the time that the World Health Organization declared COVID-19 a global pandemic in March. Samish leaders began holding daily Teams calls to discuss the nation’s situational awareness.

After several of those meetings, participants began to get a better feel for Teams functions, including custom and blurred backgrounds.

One morning, a Samish leader entered the remote meeting with a new background: The bridge of the “Enterprise,” the spaceship from the TV series “Star Trek.”

“It added relief to a situation that was feeling super stressful,” Walters says. “At the time, we didn’t know what was happening. Schools were closing. We were talking about what we were going to do as a government.

“Then someone just changes their background, and it brings a little joy,” he adds. “It lightened the mood and it made things better.”

Top photo: Tom Wooten, chairman of the Samish Indian Nation, stands along the shoreline in Anacortes, Washington. 

All photos by Dan DeLong.

Posted on Leave a comment

Shrinking the ‘data desert’: making AI systems more inclusive of people with disabilities

“We are in a data desert,” said Mary Bellard, principal innovation architect lead at Microsoft who also oversees the AI for Accessibility program. “There’s a lot of passion and energy around doing really cool things with AI and people with disabilities, but we don’t have enough data.”

“It’s like we have the car and the car is packed and ready to go, but there’s no gas in it. We don’t have enough data to power these ideas.”

To begin to shrink that data desert, Microsoft researchers have been working for the past year and a half to investigate and suggest ways to make AI systems more inclusive of people with disabilities. The company is also funding and collaborating with AI for Accessibility grantees to create or use more representative training datasets, such as ORBIT and the Microsoft Ability Initiative with University of Texas at Austin researchers.

Mary Bellard sits on outdoor steps
Mary Bellard, principal innovation architect lead at Microsoft who oversees the AI for Accessibility program. Photo provided by Bellard.

Today, Team Gleason announced it is partnering with Microsoft on Project Insight, which will create an open dataset of facial imagery of people living with ALS to help advance innovation in computer vision and train those AI models more inclusively.

It’s an industry-wide problem that won’t be solved by one project or organization alone, Microsoft says. But new collaborations are beginning to address the issue.

A research roadmap on AI Fairness and Disability published by Microsoft Research and a workshop on Disability, Bias and AI hosted last year with the AI Now Institute at New York University found a host of potential areas in which mainstream AI algorithms that aren’t trained on inclusive data either don’t work well for people with disabilities or can actively harm them.

If a self-driving car’s pedestrian detection algorithms haven’t been shown examples of people who use wheelchairs or whose posture or gait is different due to advanced age, for example, they may not correctly identify those people as objects to avoid or estimate how much longer they need to safely cross a street, researchers noted.

AI models used in hiring processes that try to read personalities or interpret sentiment from potential job candidates can misread cues and screen out qualified candidates with autism or who emote differently. Algorithms that read handwriting may not be able to cope with examples from people who have Parkinson’s disease or tremors. Gesture recognition systems may be confused by people with amputated limbs or different body shapes.

It’s fairly common for some people with disabilities to be early adopters of intelligent technologies, yet they’ve often not been adequately represented in the data that informs how those systems work, researchers say.

“When technologies are so desired by a community, they’re often willing to tolerate a higher rate of errors,” said Meredith Ringel Morris, senior principal researcher who manages the Microsoft Research Ability Team. “So imperfect AI systems still have value, but they could provide so much more and work so much better if they were trained on more inclusive data.”

‘Pushing the state of the art’

Danna Gurari, an AI for Accessibility grantee and assistant professor at the University of Texas at Austin, had that goal in mind when she began developing the VizWiz datasets. They include tens of thousands of photographs and questions submitted by people who are blind or have low vision to an app originally developed by researchers at Carnegie Mellon University.

The questions run the gamut: What is the expiration date on this milk? What does this shirt say? Do my fingertips look blue? Do these clouds look stormy? Do the charcoal briquettes in this grill look ready? What does the picture on this birthday card look like?

The app originally crowdsourced answers from people across the internet, but Gurari wondered if she could use the data to improve how computer vision algorithms interpret photos taken by people who are blind.

Danna Guarari stands outside
AI for Accessibility grantee Danna Gurari, assistant professor at the University of Texas at Austin who developed the VizWiz dataset and directs the School of Information’s Image and Video Computing Group.

Many of those questions require reading text, such as determining how much of an over-the-counter medicine is safe to take. Computer vision research has often treated that as a separate problem, for example, from recognizing objects or trying to interpret low-quality photos. But successfully describing real-world photos requires an integrated approach, Gurari said.

Moreover, computer vision algorithms typically learn from large image datasets of pictures downloaded from the internet. Most are taken by sighted people and reflect the photographer’s interest, with items that are centered and in focus.

But an algorithm that’s only been trained on perfect images is likely to perform poorly in describing what’s in a photo taken by a person who is blind; it may be blurry, off center or backlit. And sometimes the thing that person wants to know hinges on a detail that a person who is sighted might not think to label, such as whether a shirt is clean or dirty.

“Often it’s not obvious what is meaningful to people, and that’s why it’s so important not just to design for — but design these technologies with — people who are in the blind and low vision community,” said Gurari, who also directs the School of Information’s Image and Video Computing Group at the University of Texas at Austin.

Her team undertook the massive task of cleaning up the original VizWiz dataset to make it usable for training machine learning algorithms — removing inappropriate images, sourcing new labels, scrubbing personal information and even translating audio questions into text to remove the possibility that someone’s voice could be recognized.

Working with Microsoft funding and researchers, Gurari’s team has developed a new public dataset to train, validate and test image captioning algorithms. It includes more than 39,000 images taken by blind and low vision participants and five possible captions for each. Her team is also working on algorithms that can recognize right off the bat when an image someone has submitted is too blurry, obscured or poorly lit and suggest how to try again.

Earlier this year, Microsoft sponsored an open challenge to other industry and academic researchers to test their image captioning algorithms on the VizWiz dataset. In one common evaluation metric, the top performing algorithm posted a 33% improvement over the prior state of the art.

“This is really pushing the state of the art in captioning for the blind community forward,” said Seeing AI lead engineer Shaikh, who is working with AI for Accessibility grantees and their datasets to develop potential improvements for the app.

Posted on Leave a comment

Autonomous systems 101: Optimizing manufacturing processes with AI

This is one part of a four-part introductory series about autonomous systems. To learn more, read the rest of the series and the free e-book linked at the bottom of this post. 

In manufacturing, the name of the game is optimization. With many complex processes working together to build a product or solve a logistical issue, cutting even a little waste can turn into big savings over days, weeks and months.

A major revolution in manufacturing optimization was automation, in which machines carried out the process of building products without as much human intervention. As the manufacturing sector grows, businesses look for more ways to make building products easier and more cost-effective.

That’s where autonomous systems come in. By adding AI to machines, processes or lines, manufacturers have a new weapon in their fight for optimization and efficiency. Engineers are using these new tools that can adjust to their environments and adapt in real time to better meet operational objectives. This stands to disrupt not only how we think about industrial processes, but how humans and machines can work together for maximum impact.

Autonomy, control systems and advancements in AI

Manufacturers know that there are limits to automation—namely, that it’s limited in flexibility and requires high levels of precision and unchanging operating conditions. Think about a modern car assembly line: As a car moves down the line, everything must be just right: location, angle, position. If something is even slightly off, the machines responsible for any aspect of that car can, and probably will, miss their mark (or signal for a human to intervene).

The next step in the evolution of these automated systems is “autonomous” systems. Instead of the automation of a predetermined set of steps, autonomous systems use AI to learn from their environments and engage them dynamically. These machines adopt strategies, rather than rote recipes for action, and then execute them in response to their environment.

The Microsoft approach to autonomous systems

Intelligent, autonomous machines can learn to account for challenges like resource scarcity, managing fail-safe conditions and adaptive object manipulation. To accelerate this learning, Microsoft has pioneered a method that allows control and process experts to teach the AI agent using their own expertise. This allows the AI agent to learn similarly to how a human would: a little at a time, based on a lesson plan from an expert. This includes getting feedback and using that feedback to improve. This AI training methodology allows the most expert engineers to impart their wisdom to the AI agent without needing a data science background, to get the AI agent trained quickly.

The AI agent can then be deployed to assist or advise humans or even work autonomously to optimize a system or process to significantly reduce waste, cost and time.

Read more about autonomous systems: